f\ ^™ F^ A
> c PA
United States
    Proleclton
    Air Quality Criteria for
    Ozone and Related
    Photochemical Oxidants
    (First External Review Draft)
    Volume II of

-------
                                                 EPA/600/R-05/004bA
                                                      January 2005
Air Quality Criteria for Ozone and Related
           Photochemical Oxidants
                    Volume II
         National Center for Environmental Assessment-RTF Office
                Office of Research and Development
               U.S. Environmental Protection Agency
                  Research Triangle Park, NC

-------
                                   DISCLAIMER

     This document is an external review draft for review purposes only and does not
constitute U.S. Environmental Protection Agency policy. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
                                     PREFACE

     National Ambient Air Quality Standards (NAAQS) are promulgated by the United States
Environmental Protection Agency (EPA) to meet requirements set forth in Sections 108 and 109
of the U.S. Clean Air Act (CAA). Sections 108 and 109 require the EPA Administrator (1) to
list widespread air pollutants that reasonably may be expected to endanger public health or
welfare; (2) to issue air quality criteria for them that assess the latest available scientific
information on nature and effects of ambient exposure to them; (3) to set "primary" NAAQS to
protect human health with adequate margin of safety and to set "secondary" NAAQS to protect
against welfare effects (e.g., effects on vegetation, ecosystems, visibility, climate, manmade
materials, etc); and (5) to periodically review and revise,  as appropriate, the criteria and NAAQS
for a given listed pollutant or class of pollutants.
     In 1971, the U.S. Environmental Protection Agency (EPA) promulgated National Ambient
Air Quality Standards (NAAQS) to protect the public  health and welfare from adverse effects of
photochemical oxidants. The EPA promulgates the NAAQS on the basis of scientific
information contained in air quality criteria issued under  Section 108 of the Clean Air Act.
Following the review of criteria as contained in the EPA  document, Air Quality Criteria for
Ozone and other Photochemical Oxidants published in 1978, the chemical designation of the
standards was changed from photochemical oxidants to ozone (O3) in 1979 and a 1-hour O3
NAAQS was set. The 1978 document focused primarily  on the scientific air quality criteria for
O3 and, to a lesser extent, on those for other photochemical oxidants such as hydrogen peroxide
and the peroxyacyl nitrates, as have subsequent revised versions of the ozone document.
     To meet Clean Air Act requirements noted above for periodic review of criteria and
NAAQS, the O3 criteria document, Air Quality Criteria for Ozone and Other Photochemical
                                          Il-ii

-------
Oxidants, was next revised and then released in August 1986; and a supplement, Summary of
Selected New Information on Effects of Ozone on Health and Vegetation, was issued in January
1992. These documents were the basis for a March 1993 decision by EPA that revision of the
existing 1-h NAAQS for O3 was not appropriate at that time. That decision, however, did not
take into account some of the newer scientific data that became available after completion of the
1986 criteria document.  Such literature was assessed in the next periodic revision of the O3 air
quality criteria document, which was completed in 1996 and provided scientific bases supporting
the setting by EPA in 1997 of an 8-h O3 NAAQS that is currently in force together with the 1-h
O3 standard.
     The purpose of this revised air quality criteria document for O3 and related photochemical
oxidants is to critically evaluate and assess the latest scientific information published since that
assessed in the above 1996 Ozone Air Quality Criteria Document (O3 AQCD), with the main
focus being on pertinent new information useful in evaluating health and environmental effects
data associated with ambient air O3 exposures. However, some other scientific data are also
presented and evaluated in order to provide a better understanding of the nature,  sources,
distribution, measurement, and  concentrations of O3 and related photochemical oxidants and
their precursors in the environment. The document assesses pertinent literature available
through 2004.
     The present draft document (dated January 2005) is being released for public comment and
review by the Clean Air Scientific Advisory Committee (CASAC) to obtain comments on the
organization and structure of the document, the issues addressed, the approaches employed in
assessing and interpreting the newly available information on O3 exposures and effects, and the
key findings and conclusions arrived at as a consequence of this assessment.  Public comments
and recommendations will be taken into account making any appropriate further revisions to this
document for incorporation into a Second External Review Draft. That draft will be released for
further public comment and CASAC review before last revisions are made in response and
incorporated into a final version to be completed by early 2006. Evaluations contained in the
present document will be drawn on to provide inputs to associated PM Staff Paper analyses
prepared by EPA's Office of Air Quality Planning and Standards (OAQPS) to pose options for
consideration by the EPA Administrator with regard to proposal and,  ultimately, promulgation of
decisions on potential retention or revision, as appropriate, of the current O3 NAAQS.

-------
     Preparation of this document was coordinated by staff of EPA's National Center for
Environmental Assessment in Research Triangle Park (NCEA-RTP). NCEA-RTP scientific
staff, together with experts from other EPA/ORD laboratories and academia, contributed to
writing of document chapters.  Earlier drafts of document materials were reviewed by non-EPA
experts in peer consultation workshops held by EPA. The document describes the nature,
sources, distribution, measurement, and concentrations of O3  in outdoor (ambient) and indoor
environments.  It also evaluates the latest data on human exposures to ambient O3 and
consequent health effects in exposed human populations, to support decision making regarding
the primary, health-related O3 NAAQS. The document also evaluates ambient O3 environmental
effects on vegetation and ecosystems, man-made materials, and surface level solar UV radiation
flux and global climate change, to support decision making on secondary O3 NAAQS.
     NCEA acknowledges the valuable contributions provided by authors, contributors, and
reviewers and the diligence of its staff and contractors in the preparation of this draft document.
                                         Il-iv

-------
           Air Quality Criteria for Ozone and Related
                   Photochemical Oxidants
                 (First External Review Draft)


                         VOLUME I


Executive Summary	E-l

1.   INTRODUCTION  	1-1

2.   PHYSICS AND CHEMISTRY OF OZONE IN THE ATMOSPHERE  	2-1

    CHAPTER 2 ANNEX (ATMOSPHERIC PHYSICS/CHEMISTRY)	AX2-1

3.   ENVIRONMENTAL CONCENTRATIONS, PATTERNS, AND
    EXPOSURE ESTIMATES	3-1

    CHAPTER 3 ANNEX (AIR QUALITY AND EXPOSURE) 	AX3-1
                             II-v

-------
           Air Quality Criteria for Ozone and Related
                   Photochemical Oxidants
                 (First External Review Draft)
                            (cont'd)

                         VOLUME II
4.   DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY, AND
    ANIMAL-TO-HUMAN EXTRAPOLATION	4-1

    CHAPTER 4 ANNEX (DOSIMETRY)	AX4-1

5.   TOXICOLOGICAL EFFECTS OF OZONE AND RELATED
    PHOTOCHEMICAL OXIDANTS IN LABORATORY ANIMALS
    AND IN VITRO TEST SYSTEMS  	5-1

    CHAPTER 5 ANNEX (ANIMAL TOXICOLOGY)	AX5-1

6.   CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
    RELATED PHOTOCHEMICAL OXIDANTS 	6-1

    CHAPTER 6 ANNEX (CONTROLLED HUMAN EXPOSURE)  	AX6-1

7.   EPIDEMIOLOGICAL STUDIES OF HUMAN HEALTH EFFECTS
    ASSOCIATED WITH AMBIENT OZONE EXPOSURE	7-1

    CHAPTER 7 ANNEX (EPIDEMIOLOGY)	AX7-1

8.   INTEGRATIVE SYNTHESIS	8-1
                             Il-vi

-------
           Air Quality Criteria for Ozone and Related
                   Photochemical Oxidants
                 (First External Review Draft)
                           (cont'd)
                         VOLUME
9.   ENVIRONMENTAL EFFECTS: OZONE EFFECTS ON
    VEGETATION AND ECOSYSTEMS  	9-1

10.  TROPOSPHERIC OZONE EFFECTS ON UV-B FLUX AND
    CLIMATE CHANGE PROCESSES 	10-1

11.  EFFECT OF OZONE ON MAN-MADE MATERIALS	11-1

-------
                                Table of Contents

                                                                             Page

List of Tables	I-xviii
List of Figures 	I-xxi
Authors, Contributors, and Reviewers	  I-xxv
U.S. Environmental Protection Agency Project Team for Development of Air
     Quality Criteria for Ozone and Related Photochemical Oxidants 	I-xxvi
U.S. Environmental Protection Agency Science Advisory Board (SAB)
     Staff Office Clean Air Scientific Advisory Committee (CASAC)
     Ozone Review Panel	  I-xxvii
Abbreviations and Acronyms	I-xxviii

4.     DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY, AND
      ANIMAL-TO-HUMAN EXTRAPOLATION	 4-1
      4.1      INTRODUCTION	 4-1
      4.2      DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT 	 4-2
               4.2.1      Bolus-Response Studies  	 4-3
               4.2.2      General Uptake Studies	 4-5
               4.2.3      Dosimetry Modeling	 4-6
               4.2.4      Summary and Conclusions - Dosimetry 	 4-9
      4.3      SPECIES HOMOLOGY, SENSITIVITY, AND ANIMAL-TO-
               HUMAN EXTRAPOLATION 	  4-11
               4.3.1      Summary and Conclusions:  Species Homology, Sensitivity,
                        and Animal-to-Human Extrapolation  	  4-17
      REFERENCES  	  4-18

AX4.  DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT	  AX4-1
      AX4.1    INTRODUCTION	  AX4-1
      AX4.2   EXPERIMENTAL  OZONE DOSIMETRY INVESTIGATIONS	  AX4-3
               AX4.2.1   Bolus-Response Studies  	  AX4-3
               AX4.2.2   General Uptake Studies	  AX4-9
      AX4.3    DOSIMETRY MODELING  	  AX4-10
      AX4.4   SPECIES HOMOLOGY, SENSITIVITY AND ANIMAL-TO-
               HUMAN EXTRAPOLATION	  AX4-16
      REFERENCES  	  AX4-18

5.     TOXICOLOGICAL EFFECTS OF OZONE AND RELATED PHOTOCHEMICAL
      OXIDANTS IN LABORATORY ANIMALS AND IN VITRO TEST SYSTEMS	 5-1
      5.1      INTRODUCTION	 5-1
      5.2      RESPIRATORY TRACT EFFECTS OF OZONE  	 5-2
               5.2.1      Biochemical Effects	 5-2
                        5.2.1.1      Cellular Targets of O3 Interaction 	 5-2
                        5.2.1.2      Monooxygenases	 5-3
                        5.2.1.3      Antioxidants, Antioxidant Metabolism, and
                                   Mitochondrial Oxygen Consumption	 5-3
                        5.2.1.4      Lipid Metabolism and Content of the Lung	 5-5
                        5.2.1.5      Protein Synthesis  	 5-6

-------
                             Table of Contents
                                  (cont'd)
                                                                              Page
                    5.2.1.6       Gene Expression  	  5-7
                    5.2.1.7       Summary and Conclusions - Biochemical
                                 Effects  	  5-7
         5.2.2      Lung Host Defenses	  5-9
                    5.2.2.1       Clearance	  5-9
                    5.2.2.2       Alveolar Macrophages	  5-10
                    5.2.2.3       Immune System	  5-11
                    5.2.2.4       Interactions with Infectious Microorganisms	  5-14
                    5.2.2.5       Summary and Conclusions - Lung Host
                                 Defenses	  5-15
         5.2.3      Inflammation and Lung Permeability Changes	  5-16
                    5.2.3.1       Time Course of Inflammation and Lung
                                 Permeability Changes	  5-16
                    5.2.3.2       Concentration and Time of Exposure	  5-17
                    5.2.3.3       Susceptibility Factors	  5-17
                    5.2.3.4       Mediators of Inflammatory Response
                                 and Injury	  5-20
                    5.2.3.5       The Role of Nitric Oxide Synthase and
                                 Reactive Nitrogen in Inflammation  	  5-23
                    5.2.3.6       Summary and Conclusions - Inflammation
                                 and Permeability Changes  	  5-24
         5.2.4      Morphological Effects	  5-26
                    5.2.4.1       Short Term Exposure Effects	  5-27
                    5.2.4.2       Summary of Short-Term Morphological Effects  ..  5-30
                    5.2.4.3       Long Term Exposure Effects	  5-30
                    5.2.4.4       Summary and Conclusions - Long-Term
                                 Morphological Effects	  5-34
         5.2.5      Effects on Pulmonary Function	  5-35
                    5.2.5.1       Acute and Short-Term Exposure Effects on
                                 Pulmonary Function	  5-35
                    5.2.5.2       Summary and Conclusions - Short- and
                                 Long-Term Effects on Pulmonary Function 	  5-36
                    5.2.5.3       Ozone Effects on Airway Responsiveness	  5-37
                    5.2.5.4       Summary and Conclusions - Effects on
                                 Airway Responsiveness	  5-46
         5.2.6      Genotoxicity Potential of Ozone	  5-47
                    5.2.6.1       Summary and Conclusions - Genotoxicity
                                 Potential of Ozone	  5-48
5.3       SYSTEMIC EFFECTS OF OZONE EXPOSURE 	  5-48
         5.3.1      Neurobehavioral Effects  	  5-48
         5.3.2      NeuroendocrineEffects	  5-49
         5.3.3      Cardiovascular Effects	  5-50
         5.3.4      Reproductive and Developmental Effects	  5-52
         5.3.5      Effects on the Liver, Spleen, and Thymus	  5-54
                                    Il-ix

-------
                                  Table of Contents
                                        (cont'd)
                                                                                 Page
                5.3.6     Effects on Cutaneous and Ocular Tissues	  5-54
                5.3.7     Summary and Conclusions - Systemic Effects of Ozone  	  5-5 5
       5.4       INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
                POLLUTANTS  	  5-56
                5.4.1     Ozone and Nitrogen Oxides  	  5-5 6
                5.4.2     Ozone and Other Copollutants  	  5-57
                5.4.3     Complex (Multicomponent) Mixtures Containing Ozone	  5-60
                5.4.4     Summary and Conclusions - Interactions of Ozone with
                         other Co-occurring Pollutants	  5-67
       5.5       EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS  	  5-68
                5.5.1     Summary and Conclusions - Effects of Other Photochemical
                         Oxidants	  5-69
                REFERENCES	  5-70

AX5.   ANNEX TO CHAPTER 5 OF OZONE AQCD  	  AX5-1
       AX5.1    INTRODUCTION	  AX5-1
       AX5.2    RESPIRATORY TRACT EFFECTS OF OZONE  	  AX5-1
                AX5.2.1  Biochemical Effects	  AX5-1
                         AX5.2.1.1    Cellular Targets of Ozone Interaction 	  AX5-1
                         AX5.2.1.2    Monooxygenases	  AX5-3
                         AX5.2.1.3    Antioxidants, Antioxidant Metabolism,
                                      and Mitochondrial Oxygen Consumption  	  AX5-5
                         AX5.2.1.4    Lipid Metabolism and Content of
                                      the Lung	  AX5-7
                         AX5.2.1.5    Protein Synthesis  	  AX5-8
                         AX5.2.1.6    Gene Expression  	  AX5-8
                AX5.2.2  Lung Host Defenses	  AX5-9
                         AX5.2.2.1    Clearance	  AX5-9
                         AX5.2.2.2    Alveolar Macrophages  	 AX5-18
                         AX5.2.2.3    Immune System	 AX5-20
                         AX5.2.2.4    Interaction with Infectious Microorganisms  .. AX5-24
                AX5.2.3  Inflammation and Lung Permeability Changes	 AX5-24
                         AX5.2.3.1    Time Course of Inflammation and Lung
                                      Permeability Changes	 AX5-25
                         AX5.2.3.2    Concentration and Time of Exposure	 AX5-34
                         AX5.3.3.3    Susceptibility Factors	 AX5-34
                         AX5.2.3.4    Mediators of Inflammatory Response
                                      and Injury	 AX5-38
                         AX5.2.3.5    Role of Nitric Oxide Synthase and
                                      Reactive Nitrogen in Inflammation 	 AX5-41
                AX5.2.4  Morphological Effects	 AX5-42
                         AX5.2.4.1    Introduction 	 AX5-42
                         AX5.2.4.2    Short-Term Exposure Effects on
                                      Morphology	 AX5-43
                                         II-x

-------
                                 Table of Contents
                                      (cont'd)
                                                                               Page
                         AX5.2.4.3    Long-Term Exposure Effects on
                                     Morphology	 AX5-49
               AX5.2.5   Effects on Pulmonary Function	 AX5-55
                         AX5.2.5.1    Introduction  	 AX5-55
                         AX5.2.5.2    Acute and Short-Term Exposure Effects
                                     on Pulmonary Function 	 AX5-55
                         AX5.2.5.3    Long-Term Exposure Effects on
                                     Pulmonary Function	 AX5-57
                         AX5.3.5.4    Acute and Chronic Exposure Effects on
                                     Airway Responsiveness	 AX5-57
               AX5.2.6   Genotoxicity Potential of Ozone	 AX5-67
      AX5.3   SYSTEMIC EFFECTS OF OZONE EXPOSURE  	 AX5-68
               AX5.3.1   Neurobehavional Effects	 AX5-69
               AX5.3.2   Neuroendocrine Effects	 AX5-77
               AX5.3.3   Cardiovascular Effects	 AX5-78
               AX5.3.4   Reproductive and Developmental Effects	 AX5-79
               AX5.3.5   Effects on the Liver, Spleen, and Thymus	 AX5-80
               AX5.3.6   Ozone Effects on Cutaneous and Ocular Tissues 	 AX5-81
      AX5.4   INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
               POLLUTANTS  	 AX5-82
               AX5.4.1   Ozone and Nitrogen Oxides  	 AX5-83
               AX5.4.2   Ozone and Other Copollutants  	 AX5-85
               AX5.4.3   Complex (Multicomponent) Mixtures Containing Ozone .... AX5-90
      AX5.6   EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS 	 AX5-96
      REFERENCES 	 AX5-98

6.     CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
      RELATED PHOTOCHEMICAL OXIDANTS  	 6-1
      6.1      INTRODUCTION	 6-1
      6.2      PULMONARY FUNCTION EFFECTS  OF OZONE EXPOSURE
               IN HEALTHY SUBJECTS 	 6-3
               6.2.2      Acute Exposure for Up to 2 h 	 6-3
               6.2.3      Prolonged Ozone Exposures	 6-5
                         6.2.3.1       Effect of Exercise Ventilation Rate on FEVj
                                     Response to 6.6 h Ozone Exposure	 6-6
                         6.2.3.2       Exercise Ventilation Rate as a Function of
                                     Body/Lung Size on FEV] Response to 6.6 h
                                     Ozone Exposure	 6-6
                         6.2.3.3       Comparison of 2 h IE to 6.6 h O3 Exposure
                                     Effects on Pulmonary Function	 6-7
               6.2.4      Triangular Ozone Exposures	 6-7
               6.2.5      Mechanisms of Pulmonary Function Responses	 6-8
                         6.2.5.1       Pathophysiologic Mechanisms	 6-10
                         6.2.5.2       Mechanisms at a Cellular and Molecular Level ... 6-12
                                        Il-xi

-------
                                 Table of Contents
                                      (cont'd)
                                                                             Pas
      6.3       SUBJECTS WITH PREEXISTING DISEASE	 6-14
               6.3.1      Subjects with Chronic Obstructive Pulmonary Disease  	 6-14
               6.3.2      Subjects with Asthma 	 6-14
               6.3.3      Subjects with Allergic Rhinitis 	 6-16
               6.3.4      Subjects with Cardiovascular Disease	 6-18
      6.4       INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
               OF RESPONSE 	 6-18
      6.5       FACTORS MODIFYING RESPONSIVENESS TO OZONE  	 6-20
               6.5.1      Influence of Age  	 6-20
               6.5.2      Gender and Hormonal Influences 	 6-21
               6.5.3      Racial, Ethnic and Socioeconomic Status Factors	 6-22
               6.5.4      Influence of Physical Activity	 6-22
               6.5.5      Environmental Factors	 6-23
               6.5.6      Oxidant-Antioxidant Balance  	 6-24
               6.5.7      Genetic and Other Factors	 6-24
      6.6       REPEATED O3 EXPOSURE EFFECTS	 6-25
      6.7       EFFECTS ONEXERCISEPERFORMANCE 	 6-25
      6.8       EFFECTS ON AIRWAY RESPONSIVENESS  	 6-26
      6.9       INFLAMMATION AND HOST DEFENSE EFFECTS	 6-27
               6.9.1      Introduction	 6-27
               6.9.2      Inflammatory Response in the Upper Respiratory Tract	 6-28
               6.9.3      Inflammatory Response in the Lower Respiratory Tract	 6-29
               6.9.4      Effects of Repeated Exposures and Adaptation of Responses  ... 6-34
               6.9.5      Effect of Anti-Inflammatory and other Mitigating Agents	 6-36
               6.9.6      Changes in Host Defense Capability Following
                        Ozone Exposures	 6-36
      6.10     EXTRAPULMONARY EFFECTS OF OZONE	 6-38
      6.11     EFFECTS OF OZONE MIXED WITH OTHER POLLUTANTS  	 6-38
      6.12     CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES	 6-39
               6.12.1     Mobile Laboratory Studies 	 6-40
               6.12.2     Aircraft Cabin Studies	 6-40
      6.13     SUMMARY	 6-41
      REFERENCES  	 6-43

AX6.  CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
      RELATED PHOTOCHEMICAL OXIDANTS  	 AX6-1
      AX6.1    INTRODUCTION	 AX6-1
      AX6.2    PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
               IN HEALTHY SUBJECTS  	 AX6-2
               AX6.2.1   Introduction	 AX6-2
               AX6.2.2   Acute Ozone Exposures for Up to 2 Hours	 AX6-3
               AX6.2.3   Prolonged Ozone Exposures	 AX6-8

-------
                           Table of Contents
                                (cont'd)
                                                                         Page
                  AX6.2.3.1   Effect of Exercise Ventilation Rate on
                              FEVj Response to 6.6 h Ozone Exposure  .... AX6-11
                  AX6.2.3.2   Exercise Ventilation Rate as a Function of
                              Body/Lung Size on FEVj Response to 6.6 h
                              Ozone Exposure	 AX6-12
                  AX6.2.3.3   Comparison of 6.6 h Ozone Exposure
                              Pulmonary Responses to Those Observed in
                              2 h Intermittent Exercise Ozone Exposures .  . . AX6-14
         AX6.2.4  Triangular Ozone Exposures	 AX6-15
         AX6.2.5  Mechanisms of Pulmonary Function Responses	 AX6-18
                  AX6.2.5.1   Pathophysiologic Mechanisms	 AX6-19
                  AX6.2.5.2   Mechanisms at a Cellular and Molecular
                              Level	 AX6-25
AX6.3    PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
         IN SUBJECTS WITH PREEXISTING DISEASE 	 AX6-26
         AX6.3.1  Subjects with Chronic Obstructive Pulmonary Disease  	 AX6-26
         AX6.3.2  Subjects with Asthma  	 AX6-31
         AX6.3.3  Subjects with Allergic Rhinitis 	 AX6-35
         AX6.3.4  Subjects with Cardiovascular Disease	 AX6-37
AX6.4    INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
         OF RESPONSE 	 AX6-38
AX6.5    INFLUENCE OF AGE, GENDER, ETHNIC, ENVIRONMENTAL
         AND OTHER FACTORS 	 AX6-42
         AX6.5.1  Influence of Age 	 AX6-42
         AX6.5.2  Gender and Hormonal Influences 	 AX6-48
         AX6.5.3  Racial, Ethnic and Socioeconomic Status Factors	 AX6-55
         AX6.5.4  Influence of Physical Activity	 AX6-56
         AX6.5.5  Environmental Factors	 AX6-57
         AX6.5.6  Oxidant-Antioxidant Balance  	 AX6-62
         AX6.5.7  Genetic and Other Factors	 AX6-65
AX6.6    REPEATED EXPOSURES TO OZONE	 AX6-66
AX6.7    EFFECTS ON EXERCISE PERFORMANCE 	 AX6-78
         AX6.7.1  Introduction	 AX6-78
         AX6.7.2  Effect on Maximal  Oxygen Uptake	 AX6-78
         AX6.7.3  Effect on Endurance Exercise Performance 	 AX6-80
AX6.8    EFFECTS ON AIRWAY RESPONSIVENESS  	 AX6-82
AX6.9    EFFECTS ON INFLAMMATION AND HOST DEFENSE	 AX6-93
         AX6.9.1  Introduction	 AX6-93
         AX6.9.2  Inflammatory Responses in the Upper Respiratory Tract ... AX6-108
         AX6.9.3  Inflammatory Responses in the Lower Respiratory Tract ... AX6-110
         AX6.9.4  Effects of Repeated Exposures and Adaptation
                  of Responses 	 AX6-115
         AX6.9.5  Effect of Anti-Inflammatory and other Mitigating Agents .. AX6-117

-------
                                  Table of Contents
                                       (cont'd)
                                                                                 Pas
                AX6.9.6  Changes in Host Defense Capability Following
                         Ozone Exposure	 AX6-118
       AX6.10   EXTRAPULMONARY EFFECTS OF OZONE	 AX6-121
       AX6.11   OZONE MIXED WITH OTHER POLLUTANTS  	 AX6-122
                AX6.11.1  Ozone and Sulfur Oxides	 AX6-122
                AX6.11.2  Ozone and Nitrogen-Containing Pollutants	 AX6-128
                AX6.11.3  Ozone and Other Pollutant Mixtures Including
                         Particulate Matter  	 AX6-130
       AX6.12   CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES	 AX6-131
                AX6.12.1  Mobile Laboratory Studies 	 AX6-131
                AX6.12.2  Aircraft Cabin Studies	 AX6-132
       REFERENCES 	 AX6-135

7.      EPIDEMIOLOGICAL STUDIES OF HUMAN HEALTH EFFECTS
       ASSOCIATED WITH AMBIENT OZONE EXPOSURE	  7-1
       7.1       INTRODUCTION	  7-1
                7.1.1     Approach to Identifying O3 Epidemiologic Studies 	  7-2
                7.1.2     Approach to Assessing Epidemiologic Evidence 	  7-2
                7.1.3     Study Designs and Analysis Methods Used to Assess
                         O3 Health Effects	  7-6
                         7.1.3.1      Exposure Assessment in Epidemiologic Studies  ...  7-6
                         7.1.3.2      O3 Exposure Indices Used	  7-7
                         7.1.3.3      Lag of O3 Exposure Used	  7-8
                         7.1.3.4      Model Specification Issues  	  7-9
                         7.1.3.5      Controlling for Temporal Trends and
                                     Meteorologic Effects  	 7-10
                         7.1.3.6      Confounding Effects of Copollutants	 7-12
                         7.1.3.7      Model Uncertainty and Multiple Testing	 7-13
                         7.1.3.8      Impact of GAM Convergence Issue on O3
                                     Risk Estimates  	 7-14
                7.1.4     Approach to Presenting O3 Epidemiologic Evidence 	 7-16
       7.2       FIELD STUDIES ADDRESSING ACUTE EFFECTS OF OZONE	 7-17
                7.2.1     Summary of Key Findings on Field Studies of Acute Effects
                         From the 1996 O3  AQCD	 7-17
                7.2.2     Introduction to Recent Field Studies of Acute O3 Effects	 7-18
                7.2.3     Acute O3 Exposure and Lung Function	 7-19
                         7.2.3.1      Acute O3 Studies with Spirometry (FEVj)	 7-20
                         7.2.3.2      Acute O3 Studies of PEF  	 7-28
                7.2.4     Respiratory Symptoms	7-35
                7.2.5     Acute Airway Inflammation 	 7-41
                7.2.6     Acute O3 Exposure and School Absences	 7-43
                7.2.7     Cardiac Physiologic Endpoints	 7-44
                7.2.8     Summary of Field Studies Assessing Acute O3 Effects	 7-45

-------
                            Table of Contents
                                  (cont'd)
                                                                            Page
7.3       EFFECTS OF OZONE ON DAILY EMERGENCY DEPARTMENT
         VISITS AND HOSPITAL ADMISSIONS  	  7-46
         7.3.1      Summary of Key Findings on Studies of Emergency
                   Department Visits and Hospital Admissions from the
                   1996 O3 AQCD	  7-46
         7.3.2      Review of Recent Studies of Emergency Department Visits
                   for Respiratory Diseases  	  7-47
         7.3.3      Studies of Hospital Admissions for Respiratory Diseases	  7-51
         7.3.4      Association of O3 with Hospital Admissions for
                   Cardiovascular Disease	  7-58
         7.3.5      Summary of Acute O3 Effects on Daily Emergency
                   Department Visits and Hospital Admissions	  7-58
7.4       ACUTE EFFECTS OF OZONE ON MORTALITY	  7-59
         7.4.1      Summary of Key Findings on Acute Effects of O3 on
                   Mortality From the 1996 O3 AQCD 	  7-59
         7.4.2      Introduction to Assessment of Current O3-Mortality Studies  ....  7-60
         7.4.3      Single-Pollutant Model O3-Mortality Risk Estimates	  7-61
         7.4.4      Seasonal Variation in O3-Mortality Risk Estimates	  7-65
         7.4.5      O3-Mortality Risk Estimates Adjusting for PM Exposure	  7-67
         7.4.6      O3 Risk Estimates for Specific Causes of Mortality  	  7-71
         7.4.7      O3-Mortality Risk Estimates for Specific Subpopulations  	  7-72
         7.4.8      Summary of Acute O3 Effects on Mortality 	  7-74
7.5       CHRONIC EFFECTS OF OZONE	  7-75
         7.5.1      Summary of Key Findings on Studies of Health Effects and
                   Chronic O3 Exposure from the 1996 O3 AQCD 	  7-75
         7.5.2      Introduction to Morbidity Effects of Chronic O3 Exposure  	  7-75
         7.5.3      Seasonal O3 Effects on Lung Function 	  7-76
         7.5.4      Chronic O3 Effects on Lung Function	  7-78
         7.5.5      Chronic O3 Exposure and Respiratory Inflammation 	  7-81
         7.5.6      Risk of Asthma Development	  7-82
         7.5.7      Respiratory Effects of Chronic O3 Exposure on Susceptible
                   Populations  	  7-84
         7.5.8      Mortality Effects of Chronic O3 Exposure  	  7-85
         7.5.9      Summary of Chronic O3 Effects on Morbidity  and Mortality ....  7-87
7.6       INTERPRETATIVE ASSESSMENT OF THE EVIDENCE IN
         EPIDEMIOLOGIC STUDIES OF OZONE HEALTH EFFECTS  	  7-87
         7.6.1      Introduction	  7-87
         7.6.2      Exposure Assessment 	  7-88
                   7.6.2.1       Relationship between Ambient Concentrations
                                and Personal Exposure to O3 	  7-89
                   7.6.2.2       Factors Affecting the Relationship between
                                Ambient Concentrations and Personal Exposures
                                to O3	  7-90
                                   II-xv

-------
                                   Table of Contents
                                        (cont'd)
                                                                                  Page
                          7.6.2.3       Assessing Chronic Exposure to O3	  7-92
                7.6.3      O3 Exposure Indices  	  7-92
                7.6.4      Lag Time:  Period between O3 Exposure and Observed
                          Health Effect	  7-94
                7.6.5      Confounding by Temporal Trends and Meteorologic Effects ....  7-99
                          7.6.5.1       Assessment of O3 Effects after Adjusting for
                                      Temporal Trends and Meteorologic Effects	  7-99
                          7.6.5.2       Importance of Season-Specific Estimates of
                                      O3 Health Effects	 7-103
                7.6.6      Assessment of Confounding by Copollutants	 7-110
                          7.6.6.1       Relationship between Personal Exposure to
                                      O3 and Copollutants	 7-111
                          7.6.6.2       Assessment of Confounding Using
                                      Multipollutant Regression Models	 7-112
                7.6.7      Issues of Model Uncertainty and Multiple Hypothesis Testing  . 7-116
                7.6.8      Concentration-Response Function and Threshold	 7-118
                7.6.9      Spatial Variability in O3 Effect	 7-122
                7.6.10     Health Effects of O3 in Susceptible Populations	 7-124
                          7.6.10.1      Health Effects Associated with Ambient
                                      O3 Exposure in Asthmatics	 7-125
                          7.6.10.2      Age-Related Differences in O3 Effects	 7-128
                7.6.11     Summary of Key Findings and Conclusions Derived From
                          O3 Epidemiologic Studies 	 7-130
       REFERENCES 	 7-134

AX7.   STUDIES OF HUMAN HEALTH EFFECTS ASSOCIATED WITH
       AMBIENT OZONE EXPOSURE	 AX7-1

8.      INTEGRATIVE SYNTHESIS  	  8-1
       8.1       INTRODUCTION	  8-1
                8.1.1      Chapter Organization	  8-2
                8.1.2      Current Standards  	  8-3
       8.2       TRENDS IN UNITED STATES OZONE AIR QUALITY	  8-4
                8.2.1      Ozone Concentrations, Patterns	  8-4
                8.2.2      Seasonal Variations	  8-6
                8.2.3      Long-Term Trends	  8-7
                8.2.4      Ozone Interactions with Other Ambient Pollutants	  8-7
       8.3       AMBIENT OZONE EXPOSURE ASSESSMENTS	  8-8
                8.3.1      Personal Exposure	  8-8
                8.3.2      Indoor Concentrations	  8-8
       8.4       SYNTHESIS OF AVAILABLE INFORMATION ON OZONE-
                RELATED HEALTH EFFECTS	  8-9
                8.4.1      Assessment of Epidemiological Evidence	  8-10
                8.4.2      Strength of Epidemiological Associations	  8-11

-------
                              Table of Contents
                                   (cont'd)
                                                                                Page
          8.4.3      Acute Exposure Studies	  8-12
                    8.4.3.1       Panel Studies 	  8-12
                    8.4.3.2       Asthma Panels 	  8-13
                    8.4.3.3       School Absences  	  8-14
                    8.4.3.4       Field Studies on Cardiovascular Effects  	  8-15
          8.4.4      Emergency Department Visits and Hospital Admissions 	  8-15
          8.4.5      Acute Effects of Ozone on Mortality	  8-16
          8.4.6      Chronic Ozone Exposure Studies  	  8-20
          8.4.7      Robustness of Epidemiological Associations  	  8-21
                    8.4.7.1       Exposure Issues:  Ambient versus Personal	  8-21
                    8.4.7.2       Confounding by Temporal Trends and
                                 Meteorologic Effects  	  8-22
                    8.4.7.3       Assessment of Confounding by Copollutants ....  8-22
                    8.4.7.4       Lag Period between Ozone Exposure and
                                 Health Response	  8-23
                    8.4.7.5       Concentration-Response Functions and
                                 Threshold	  8-24
                    8.4.7.6       Summary and Conclusions for Epidemiology
                                 Findings	  8-24
          8.4.8      Integration of Experimental and Epidemiologic Evidence	  8-25
                    8.4.8.1       Background on Cross-Cutting Issues	  8-27
                    8.4.8.2       Approaches to Experimental Evaluation of
                                 Ozone Health Effects	  8-27
                    8.4.8.3       Interspecies Comparison of Experimental
                                 Effects — Dosimetry Considerations	  8-28
                    8.4.8.4       Integrated Critical Analysis of Physiological,
                                 Biochemical Effects	  8-29
          8.4.9      Preexisting Disease as a Potential Risk Factor 	  8-37
          8.4.10     Biological Plausibility and Coherence of Evidence for
                    Adverse Respiratory Health Effects  	  8-38
                    8.4.10.1      Pulmonary Function	  8-39
                    8.4.10.2      Lung Injury, Inflammation, and Host  Defense ....  8-42
          8.4.11     Coherence Between Epidemiological and Experimental
                    Evidence  for Respiratory Health Effects	  8-44
          8.4.12     Summary and Conclusions for Ozone Health Effects	  8-45
REFERENCES                                                                   8-47
_I_\J-J_L J-j_l_VJ-j_l_ >l v_xJ—'k-J  	  OT^/

-------
                                     List of Tables

Number                                                                            Page

AX4-1      New Experimental Human Studies on Ozone Dosimetry 	 AX4-4

AX4-2      New Ozone Dosimetry Model Investigations 	  AX4-12

AX5-1      Effects of Ozone on Lung Monooxygenases	 AX5-4

AX5-2      Effects of Ozone on Lung Host Defenses 	  AX5-10

AX5-3      Effects of Ozone on Lung Permeability and Inflammation	  AX5-26

AX5-4      Inbred Mouse Strain Susceptibility  	  AX5-37

AX5-5      Effects of Ozone on Lung Structure: Short-Term Exposures	  AX5-44

AX5-6      Effects of Ozone on Lung Structure: Long-Term Exposures	  AX5-50

AX5-7      Effects of Ozone on Airway Responsiveness 	  AX5-61

AX5-8      Systemic Effects of Ozone	  AX5-70

AX5-9      Interactions of Ozone With Nitrogen Dioxide	  AX5-84

AX5-10     Interactions Of Ozone With Particles	  AX5-86

AX6-1      Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours
            during Exercise  	 AX6-5

AX6-2      Pulmonary Function Effects after Prolonged Exposures to Ozone  	 AX6-9

AX6-3      Ozone Exposure in Subjects with Preexisting Disease  	  AX6-27

AX6-4      Classification of Asthma Severity  	  AX6-34

AX6-5      Age Differences in Pulmonary Function Responses to Ozone 	  AX6-43

AX6-6      Gender and Hormonal Differences in Pulmonary Function Responses to Ozone AX6-49

AX6-7      Influence of Ethnic, Environmental, and Other Factors	  AX6-59

AX6-8      Changes in Forced Expiratory Volume in One Second After Repeated Daily
            Exposure to Ozone	  AX6-67

AX6-9      Pulmonary Function Effects with Repeated Exposures to Ozone	  AX6-69

-------
                                     List of Tables
                                         (cont'd)

Number                                                                             Page

AX6-10     Ozone Effects on Exercise Performance  	  AX6-79

AX6-11     Airway Responsiveness Following Ozone Exposures	  AX6-84

AX6-12     Studies of Respiratory Tract Inflammatory Effects from Controlled
            Human Exposure to Ozone  	  AX6-95

AX6-13     Studies of Effects on Host Defense, on Drug Effects and Supportive
            In Vitro Studies Relating to Controlled Human Exposure to Ozone	  AX6-102

AX6-14     Ozone Mixed with Other Pollutants	  AX6-123

AX6-15     Acute Effects of Ozone in Ambient Air in Field Studies With a
            Mobile Laboratory	  AX6-133

7-la        Field Studies that Investigated the Association between Acute Ambient
            O3 Exposure and Changes in FEV] 	 7-20

7-lb        Changes in FEVj (95% CI) Associated with Acute Ambient O3 Exposures,

            Ordered by Size of the Estimate	 7-21

7-lc        Cross-day Changes in FEVj Associated with Acute Ambient O3 Exposures,

            Ordered by Size of the Estimate	 7-23

7-2         Estimated O3 External Doses and Changes in FEVj Associated with Acute
            Ambient O3 Exposures in Outdoor Workers and Athletes	 7-26

AX7-1      Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms
            in Field Studies	 AX7-2

AX7-2      Effects of Acute O3 Exposure on Cardiovascular Outcomes in
            Field Studies	  AX7-24

AX7-3      Effects of O3 on Daily Emergency Department Visits	  AX7-28

AX7-4      Effects of O3 on Daily Hospital Admissions 	  AX7-36

AX7-5      Effects of Acute O3 Exposure on Mortality	  AX7-46

AX7-6      Effects of Chronic O3 Exposure on Respiratory Health	  AX7-64

AX7-7      Effects of Chronic O3 Exposure on Mortality 	  AX7-78

-------
                                      List of Tables
                                         (cont'd)

Number                                                                               Page

8-1         Current National Ambient Air Quality Standards (NAAQS) in the
            United States  	 8-3

8-2         Gradation of Individual Responses to Short-Term Ozone Exposure in
            Healthy Persons	 8-4

8-3         Gradation of Individual Responses to Short-Term Ozone Exposure in
            Persons with Impaired Respiratory Systems 	 8-5
                                           II-xx

-------
                                      List of Figures

Number                                                                               Page

AX5-1      Major secondary products of ozone interaction with lung cells	  AX5-2

6-1         Predicted O3-induced decrements in FEVj as a function of exposure
            duration and O3 concentration in young healthy adults (20 yrs of age)
            during moderate IE (VE = 30 L/min)	 6-5

6-2         Recovery of FEVj responses following a 2 h exposure to 0.4 ppm O3 with IE  ...  6-10

6-3         Predicted O3-induced decrements in FEV] as a function of exposure duration
            and level of IE (line labels are VE levels) in young healthy adults (20 yrs of
            age) exposed to 0.3 ppm O3	  6-22

AX6-1      FEV] decrements as a function of O3 concentration following a 2 h exposure
            with incremental exercise (15 min intervals) or rest  	  AX6-7

AX6-2      Average FEVj decrements (±SE) for prolonged 6.6 h exposures to 0.12 ppm
            O3 as a function of exercise VE  	  AX6-13

AX6-3      The forced expiratory volume in 1 s (FEVj) is shown in relation to exposure
            duration (hours) under three exposure conditions	  AX6-16

AX6-4a,b   Recovery of spirometric responses following a 2 h exposure to 0.4 ppm
            O3 with IE	  AX6-20

AX6-5      Plot of the mean FEV] (% baseline) vs. time for ozone exposed cohorts 	  AX6-23

AX6-6      Frequency distributions of percent decrements in FEVj for 6.6-h exposure
            to four concentrations of ozone	  AX6-39

AX6-7      Effect of O3 exposure (0.42 ppm for 1.5 h with IE) on FEV]  as a function
            of subject age	  AX6-47

AX6-8      Regression curves were fitted to day-by-day postexposure FEV] values
            obtained after repeated daily acute exposures to O3 for 2 to 3 h with
            intermittent exercise at a VE  of 24 to 43 L/min (adaptation studies)	  AX6-77

7-la        Percent change (95% CI) in  morning PEF in children per 40 ppb increase
            in 1-h max O3 or equivalent, arranged by size of the effect estimate	  7-29

7-lb        Percent change (95% CI) in  afternoon PEF in children per 40 ppb increase
            in 1-h max O3 or equivalent, arranged by size of the effect estimate	  7-30

7-2         Density curves of the % change in PEF per 30 ppb increase in 8-h avg O3
            with a cumulative lag of 1 to 5 days for the individual eight NCICAS cities
            and the pooled average of all cities 	  7-32

-------
                                       List of Figures

Number                                                                                 Page

7-3         Density curves of the odds ratios for the prevalence of cough among
            asthmatic children  	  7-36

7-4         Density curves of the odds ratios for the prevalence of extra bronchodilator
            use among asthmatic children  	  7-37

7-5         Density curves of the odds ratios for the incidence of symptoms per 30 ppb
            increase in 8-h avg O3 with a cumulative lag of 1 to 4 days for the individual
            eight cities and the pooled average of all cities	  7-38

7-6         Ozone-associated % change (95% CI) in emergency department visits for
            asthma per 40 ppb increase in 1-h max O3 or equivalent	  7-48

7-7         Ozone-associated % change (95% CI) in total respiratory hospitalizations
            for all year analyses per 40 ppb increase in 1-h max O3 or equivalent  	  7-52

7-8         Ozone-associated % change (95% CI) in total respiratory hospitalizations
            by season per 40 ppb increase in 1-h max O3 or equivalent	  7-53

7-9         Ozone-associated % change (95% CI) in total respiratory hospitalizations with
            adjustment for PM indices per 40 ppb increase in 1-h max O3 or equivalent 	  7-54

7-10        All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) for all year
            analyses per 40 ppb increase in 1-h max O3 or equivalent	  7-62

7-11        Bayesian city-specific and national average estimates for the % change
            (95% CI) in daily mortality per 10 ppb increase in 24-h avg O3 in the previous
            week using a constrained distributed lag model  for 95 U.S. communities
            (NMMAPS), arranged by size of the effect estimate	  7-64

7-12        All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) by
            season per 40 ppb increase in 1-h max O3 or equivalent	  7-66

7-13        All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) with
            adjustment for PM indices for all year analyses  per 40 ppb increase in 1-h max
            O3 or equivalent	  7-69

7-14        All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
            with adjustment for PM indices by season per 40 ppb increase in 1-h max
            O3 or equivalent 	  7-70

7-15        Comparison of single-day lags (0-, 1-, 2-, and 3-day) to a cumulative multiday
            lag (0- to 6-day) for % changes in all cause mortality per 20 ppb increase in
            24-h avg O3 in all ages	  7-96

-------
                                       List of Figures

Number                                                                                 Page

7-16         Comparison of single-day lags (0-, 1-, 2-, 3-, 4-, and 5-day) to a cumulative
             multiday lag (0- to 4-day) for % changes in total respiratory hospitalizations
             per 40 ppb increase in 1-h max O3 in children less than two years of age	 7-97

7-17         Comparison of single-day lags (1-, 2-, 3-, 4-, 5-, and 6-day) to a cumulative
             multiday lag (1- to 5-day) for % changes in PEF per 30 ppb increase in 8-h
             avg O3 in urban children	 7-98

7-18         Daily nonaccidental mortality in Montreal, Canada as a function of mean
             temperature, using natural splines with two degrees of freedom	 7-100

7-19         Scatterplots of daily levels of O3 with TSP and SO2 in Philadelphia, PA
             by season  	 7-104

7-20         Estimated total (nonaccidental) mortality relative risk per 100 ppb increase in
             24-h avg O3 of reach seasonal data set	 7-105

7-21         Marginal posterior distributions of the national average estimates of O3
             effects on total mortality per 10 ppb increase in 24-h avg O3 at a 1-day lag
             for all year, summer (June-August), and winter (December-February)
             analyses in 90 U.S. cities	 7-106

7-22         Summary density curves of the % change in all cause mortality for all year
             data and by season per 40 ppb increase in 1-h max O3 or  equivalent 	 7-108

7-23         Summary density curves of the % change in total respiratory hospital
             admissions for all year data and by season per 40 ppb increase in 1-h
             max O3 or equivalent	 7-109

7-24         Posterior means and 95% posterior intervals of the national average estimate
             of PM10 effects on total  mortality from non-external causes per 10 ^ig/m3
             increase in 24-h avg PM10 at a 1-day lag within sets of 90 U.S. cities with
             pollutant data available   	 7-113

7-25         Posterior means and 95% posterior intervals of the national average
             estimate of O3 effects  on total mortality from non-external causes per
             10 ppb increase in 24-h  avg O3 at 0-, 1-, and 2-day lags within sets of
             90 U.S. cities with pollutant data available 	 7-113

7-26         Maximum likelihood  estimates of O3-mortality for 95 U.S. communities,
             determined using a constrained distributed lag model for lags 0 through 6 days  . 7-114

8-1          Ozone-associated percent change  (95% CI) in emergency room visits
             for asthma	 8-16

-------
                                       List of Figures

Number                                                                                 Page

8-2         Ozone associated percent change (95% CI) in total respiratory
            hospitalizations (95% CI) for all year and for by season	  8-17

8-3         All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
            for all year and for by season per 40 ppb increase in 1-h maximum O3
            or equivalent	  8-18

8-4
All cause (nonaccidental) O3 excess mortality risk estimates (95% CI) with
adjustment for PM indices for all analyses  	
                                                                                        8-19
                                           II-xxiv

-------
                        Authors, Contributors, and Reviewers
To be inserted in Second External Review draft.
                                        II-xxv

-------
                U.S. ENVIRONMENTAL PROTECTION AGENCY
       PROJECT TEAM FOR DEVELOPMENT OF AIR QUALITY CRITERIA
                                 FOR OZONE
To be inserted in Second External Review draft.
                                   II-xxvi

-------
               U.S. ENVIRONMENTAL PROTECTION AGENCY
              SCIENCE ADVISORY BOARD (SAB) STAFF OFFICE
          CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE (CASAC) OZONE
                             REVIEW PANEL
To be inserted in Second External Review draft.
                                 II-xxvii

-------
                             Abbreviations and Acronyms
AA
ascorbic acid
ACh
acetylcholine
ADs
alveolar entrance rings
ADSS
aged and diluted sidestream cigarette smoke
AED
aerodynamic diameter
AER
air exchange rate
AF
adsorbed fraction
AH9
ascorbic acid
AHR
airway hyperreactivity
AHSMOG
ALI
Adventist Health Study on Smog
AIRS             Aerometric Information Retrieval System (U.S. Environmental Protection
                  Agency)
air-liquid interface
AM
alveolar macrophage
A,,
cros-sectional area of peripheral lung
AP
alkaline phosphatase
APHEA
Air Pollution on Health: European Approach
AQCD
Air Quality Criteria Document
ATR
atrial natriuretic factor
BAL
bronchoalveolar lavage
BALF
bronchoalveolar lavage fluid
BALT
bronchus-associated lymphoid tissue
B[a]P
benzo[a]pyrene
BHR
bronchial hyperresponsiveness
b.i.d.
twice a day
BMZ
basement membrane zone
BP
blood pressure
BrdU
bromodeoxyuridine

-------
BS
BSA
C
CxT
C3a
CAPs
CAR
CARB
CAT
CC16
CCh
CCSP
Cdyn
CE
CFD
CHO
CI
CINC
CIU, CBU
CMD
CO
C02
ConA
COPD
cw
Cyt
A
DD
DHBA
black smoke
body surface area
concentration
concentration times duration of exposure
complement protein fragment
concentrated ambient particles
centriacinar region
California Air Resources Board
cell antioxidant capacity
Clara cell secretory protein
carbachol
Clara cell secretory protein
dynamic lung compliance
continuous exercise
computational fluid dynamics
Chinese hamster ovary
confidence interval
cytokine-induced neutrophil chemoattractant
cumulative inhalation unit
count mean diameter
carbon monoxide
carbon dioxide
concanavalin A
chronic obstructive pulmonary disease
chest wall compliance
cytochrome
mean change in a variable
doubling dose
2,3 -deny droxybenzoic acid
II-xxix

-------
DNA
deoxyribonucleic acid
DPPC
dipalmitoylglycero-3-phosphocholine
DR
disulfide reductase
                  convergence precision
EEG
electroencephalographic
ELF
epithelial lining fluid
EM
electron microscopy
ENA
epithelial cell-derived neutrophil-activating peptide
EPA
U.S. Environmental Protection Agency
ETS
environmental tobacco smoke
EU
                  frequency of breathing
                  female
F344
Fischer 344
FA
filtered air
FA
fatty acid
FA
fractional absorption; absorbed fraction
FAA
Federal Aviation Administration
FEF
forced expiratory flow
FEF9
    25
forced expiratory flow after 25% vital capacity
FEF9
    25-75
forced expiratory flow between 25 and 75% of vital capacity
FEF,
forced expiratory flow after 50% vital capacity
FEF,
    6oP
FEF75
forced expiratory flow after 75% vital capacity
FEV,
    0.75
forced expiratory volume in 0.75 s
FEV,
forced expiratory volume in 1 s
FFA
free fatty acid
FGF



FGFR
                                         II-XXX

-------
FIVC
forced inspiratory vital capacity
Fn
fibronectin
FP
fluticasone propionate
FRC
functional residual capacity
FS
field stimulation
FVC
forced vital capacity
GAM
General Additive Model
GDI
glutathione-disulfide transhydrogenase
GEE
(model)
GLM
General Linear Model
GM-CSF
granulocyte-macrophage colony stimulating factor
G6PD
glucose-6-phosphate dehydrogenase
GR
glutathione reductase
GSH
glutathione
GSHPx, GPx
glutathione peroxidase
GST
glutathione-^-transferase
                  hydrogen ion
H2CO, HCHO      formaldehyde
HDMA
house dust mite allergen
                  hydrogen peroxide
H2S04
sulfuric acid
HEI
Health Effects Institute
HHP-C9
1 -hy droxy-1 -hy droperoxynonane
HIST
histamine
HLA
human lymphocyte antigen
HNE
4-hydroxynonenal
HR
heart rate
HSP
heat shock protein
HSV
herpes simplex virus
                                        II-xxxi

-------
5-HT
IAS
1C
ICAM
ICRP
ICS
IE
Ig
IL
IN
INF
inh
iNOS
5-hydroxytryptamine
interalveolar septum
inspiratory capacity
intracellular adhesion molecule
International Commission on Radiological Protection
inhaled steroids
intermittent exercise
immunoglobulin (IgA, IgE, IgG, IgM)
interleukin (IL-1, IL-6, IL-8)
intranasal
interferon
inhalation
inducible nitric oxide synthase
ip intraperitoneal
IT
IU
iv
intratracheal
International Units
intravenous
Ka intrinsic mass transfer coefficient/parameter
Kg mass transfer coefficient for gas phase
KTB
K,
tracheobronchial region overall mass transfer coefficient
mass transfer coefficient for liquid phase
Kj reaction rate constant
A
LDH
LIS
LM
LOESS
LPS
LRT
ozone uptake efficiency
lactate dehydrogenase
lateral intercellular space
light microscopy
locally estimated smoothing splines
lipopolysaccharide
lower respiratory tract
II-xxxii

-------
LT
leukotriene (LTB4, LTC4, LTD4, LTE4)
M
male
M
maximum number of iterations
MCB
monochlorobimane
MCh
methacholine
MCP
monocyte chemotactic protein
MDA
malondi aldehyde
MHC
major histocompatibility
MIP
macrophage inflammatory protein
MLN
mediastinal lymph node
MMAD
mass median aerodynamic diameter
mRNA
messenger ribonucleic acid
MSA
metropolitan statistical area
Mt
metallothionein
n,N
number
NAAQS
National Ambient Air Quality Standards
NADH
reduced nicotinamide adenine dinucleotide
NADPH
NAG
reduced nicotinamide adenine dinucleotide phosphate
NADPH-CR       cytochrome c reductase pertaining to nicotinamide adenine dinucleotide
                  phosphate activity
N-acetyl-p-t/-glucosamine
NB-KB
NCICAS
nuclear factor kappa B
NCEA-RTP       National Center for Environmental Assessment Division in Research Triangle
                  Park, NC
National Cooperative Inner-City Asthma Study
NHBE
cultured human bronchial epithelial (cells)
(NH4)2S04
ammonium sulfate
Nffl
National Institutes of Health
NIST
National Institute of Standards and Technology
NK
natural killer (cells)

-------
NL
NLF
NMMAPS
NO
NO2
NOX
NOS
NS
NSAID
NSBR
NTP
NTS
02
03
OAQPS
8-OHdG
OVA
P
PAF
PAN
PAR
PBPK
PC20
PC-ALF
PCI
PD
PDioo
nasal lavage
nasal lavage fluid
National Morbidity, Mortality and Air Pollution Study
nitric oxide
nitrogen dioxide
nitrogen oxides
nitric oxide synthase
nonsmoker
non-steroidal anti-inflammatory agent
nonspecific bronchial respnsiveness
National Toxicology Program
nucleus tractus solitarius
superoxide
ozone
Office of Air Quality Planning and Standards
8-hydroxy-2 '-deoxyguanosine
ovalbumin
probability
platelet-activating factor
peroxyacetyl nitrate
proximal alveolar region
physiologically based pharmacokinetic
provocative concentration that produces a 20% decrease in forced expiratory
volume in 1 s
l-palmitoyl-2-(9-oxononanoyl)-sn-glycero-3-phosphocholine
picryl chloride
potential difference
provocative dose that produces a 100% decrease in forced expiratory volume
in 1 s
II-xxxiv

-------
PD20 provocative dose that produces a 20% decrease in forced expiratory volume in
Is
PE
PEF
"enh
PG
6PGD
PGP
pH
PHA
PIF
PM
PM10
PM15
PM25
PMN
PND
PPN
PSA
PUFA
PUL
PWM
QCE
r
R
9
r
R2
rALP
RANTES
Raw
postexposure
peak expiratory flow
enhanced pause
prostaglandm (PGD2, PGE, PGE1; PGE2 PGFla, PGF2a)
6-phosphogluconate dehydrogenase
protein gene product
hydrogen ion concentration
phytohemagglutinin
peak inspiratory flow
particulate matter
particulate matter of mass median aerodynamic diameter < 10 j^m
particulate matter of mass median aerodynamic diameter < 1 5 j^m
particulate matter of mass median aerodynamic diameter < 2.5 j^m
poly morphonucl ear neutrolphil leukocyte (also called neutrophil)
post natal day
peroxypropionyl nitrate
picryl sulfonic acid
polyunsaturated fatty acid
pulmonary
pokeweed mitogen
quasi continuous exercise
linear regression correlation coefficient
intraclass correlation coefficient
correlation coefficient
multiple correlation coefficient
recombinant antileukoprotease
regulated on activation, normal T cell-expressed and -secreted
airway resistance
II-XXXV

-------
KB               respiratory bronchiole
RER              rough endoplasmic reticulum
RFS
RH               relative humidity
RL                total pulmonary resistance
ROI              reactive oxygen intermediate
ROS              reactive oxygen species
RT               respiratory tract
RT                total respiratory resistance
a                 geometric standard deviation
                  smoker
SAC              Staphylococcus aureus Cowan 1 strain
SAW             small airway function
sc                subcutaneous
SC               stratum corneum
SD, S-D           Sprague-Dawley
SD               standard deviation
SE               standard error
SES              socioeconomic status
SGaw              specific airway conductance
SNPs             single nucleotide polymorphisms
SO2               sulfur dioxide
SO42              sulfate ion
SO42              sulfate
SOA              secondary organic aerosol
SOD              superoxide dismutase
SP                substance P
SP                surfactant protein (SP-A, SP-D)
SRaw              specific airway resistance
                                        II-xxxvi

-------
SRBC
sheep red blood cell
                  temperature
                  time (duration of exposure)
                  triiodothyronine
                  thyroxine
TB
tracheobronchial
TEA
thiobarbituric acid
99m-
  Tc-DTPA
radiolabeled diethylene triamine pentaacetic acid
T
ico
core body temperature
T
J-
cytotoxic T-lymphocytes
Th
helper T-lymphocyte
TLC
total lung capacity
TLR
Toll-like receptor
TNF
tumor necrosis factor
TNFR
tumor necrosis factor receptor
TSH
thyroid-stimulating hormone
TSP
total suspended particulate
TWA
time-weighted average
TX
thromboxane (A2, B2)
URT
upper respiratory tract
UV
ultraviolet
V
volume
VC
vital capacity
VCAM
VD
dead space
                  minute ventilation; expired volume per minute
V
  Emax
maximum minute ventilation
VT
average inspiratory flow
V
  max25%
maximum expiratory flow at 25% of the vital capacity
                                        II-xxxvii

-------
* max50%
                  maximum expiratory flow at 50% of the vital capacity
 'max50%TLC
                  maximum expiratory flow at 50% of the total lung capacity
V
  max75%
maximum expiratory flow at 75% of the vital capacity
VMD
volume mean diameter
V09
oxygen uptake by the body
VO
   2max
maximal oxygen uptake (maximal aerobic capacity)
VOCs
volatile organic compounds
VT
tidal volume
VT
tracheal transepithelial potential
 ' TB
                  tracheobronchial region volume
  Tmax
                  maximum tidal volume
WT
wild-type
                                       II-xxxviii

-------
 i      4.  DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY,
 2            AND ANIMAL-TO-HUMAN EXTRAPOLATION
 3
 4
 5     4.1    INTRODUCTION
 6          The dosimetry of ozone (O3) in humans has been examined in a series of studies published
 7     in the past decade.  This important new information further characterizes the dose of O3
 8     delivered to various sites in the respiratory tract (RT).  Ozone, classified as a reactive gas,
 9     interacts with surfactant, antioxidants, and other compounds in the epithelial lining fluid (ELF).
10     Researchers have attempted to obtain a greater understanding of how these complex interactions
11     affect uptake of O3, which eventually link to O3-induced injury. New work has also been
12     completed evaluating species differences in responses to O3 exposures, which allow more
13     accurate quantitative extrapolation from animal to human.
14          This chapter is not intended to be a complete overview of O3 dosimetry and animal-to-
15     human comparisons, but rather, it is an update of the dosimetry/extrapolation chapter from the
16     last O3 criteria document (U.S. Environmental Protection Agency, 1996), or 1996 O3 AQCD, and
17     other reviews of the earlier published literature.  The framework for presenting this chapter is
18     first a discussion in Section 4.2 of general concepts of the dosimetry of O3 in the RT. Bolus-
19     response studies are then presented in Section 4.2.1 followed by general uptake studies in
20     Section 4.2.2. Dosimetry modeling is presented in Section 4.2.3 followed by the summary and
21     conclusions for the dosimetry material in Section 4.2.4. The chapter continues in Section 4.3
22     with a discussion of species comparisons and ends with a discussion of animal-to-human
23     extrapolation. More detailed discussions of the studies are presented in the supporting material
24     to this chapter (Annex AX4). The toxicological  effects of O3 in laboratory animals and in vitro
25     test systems are discussed in Chapter 5 and direct effects of O3 in humans are discussed in
26     Chapter 6. The historical O3 literature is very briefly summarized in this chapter, providing a
27     very concise overview of previous work. The reader is referred to the 1996 O3 AQCD for more
28     detailed discussion of the literature prior to the early 1990s.
29
       January 2005                             4-1         DRAFT-DO NOT QUOTE OR CITE

-------
 1      4.2   DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT
 2           Dosimetry refers to measuring or estimating the quantity of or rate at which a chemical is
 3      absorbed by target sites within the RT. The compound most directly responsible for toxic effects
 4      may be the inhaled gas O3 or one of its chemical reaction products. Complete identification of
 5      the actual toxic agents and their integration into dosimetry is a complex issue that has not been
 6      resolved. Thus, most dosimetry investigations are concerned with the dose of the primary
 7      inhaled chemical. The actual meaning of the word 'dose' is somewhat subjective. Dahl (1990)
 8      classified O3 as a reactive gas and discussed dose in terms of: inhaled O3 concentration; amount
 9      of O3 inhaled as determined by minute volume, vapor concentration, and exposure duration;
10      uptake or the amount of O3 retained (i.e., not exhaled); O3 or its active metabolites delivered to
11      target cells or tissues; O3 or its reactive metabolites delivered to target biomolecules or
12      organelles; and O3 or its metabolites participating in the ultimate toxic reactions - the effective
13      dose. Understanding dosimetry as it relates to O3-induced injury is complex due to the fact that
14      O3 interacts primarily with the ELF which contains surfactant and antioxidants.  In the upper
15      airways ELF is thick and highly protective against oxidant injury. In lower airways ELF is
16      thinner, has lower levels of antioxidants, and thus, allows more cellular injury.  Adding to the
17      complexity is the fact that O3 can react with molecules in the ELF to create even more reactive
18      metabolites, which can then diffuse within the lung or be transported out of the  lung to generate
19      systemic effects.
20           Since the 1996 ozone criteria document (U.S. Environmental Protection Agency, 1996),
21      all new experiments have been carried out in humans to obtain direct measurements of absorbed
22      O3 in the RT, the upper RT (URT) region proximal to the tracheal entrance, and in the
23      tracheobronchial (TB) region; no uptake experiments have been identified as being performed
24      using laboratory animals.
25           In vivo dosimetry studies described in the last criteria document estimate that uptake of
26      inhaled O3 in rat is 0.50 of inhaled O3 concentration and of that about half is removed in the
27      head, the just under half in the lungs, and only a small percentage removed in the larynx/trachea.
28      There was no agreement as to whether uptake was dependent on flow. The uptake of O3 in the
29      RT of humans at rest was between 0.8 and 0.95 .  Studies in humans have shown that increasing
30      minute ventilation (VE) with exercise (by increasing both breathing frequency and tidal volume)
31      causes only a small decrease in uptake efficiency by the total RT. Mode of breathing had little

        January 2005                               4-2         DRAFT-DO NOT QUOTE OR CITE

-------
 1      effect on uptake; oral breathing had approximately 10% greater uptake efficiency than nasal
 2      breathing.  Comparing bronchoalveolar lavage (BAL) cells from rat and human, a 0.4 ppm
 3      exposure in exercising humans gave 4 to 5 times the dose of O3 (dose retained) than a rat at rest
 4      exposed to the same concentration. Overall, uptake efficiency data between humans and other
 5      species had a great deal of consistency.  Miller (1995) reviewed the major factors influencing RT
 6      uptake of O3: (1) structure of the RT region, (2) ventilation, and (3) gas transport mechanisms.
 7      In comparing rats and humans, they differ greatly in URT structure, which imparts disparate
 8      airflow streams. Rate, depth, and route of breathing all influence the amount of O3 inhaled.
 9      However, route of breathing has little biological significance.  In humans at exposures to
10      0.5 ppm (resting) or 0.2 ppm (exercising), breathing frequency increases and tidal volume
11      decreases such that minute ventilation is not altered. Miller further noted that local dose is the
12      critical link between exposure and response, and that modeling of O3 has typically discounted the
13      mucociliary layer of the URT and tracheobronchial regions.
14           In vitro dosimetry studies using isolated lung preparations showed that uptake efficiency is
15      chemical-reaction dependent, indicating the importance of reaction product formation. These
16      reaction products, created mainly by the ozonolysis of polyunsaturated fatty acids (PUFA) ,
17      included hydrogen peroxide, aldehydes, and hydroxyhydroperoxides, which are mediators of O3
18      toxicity. Other products are created by the  reaction of O3 with other ELF constituents, all of
19      which must be considered in understanding the dosimetry of O3.
20           The next two sections review the available new experimental studies on  O3 dosimetry, all
21      of which were conducted by Ultman and colleagues.  Table AX4-1 in Annex AX4 summarizes
22      theses studies.
23
24      4.2.1  Bolus-Response Studies
25           The bolus-response method has been used by the Ultman group as an approach to explore
26      the absorption of O3 by humans.  This non-invasive method consists of an injection of a known
27      volume and concentration of O3 during inspiration. Ozone uptake is the amount of O3 absorbed
28      during a single inspiration relative to the amount contained the inhaled bolus.  Exposure to
29      nitrogen dioxide (NO2) or sulfur dioxide (SO2) just prior to O3 exposure has an effect on the
30      amount of O3 absorbed in conducting airways.  Asplund  et al. (1996) used continuous O3
31      exposure, which consisted of 2 hours of O3at 0.0, 0.12, or 0.36 ppm. Rigas et al. (1997) used

        January 2005                              4-3         DRAFT-DO NOT QUOTE OR CITE

-------
 1      continuous 2 h exposures of 0.0, 0.36, or 0.75 ppm NO2, or 0.0 or 0.36 SO2.  In both experiments
 2      the continuous exposure was interrupted every 30 min by a series of bolus breaths of 2 ppm O3 at
 3      250 mL/s, targeting the lower conducting airways. With continuous O3 exposure, the absorbed
 4      fraction (FA) of the bolus decreased, which suggested to the authors that the continuous O3
 5      exposure possibly depleted biochemical substrates from the airways, reducing their capacity to
 6      absorb more O3. With NO2 and SO2 continuous exposures, the absorbed fraction of the bolus
 7      increased, which the authors contributed to the these same biochemical substrates being made
 8      available by the continuous exposure.
 9           The bolus-response technique was also used to ascertain differences in lung anatomy and
10      gender that  can alter the exposure-dose cascade (Bush et al., 1996a). Forced vital capacity
11      (FVC), total lung capacity (TLC) and anatomic dead space (VD) were  determined for ten male
12      and ten female subjects, who were then exposed to a 20 ml bolus of 3  ppm O3 injected into the
13      airstream. In all subjects, dosimetry differences could be explained by differences in VD.  The
14      investigators point out that the applicability of their results may be limited because of their
15      assumptions that the intrinsic mass transfer parameter (Ka) was independent of location in the
16      RT and that there was no mucous resistance.  They further suggested that the dependence of Ka
17      on flowrate  and VD be restricted to flowrates < 1000 mL/s until studies at higher rates have been
18      performed.
19           Nodelman and Ultman (1999a) used the bolus-response method  to demonstrate that the
20      uptake distributions of O3 was  sensitive to the mode of breathing and to the airflow rate.
21      As flowrates increased from 150 to 1000 mL/s, O3 penetrated deeper in to the lung and
22      penetration  was further increased with oral breathing.  The authors suggest that the switch from
23      nasal to oral breathing coupled with increases in respiratory flow as occurs during exercise
24      causes a shift in the O3 dose distribution, allowing O3 to penetrate deeper into the lung,
25      increasing the potential for damage to bronchiolar and alveolar tissues.
26           Very recent work by this group (Ultman et al., 2004) demonstrated that differences in
27      regional  O3  uptake are due, in part, to differences in lung anatomy. Using  60 male and female
28      subjects  exposed to O3 at a high minute ventilation, they measured the penetration volume at
29      which 50%  of the O3 bolus was taken up. Very little O3 was taken up  in the upper airway, thus
30      most of it was taken up in the lower conducting airways and peripheral airspaces.  This
31      significant difference in uptake suggests  to the authors that in females the smaller airways,

        January 2005                             4-4        DRAFT-DO NOT QUOTE OR CITE

-------
 1      and associated larger surface-to-volume ratio, enhance local O3 uptake and cause reduced
 2      penetration of O3 into the distal lung. Thus, these findings indicate that overall O3 uptake is not
 3      related to airway size, but that the distribution of O3 shifts distally as the size of the airway is
 4      increased.
 5           These bolus- response studies suggest that prior continuous exposure to O3 limits uptake
 6      from a bolus dose. This paradigm of exposure would have some relevance to environmental
 7      exposures in which humans may receive differing day and night concentrations of O3.  The lack
 8      of gender differences in O3 dosimetry is an important finding that is in agreement with
 9      Sarangapani  et al. (2003), discussed in Section4.2.3, who reported no gender differences in O3
10      extraction. The  new findings characterizing O3 uptake as inversely related to airflow are in
11      agreement with earlier animal studies.  Discussions about estimating mass transfer coefficients
12      and the "accuracy" of these models are contained in Annex AX4.
13
14      4.2.2   General Uptake Studies
15           Ultman and colleagues have recently completed some general uptake studies to determine
16      the ratio of O3 uptake to the quantity of O3 inhaled. Fractional absorption  (FA) was determined
17      at exposures  of 0.2 or 0.4 ppm O3 while exercising at a minute volume of approximately
18      20 L/min for 60  minutes or 40 L/min for 30 minutes in both men and women (Rigas et al., 2000).
19      Fractional absorption ranged from 0.56 to 0.98 with only large changes in concentration, minute
20      volume, and  exposure time having statistically significant effects on FA.  Intersubject
21      differences had the largest influence on FA, resulting in a variation of approximately 10%.
22      As the quantity of O3 retained by the RT is equal to FA times the quantity  of O3 inhaled,
23      relatively large changes in concentration, minute volume, or exposure time may result in
24      relatively large changes in the amount of O3 retained by the RT or absorbed locally.  The authors
25      concluded that for exposure times < 2 h,  inhaled dose is a reasonable predictor of actual uptake
26      as long as there are fixed concentrations  of O3 and fixed levels of exercise. More importantly,
27      individuals exposed to similar doses varied in the amount of actual dose received.
28           Santiago et al. (2001) studied the effects of airflow rate (3 to 15 L/min) and O3
29      concentration (0.1, 0.2, or 0.4 ppm) on O3 uptake in nasal cavities of males and females. The FA
30      in the nose was inversely related to the flowrate and the concentration of O3  in the inlet air.
31      They computed a gas-phase diffusion resistance of < 24% of overall diffusion resistance which

        January 2005                              4-5         DRAFT-DO NOT QUOTE OR CITE

-------
 1      suggested to them that simultaneously occurring diffusion and chemical reactions in the mucous
 2      layer were the limiting factors in O3 uptake.  Difference in O3 uptake ranged from 0.63 to 0.97 at
 3      flowrates of 3 L/min and 0.25 to 0.50 at 15 L/min.  The effect of flowrate, concentration, and
 4      subject on FA were statistically significant, but subject variability accounted for approximately
 5      half of the total variation in FA.  Both these general uptake studies, done at environmentally
 6      relevant O3 concentrations, indicate that inter-individual differences in fractional uptake are
 7      extremely important in O3 dose-response relationships.
 8           In the research mentioned above, Ultman et al. (2004) also completed continuous exposure
 9      studies.  The same 60 subjects were exposed continuously for 1 h to either clean air or 0.25 ppm
10      ozone while exercising at a target minute ventilation of 30 L/min.  This is the first study to assess
11      ventilatory and dosimetric parameters for an entire hour of exposure.  Additionally they
12      measured bronchial cross-sectional area available for gas diffusion in addition to other
13      ventilatory parameters.  The mean fractional O3 uptake efficiency was 0.89 ± 0.06.  They found
14      an inverse correlation between uptake and breathing frequency and a direct correlation between
15      uptake and tidal volume. The uptake efficiency decreased during the four sequential 15 minute
16      intervals of the 1 h exposure, demonstrating a general decrease in uptake efficiency with
17      increased breathing frequency and decreasing tidal volume. Ozone uptake rate correlated with
18      individual bronchial cross-sectional area, but did not correlate with individual %FEVj. Neither
19      of these parameters correlated with the penetration volume determined in the bolus studies
20      mentioned above. The authors concluded that the intersubject differences in forced respiratory
21      responses were not due to differences in O3 uptake. However, these data did partially support
22      the hypothesis that the differences in cross-sectional area available for gas diffusion induce
23      differences in O3 uptake.
24
25      4.2.3   Dosimetry Modeling
26           When all of the animal and human in vivo O3 uptake efficiency data are compared, there is
27      a good degree of consistency across data sets, which raises the level of confidence with which
28      these data sets can be used to support dosimetric model formulations.  Early models predicted
29      that net O3 dose to lung lining fluid plus tissue gradually decreases distally from the trachea
30      toward the end of the TB, and then rapidly decreases in the pulmonary region.  When the
31      theoretical dose of O3 to lung tissue is computed, it is low in the trachea, increases to a maximum

        January 2005                              4-6         DRAFT-DO NOT QUOTE OR CITE

-------
 1      in the terminal bronchioles of the first generation of the pulmonary region, and then decreases
 2      rapidly distally into the pulmonary region.  The increased VT and flow, associated with exercise
 3      in humans or CO2-stimulated ventilation increases in rats, shifts O3 dose further into the
 4      periphery of the lung, causing a disproportionate increase in distal lung dose.
 5           Localized damage to lung tissue has been modeled showing variation of O3 dose among
 6      anatomically equivalent ventilatory units as a function of path length from the trachea with
 7      shorter paths showing greater damage. More recent data indicate that the primary site of acute
 8      cell injury occurs in the conducting airways (Postlethwait et al., 2000). These data  must be
 9      considered when developing models that attempt to predict site-specific locations of O3-induced
10      injury.  The early models computed relationships between delivered regional dose and response
11      with the assumption that O3 was the active agent responsible for injury. It is now known that
12      reactive intermediates such as hydrohydroxyperoxides and aldehydes are important agents
13      mediating the response to O3 (further discussed in Section 5.3.1).  Thus, models must consider
14      O3 reaction/diffusion in the ELF and ELF-derived reactions products.
15           Table AX4-2 in the annex presents a summary of new theoretical studies of the uptake of
16      O3 by the RTs (or regions) of humans and laboratory animals that have become available since
17      the 1996 review.  They are discussed below.
18           Overton and Graham (1995) created a rat model combining multiple path anatomic models
19      and one-dimensional convection-dispersion equations which simulates transport and uptake of
20      O3 in airways and airspaces of the modeled TB region. Predictions from this model realistically
21      detail O3 transport and uptake of different but morphologically equivalent sites.
22           Using computational fluid dynamics (CFD), Cohen-Hubal et al. (1996) modeled the effect
23      of the mucus layer thickness in the nasal passage of a rat. Predictions of overall uptake were
24      within the range of measured uptake. Predicted regional O3 flux was correlated with measured
25      cell proliferation for the CFD simulation that incorporated two regions, each with a different
26      mucus thickness.  But using bolus-response data described above, Hu et al. (1994) and Bush
27      et al.  (2001) estimate a reaction rate constant that is more than 1000 times as large as that used
28      by Cohen-Hubal et al. (1996).
29           The authors acknowledge that the reaction rate constant may be underestimated in this
30      model due to the report by Pryor (1992) predicting that O3 does not penetrate a liquid lining
31      layer more than 0.1  |im thick.

        January 2005                               4-7         DRAFT-DO NOT QUOTE OR CITE

-------
 1           With a RT dosimetry model, Overton et al. (1996) investigated the sensitivity of absorbed
 2      fraction (AF), proximal alveolar region (PAR) dose, and PAR dose ratio to TB region volume
 3      (VTB) and TB region expansion in humans and rats.  The PAR was defined as the first generation
 4      distal to terminal bronchioles and the PAR dose ratio was defined as the ratio of a rat's predicted
 5      PAR dose to a human's predicted PAR dose.  This ratio relates human and rat exposure
 6      concentrations so that both species receive the same PAR dose. In rats, the PAR is a region of
 7      major damage from O3. For each species, three values of V^ were used: a mean value from the
 8      literature and the mean ± twice the SD. For both the rat and human  simulations, there were
 9      several general findings: (1) AF  and PAR both increased with decreasing V^, e.g., using the
10      highest kTB, the PAR for Vra-2SD was five times greater than the PAR for Vra+2SD, (2) AF and
11      PAR both decreased with TB expansion relative to no expansion, 3) PAR increased with tidal
12      volume, 4) PAR increased with decreasing k^, and 5) AF increased with kTB.
13           Bush et al. (2001) modified their single-path model (Bush et al.,  1996) so that simulations
14      would coincide with experimental AF data for O3 and C12 during oral and nasal breathing.
15      Relative to their original model, the Bush et al. (2001) model added lung expansion and
16      modified the mass transfer coefficients for both the gas-phase (kg) and the liquid-phase (k,).
17      Using kb a variable reaction rate constant (k,.) was defined as ^ divided by the mucus layer
18      thickness  which  was assumed to decrease in thickness with progression from the trachea to the
19      lower airways. Consistent with Overton  et al. (1996), considering expansion of the TB airways
20      reduced AF versus no expansion. As very little inhaled O3 reaches the peripheral lung, it was not
21      surprising that alveolar expansion had minimal affect on AF. Ignoring k,., the simulations for O3
22      and C12 were nearly the same since the gas-phase diffusion coefficients of O3 and C12 are similar.
23      But for a given volumetric depth  into the TB airways of the lung, experimental AF data are
24      always greater for O3 than for C12. The authors surmised that the difference between the AF for
25      these gases could be explained adequately based solely on the diffusive resistance of O3 in
26      airways surface fluid (modeled by k,.). Interestingly, k,. was lower for oral than nasal  breathing,
27      implying less antioxidant capacity in the  airway surface liquid of the oral versus the nasal cavity.
28      Qualitatively,  model simulations  also agreed well with the experimental data of Gerrity et al.
29      (1995).
30           Age- and gender-specific differences in both regional and systemic uptake in humans was
31      modeled using physiologically-based pharmacokinetic (PBPK ) approach (Sarangapani et al.,

        January 2005                               4-8         DRAFT-DO NOT QUOTE OR CITE

-------
 1     2003). The model estimated that regional extraction of O3 is relatively insensitive to age and
 2     gender and that the postnatal period is the age (in which extraction of O3 in infants is 2- to 8-fold
 3     higher than is adults) at which the largest difference in pharmacokinetics exist.
 4           A recent attempt was made (Mudway and Kelly, 2004) to model O3 dose-inflammatory
 5     response using a meta-analysis of 23 exposures in published human chamber studies.  The O3
 6     concentrations ranged from 0.08 to 0.6 ppm and the exposure durations ranged from 60 to
 7     396 minutes.  The analysis showed linear relationships between O3 dose and neutrophilia in
 8     bronchoalveolar lavage fluid (BALF). Linear relationships were also observed between O3 dose
 9     and protein leakage into BALF, which suggested to the authors that a large-scale study could
10     determine a possible O3 threshold level for these inflammatory responses.
11
12     4.2.4   Summary and Conclusions - Dosimetry
13           Ozone is a highly reactive gas and powerful oxidant with  a short half-life. Uptake occurs
14     in mucous membranes of the RT where O3 immediately reacts with components of the ELF.
15     Uptake efficiency is chemical-reaction dependent and the reaction products (hydrogen peroxide,
16     aldehydes,  and hydroxyhydroperoxides)  created by ozonolysis of polyunsaturated fatty acids
17     (PUFA) mediate of O3 toxicity. The previous literature review  found that uptake of O3 in rat is
18     about 0.50  and in humans at rest is about 0.8 to 0.95. About 0.07 of the O3 is removed in the
19     larynx/trachea, about 0.50 is removed in  the head, and about 0.43 is removed in the lungs, where
20     the primary site of damage is the centriacinar region (CAR). There was no agreement as to
21     whether uptake was dependent on flow.  Studies in humans have shown that increasing VE with
22     exercise (by increasing both breathing frequency and tidal volume) causes only a small decrease
23     in uptake efficiency by the total RT. Mode of breathing had little effect on uptake, suggesting
24     that O3 is removed equally by both mouth and nose. Comparing BAL cells from rat and human,
25     a 0.4  ppm exposure in exercising humans gave 4 to 5 times the  dose of O3 than a rat at rest
26     exposed to the same concentration.
27           New research on O3 uptake has been performed in humans, but not in laboratory animals.
28     Bolus-response studies demonstrated that a previous continuous exposure to O3 decreased the
29     absorption  of a bolus of O3, probably due to depletion of compounds able to absorb O3. Previous
30     continuous exposure to NO2 and SO2 increased absorption of a bolus of O3.  These data are of
31     some relevance to environmental exposures where humans may receive differing concentrations

       January 2005                              4-9        DRAFT-DO NOT QUOTE OR CITE

-------
 1      of O3 depending on time of day.  Another bolus-response study showed that absorption of O3 was
 2      dependent on VD, but not height, weight, age, gender, FVC, or TLC. In contrast to earlier data,
 3      the bolus-response method was used to demonstrate that the uptake distributions of O3 is
 4      sensitive to the mode of breathing and to the airflow rate. As flow rates increased from 150 to
 5      1000 mL/s, O3 penetrated deeper in to the lung and was further increased with oral breathing.
 6      This suggests that the switch from nasal to oral breathing coupled with increases in respiratory
 7      flow as occurs during exercise causes a shift in the O3 dose distribution, allowing O3 to penetrate
 8      deeper into the lung, increasing the potential of damage to bronchiolar and alveolar tissues. The
 9      finding that O3 uptake is inversely related to airflow agrees with earlier animal studies.
10           New general uptake study data demonstrate that exercising men and women receiving
11      0.2 or 0.4 ppm O3 at 20  L/min for 60 minutes or 40 L/min for 30 minutes absorb 0.56 to 0.98.
12      The absorbed fraction or FA is affected only by large changes in concentration, minute volume,
13      and exposure time.  This suggests that for exposure times < 2 h, inhaled dose is a reasonable
14      predictor of actual uptake as long as there are fixed concentrations of O3 and fixed levels  of
15      exercise.  Individuals exposed to similar concentrations vary considerably in the amount of
16      actual dose received. This intersubject variability was also demonstrated in a study of O3 uptake
17      in nasal cavities of men and women.  The FA in the nose was inversely related to the flowrate
18      and the concentration of O3 suggesting that simultaneously occurring diffusion and chemical
19      reactions in the mucous layer were the limiting factors in O3 uptake. Both these general uptake
20      studies, done at environmentally relevant O3 concentrations, indicate that inter-individual
21      differences in fractional uptake, which can range from 0.25 to 0.97, are extremely important in
22      O3 dose-response relationships.
23           The consistency of uptake data generated in animal and human studies allow a high level
24      of confidence in their use in dosimetry modeling. Early models predicted that net O3 dose to
25      ELF and tissue gradually decreases distally from the trachea toward the end of the  TB, and then
26      rapidly decreases in the pulmonary region.  Exercise-induced or CO2-stimulated increases in VT
27      and flow, shift O3 dose further into the periphery of the lung, causing a disproportionate increase
28      in distal lung dose.  Localized damage to lung tissue has been modeled showing variation of O3
29      dose among anatomically equivalent ventilatory units as a function of path length from the
30      trachea with  shorter paths showing greater damage.
        January 2005                              4-10       DRAFT-DO NOT QUOTE OR CITE

-------
 1           New models have produced some refinements of earlier models such as: (1) the use of
 2      mucus resistance and thickness in describing O3 dosimetry and determining the patterns of
 3      O3-induced lesions; (2) the shape of the dose versus generation plot along any path from the
 4      trachea to alveoli is independent of path, with the tissue dose decreasing with increasing
 5      generation index; (3) simulations sensitive to conducting airway volume but relatively
 6      insensitive to characteristics of the respiratory airspace; (4) the importance of TB region
 7      expansion; (5) the importance of dose received in the PAR both inter-individual differences and
 8      extrapolations based on dose;  (6) revaluation of mass transfer coefficients for conducting
 9      airways, and (7) extraction of O3 in infants is 2- to 8-fold higher than in adults, but the
10      differences leveled out by age 5. Additionally, more recent data indicate that the primary site of
11      acute cell injury occurs in the  conducting airways and that reactive intermediates in the ELF,
12      rather than O3 itself, are responsible for pulmonary injury. These data must be considered when
13      developing new models.
14
15
16      4.3    SPECIES HOMOLOGY, SENSITIVITY, AND ANIMAL-TO-HUMAN
17             EXTRAPOLATION
18           Basic similarities exist across human and other animals species with regard to basic
19      anatomy, physiology, biochemistry, cell biology, and disease processes.  This homology creates
20      similarities in acute O3-induced effects, especially in the respiratory tract and in lung defense
21      mechanisms. Rodents appear to have a slightly higher tachypneic response to O3, which is
22      clearly  concentration-dependent in most species and shows parallel exacerbation when
23      hyperventilation (e.g., exercise or CO2) is superimposed. What is not known is whether this is
24      evidence of pulmonary irritant sensitivity, perhaps as a prelude to toxicity, or whether tachypnea
25      is a defensive action taken by  the respiratory system to minimize distal lung O3 deposition.
26      Airway or lung resistance in humans is not affected appreciably by acute exposure to O3, except
27      under conditions of heavy exercise; animals appear to need high-level  exposures or special
28      preparations that bypass nasal scrubbing. Dynamic lung compliance (Cdyn) has been shown to
29      have small magnitude decreases in response to O3 in some studies across species, but it is
30      thought that these changes are of little biological significant for ambient exposures. Spirometric
31      changes, the hallmark of O3 response in humans, occur in  rats, but to a lesser degree. It is

        January 2005                              4-11        DRAFT-DO NOT QUOTE OR CITE

-------
 1      unclear, however, the degree to which anesthesia (rat) and the comparability of hyperventilation
 2      induced by CO2 (rat) or exercise (human) may influence this difference in responsiveness.
 3      Collectively, the acute functional response of laboratory animals to O3 appears quite homologous
 4      to that of the human.
 5           Examination of BAL constituents show that the influx of inflammatory cells and protein
 6      from the serum is influenced by species, but perhaps to less extent than by ventilation and
 7      antioxidant status. Adjustment for these factors can modulate responses to approximate animal
 8      responses to those of humans.  Unfortunately, these influential factors are rarely measured and,
 9      even less often,  controlled. Increases in protein levels in BALF with O3 exposures in guinea pigs
10      are also a factor in the species' susceptibility to the effects of O3. Species comparisons of acute
11      O3 exposures to mice, guinea pigs, rats, hamsters, and rabbits found that guinea pigs were the
12      most responsive (to > 0.2 ppm); rabbits were the least responsive (2.0 ppm only); and rats,
13      hamsters, and mice were intermediate (effects at > 1.0 ppm). Rats and humans have subtle
14      species-specific differences in inflammatory responses to O3 in terms of the timing of PMN
15      influx in the nasal and bronchoalveolar regions.
16           When humans are exposed to O3 repeatedly for several consecutive days, lung function
17      decrements subside, and normal spirometric parameters are regained. This phenomenon of
18      functional attenuation also has been demonstrated in rats, not only in terms of spirometry, but
19      also in terms of the classic tachypneic ventilatory response. Full or partial attenuation of the
20      BAL parameters also appears to occur in both rats and humans, but exposure scenario appears to
21      play a role; other cellular changes in animals do not attenuate.  Existing epidemiologic  studies
22      provide only suggestive evidence that persistent or progressive deterioration in lung function is
23      associated with long-term oxidant-pollutant exposure (See Chapter 7).  These  long-term effects
24      are thought to be expressed in the form of maximum airflow or spirometric abnormalities, but
25      the foundation for this conclusion remains weak and hypothetical.  Animal study data, although
26      suggesting that O3 has effects on lung function  at near-ambient levels, present a variable picture
27      of response that may or may not relate to technical conditions of exposure or some other, yet
28      undiscovered variable of response. Thus, a cogent interpretation of the animal findings as
29      definitive evidence of chronic deterioration of lung function would be difficult at this time.
30      However, the subtle functional defects apparent after 12 to 18 mo of exposure and the detailed
31      morphometric assessments of the O3-induced lesions do appear consistent with the modicum of

        January 2005                              4-12       DRAFT-DO NOT QUOTE OR CITE

-------
 1      studies focusing on long-term effects in human populations. Based on the apparent homology of
 2      these responses between humans and laboratory animals, animal studies may potentially provide
 3      a means to more directly assess such chronic health concerns.
 4           A species' susceptibility to the effects of O3 exposure may be due, in part, to biochemical
 5      differences among species.  Evidence for this is provided by differences in activity of SD rat and
 6      rhesus monkey CYP moonoxygenases elicited by O3 exposure (Lee et al., 1998).  Additional
 7      characterization of species- and region-specific CYP enzymes will create a better understanding
 8      of the differences in response to O3.  This will allow more accurate extrapolation from animal
 9      exposures to human exposures and toxic effects.
10           Antioxidant metabolism varies widely among species, which can greatly influence the
11      effects of O3 (discussed in greater detail in 5.2.1.3). The guinea pig appears to be the species
12      most susceptible to ozone. Early studies ranked mice > rats > guinea pigs in order of antioxidant
13      responsiveness to O3 challenge.  Guinea pigs have been shown to have lower basal levels of
14      GSH transferase activity, lower activity of GSH peroxidases, and lower levels of vitamin E
15      compared to rats. These lower levels of antioxidants combined with increases in protein levels
16      in BALF (discussed above) with O3 exposures likely explain, at least in part, the species'
17      susceptibility to the effects of O3.
18           Because cytokine and chemokine responses are so important in an animal's  defense against
19      O3 exposure, comparisons of differences in species expression and activity of these
20      inflammatory mediators is necessary. Arsalane et al. (1995) compared guinea pig and human
21      AM recovered in BALF and subsequently exposed in vitro to 0.1 to 1 ppm for 60  minutes.
22      Measurement of inflammatory cytokines showed a peak at 0.4 ppm in both species.  Guinea pig
23      AM had an increase in IL-6 and TNFa while human AM had increases in TNFa, IL-lb, IL-6  and
24      IL-8. This exposure also caused an increase in mRNA expression for TNFa, IL-lb, IL-6 and
25      IL-8 in human cells. At 0.1 ppm exposures, only TNFa secretion was increased.  These data
26      suggest similar cytokine responses in guinea pigs and humans, both qualitatively and
27      quantitatively.
28           Species differences in morphological responses to O3 exposure have been characterized by
29      Dormans et al. (1999), as discussed in previous sections. Dormans et al. (1999) continuously
30      exposed rats, mice, and male guinea pigs to filtered air, 0.2, or 0.4 ppm O3 for 3, 7, 28, and
31      56 days. The animals exposed for 28 days were examined at 3, 7, or 28 days PE.  Depending

        January 2005                              4-13        DRAFT-DO NOT QUOTE OR CITE

-------
 1      on the endpoint studied, the species varied in sensitivity.  Greater sensitivity was shown in the
 2      mouse as determined by biochemical endpoints, persistence of bronchiolar epithelial
 3      hypertrophy, and recovery time.  Guinea pigs were more sensitive in terms of the inflammatory
 4      response though all three species had increases in the inflammatory response after three days that
 5      did not decrease with exposure. These data on inflammation are in general agreement with
 6      Hatch et al., (1986), discussed above.  In all species the longest  exposure to the highest dose
 7      caused increased collagen in ductal septa and large lamellar bodies in Type II cells, but that
 8      response also occurred in rats and guinea pigs at 0.2 ppm.  No fibrosis was seen at the shorter
 9      exposure times and the authors question whether fibrosis occurs in healthy humans after
10      continuous exposure. The authors do not rule out the possibility that some of these differences
11      may be attributable to differences in total inhaled dose or dose actually reaching a target site.
12      Overall, the authors rated mice as most susceptible, followed by guinea pigs and rats.
13           Comparisons of airway effects in rats, monkeys and ferrets resulting from exposures of
14      1.0 ppm O3 for 8  h (Sterner-Kock et al. 2000) demonstrated that monkeys and ferrets had a
15      similar inflammatory responses and epithelial necrosis. The response of these two species was
16      more severe than that seen in rats.  These data suggest that ferrets are a good  animal model for
17      O3-induced airway effects due to the similarities in pulmonary structure between primates and
18      ferrets.
19           A number of species,  including nonhuman primates, dogs, cats,  rabbits, and rodents, have
20      been used to study the effects of O3 exposure on airway bronchoconstriction.  A commonly used
21      model of bronchospasm utilizes guinea pigs acutely exposed to  high O3 concentrations (2 to
22      3 ppm) to induce airway hyperreactivity (AHR). As mentioned earlier, the model is helpful for
23      determining mechanistic aspects of AHR, but is not really relevant for extrapolation to potential
24      airway responses in humans exposed to ambient levels of O3.  Additionally, guinea pigs have
25      been shown to have AHR in other studies that is very similar to human asthmatics, but the utility
26      of guinea pig data is somewhat limited by their disparity from other animal models.
27           The rat is a key species used in O3 toxicological studies, but the rat has  both behavioral
28      and physiological mechanisms that can lower core temperature in response to acute exposures,
29      thus limiting extrapolation of rat data to humans. Iwasaki et al.  (1998) evaluated cardiovascular
30      and thermoregulatory responses to O3 at exposure of 0.1, 0.3,  and 0.5  ppm O3 8 hrs/day for
31      4 consecutive days.  A dose-dependent disruption of HR and Tco were seen on the first and

        January 2005                             4-14       DRAFT-DO NOT QUOTE OR CITE

-------
 1      second days of exposure, which then recovered to control values.  Watkinson et al. (2003)
 2      exposed rats to 0.5 ppm O3 and observed this hypothermic response, which included lowered
 3      HR, lowered Tco, and increased inflammatory components in BALF.  The authors suggested that
 4      the response is an inherent reflexive pattern that can possibly attenuate O3 toxicity in rodents.
 5      They discuss the cascade of effects created by decreases in Tco, which include: (1) lowered
 6      metabolic rate, (2) altered enzyme kinetics,  (3) altered membrane function, (4) decreased oxygen
 7      consumption and demand, (5) reductions in minute ventilation, which would act to limit the dose
 8      of O3 delivered to the lungs. These effects are concurrent with changes in HR which lead to:
 9      (1) decreased CO, (2) lowered BP, and (3) decreased tissue perfusion, all of which may lead to
10      functional deficits.  They hypothermic response  has not been observed in humans except at very
11      high exposures, which  complicates extrapolation of results in rat studies to humans.
12           The importance of animal studies derives from their utilization in determining cause-effect
13      relationships between exposure and health outcome, but the animal data must be integrated with
14      epidemiological studies and controlled human clinical studies.  Animal studies can corroborate
15      both clinical and epidemiology studies and further provide important data that is impossible to
16      collect in human studies. Toxic pulmonary  and  extrapulmonary effects following ozone
17      exposure have been well-studied in rodents, nonhuman primates, and a few other species;
18      so, extrapolation, both  qualitative and  quantitative, to human exposures and consequent health
19      effects is possible. Quantitative extrapolation, required to determine what specific exposure is
20      likely to cause an effect in humans, is theoretically founded on the equivalency of mechanisms
21      across species. At the molecular level, O3 acts on the carbon-carbon double bond in
22      polyunsaturated fatty acids and on sulfhydryl groups in proteins, both of which are found within
23      cell membranes in animals and humans.  At higher levels of cellular organization, cells affected
24      in animals (e.g., AMs,  Type 1 cells) have similar functions in humans, and organ  systems (e.g.,
25      respiratory system) have major interspecies similarities. However, interspecies differences do
26      occur and confound extrapolation.
27           Quantitative extrapolation, which involves a combination of dosimetry and species
28      sensitivity, still requires more research before it  can be fully realized. Knowledge of dosimetric
29      animal-to-human extrapolation is more advanced than that of species-sensitivity, but
30      extrapolation models have not been completely validated,  and therefore, significant uncertainties
31      remain.  Mathematical  modeling of O3 deposition in the lower respiratory tract (i.e., from the

        January 2005                              4-15        DRAFT-DO NOT QUOTE OR CITE

-------
 1      trachea to alveoli) of several animal species and humans shows that the pattern of regional dose
 2      is similar, but that absolute values differ. In spite of structural and ventilatory differences
 3      between species, the greatest predicted tissue dose is to the CAR. Even though the CAR of rats
 4      has very rudimentary respiratory bronchioles, compared to well-developed ones in primates, the
 5      CAR of both rats and nonhuman primates respond similarly to O3.
 6           Experimental measurement of delivered O3 doses estimate that total respiratory uptake is
 7      -47% in laboratory animals and -87% in exercising humans, while nasopharyngeal removal is
 8      -17% in rats and -40% in humans.  The previous O3 AQCD (U.S. Environmental Protection
 9      Agency, 1996) provided the first quantitative animal-to-human extrapolation of morphological
10      changes in the proximal alveolar region using rat and monkey studies.  The extrapolation
11      predicted that a 9-year-old child would have a 20% or 75% increase in PAR tissue thickness if
12      their sensitivity to O3 was equal to that of a rat or monkey, respectively.  Adults would have
13      15 or 70% increase, suggesting the potential for chronic effects in humans. In spite of the
14      significant uncertainties, this extrapolation raises concern about the potential for chronic effects
15      in humans
16           Experiments using 0.4 ppm 18O3 suggested that exercising humans received a 4- to 5-fold
17      higher 18O3 concentrations in BAL than resting rats (Hatch et al., 1994).  That level of exposure
18      increased BAL protein and PMNs in humans, while a concentration of 2.0 ppm in rats was
19      necessary for similar effects.  Caveats in the interpretation of 18O3 studies include: (1) only a
20      very small portion of the labeled compound is recoverable to assess incorporation; and
21      (2) if species being compared differ in physiocochemical factors controlling mass transfer
22      and downstream O3 metabolism, it could cause significant differences in  the amount of inhaled
23      18O3 that is detected during subsequent tissue analysis. Further,  species differences in pulmonary
24      anatomy, ventilation, antioxidants,  and susceptibility all influence dose, repair processes, and
25      tolerance to subsequent O3 exposure. Important differences between exercising humans and
26      resting rats that can affect tissue  O3 dose include:  (1) increased ventilation and O3 delivery with
27      exercise; (2) decreased pulmonary ventilation and body temperature during O3 exposure in rats;
28      (3) diminished  dose received in rats due to their burying their noses in their fur during exposure;
29      and (4) increased concentration of antioxidants in ELF in rats compared to humans.  These
30      antioxidants are important for converting O3 to inactive products before toxicity occurs (Kari
31      et al.,  1997; Gunnison and Hatch, 1999; Plopper et al., 1998), though this quenching is not

        January 2005                             4-16       DRAFT-DO  NOT QUOTE OR CITE

-------
 1      quantitative.  These and possibly other differences between rats and humans suggest that a
 2      2 ppm exposure in nonexercising rats approximates a 0.4 ppm exposure in exercising humans.
 3      Further comparisons of exercising human exposure to 0.1 ppm for 6 hours (Devlin et al., 1991)
 4      and resting rat exposure to 0.3 ppm show inflammatory and permeability changes in humans but
 5      not rats.
 6
 7      4.3.1    Summary and Conclusions: Species Homology, Sensitivity, and
 8              Animal-to-Human Extrapolation
 9          Comparisons of acute exposures in rat and human suggest that, though both species have
10      similar qualitative responses to O3 exposure, there are interspecies mechanistic disparities that
11      necessitate careful comparisons of dose-response relationships. There is no perfect nonhuman
12      species with which to model O3 toxicity. All have limitations that must be considered when
13      attempting to extrapolate to human exposures. Awareness of these limitations, even at the level
14      of subtle strain differences within a test species,  is extremely important. The currently available
15      data suggest that LOELs in resting rats are approximately 4- to-5  fold higher than for exercising
16      humans for toxicological endpoints including BAL protein and BAL PMNs. Studies comparing
17      species-specific differences in O3-induced effects showed that guinea pigs were the most
18      susceptible, rabbits the least susceptible, and rodents intermediate in susceptibility.  The recent
19      work being done utilizing various mouse strains with differing sensitivities to  O3 will help us to
20      understand the extremely complex inter-individual differences in  human sensitivity to O3.
21
       January 2005                             4-17        DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Arsalane, K.; Cosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
  3             synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
  4             13:60-68.
  5       Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S. (1996) Longitudinal distribution of ozone absorption in
  6             the lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
  7       Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996a) Longitudinal distribution of O3
  8             absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
  9       Bush, M. L.; Raybold, T.; Abeles, S.; Hu, S.-C; Ben-Jebria, A.; Ultman, J. S.  (1996b) Longitudinal distribution of
10             ozone absorption in the lung: simulation with a single-path model. Toxicol. Appl. Pharmacol. 140: 219-226.
11       Bush, M. L.; Zhang, W.; Ben-Jebria, A.; Ultman, J. S. (2001) Longitudinal distribution of ozone and chlorine in the
12             human respiratory tract: simulation of nasal and oral breathing with the single-path diffusion model. Toxicol.
13             Appl. Pharmacol. 173: 137-145.
14       Cohen-Hubal, E. A.; Kimbell, J. S.; Fedkiw, P. S. (1996) Incorporation of nasal-lining mass-transfer resistance into a
15             CFD model for prediction of ozone dosimetry in the upper respiratory tract. Inhalation Toxicol. 8: 831-857.
16       Dahl, A. R. (1990) Dose concepts for inhaled vapors and gases. Toxicol. Appl. Pharmacol.  103: 185-197.
17       Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991)
18             Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the
19             lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
20       Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
21             time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
22       Gerrity, T. R.; Biscardi, F.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995) Bronchoscopic
23             determination of ozone uptake in humans. J. Appl. Physiol. 79: 852-860.
24       Gunnison, A. F.; Hatch, G. E. (1999) O3-induced inflammation in prepregnant, pregnant, and lactating rats correlates
25             with O3 dose estimated by 18O. Am. J.  Physiol. 276: L332-L340.
26       Hatch, G. E.; Slade, R.; Stead, A. G.; Graham, J. A. (1986) Species comparison of acute inhalation toxicity of ozone
27             and phosgene.  J. Toxicol. Environ. Health 19: 43-53.
28       Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
29             respiratory flow. J. Appl. Physiol. 77:  574-583.
30       Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
31             conscious rats. Ind. Health 36: 57-60.
32       Kari, F.; Hatch, G.; Slade, R.; Crissman, K.; Simeonova, P. P.; Luster, M. (1997) Dietary restriction mitigates
33             ozone-induced lung inflammation in rats: a role for endogenous antioxidants. Am. J. Respir. Cell Mol. Biol.
34             17:740-747.
35       Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
3 6             differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
37             and induction in rats. Drug Metab. Dispos. 26: 396-400.
3 8       Miller, F. J. (1995) Uptake and fate of ozone in the respiratory tract. Toxicol. Lett. 82/83: 277-285.
39       Mudway, I. S.; Kelly, F. J. (2004) An investigation of inhaled ozone dose and the magnitude of airway inflammation
40             in healthy adults. Am.  J. Respir. Crit. Care Med. 169: 1089-1095.
41       Nodelman, V.; Ultman, J. S. (1999a) Longitudinal distribution of chlorine absorption in human airways: a
42             comparison to  ozone absorption. J. Appl. Physiol. 87: 2073-2080.
43       Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat respiratory tract model with
44             an asymmetric tracheobronchial region patterned on complete conducting airway cast data. Comput. Biomed.
45             Res. 28: 171-190.
46       Overton, J. H.; Graham, R. C.; Menache, M.  G.; Mercer, R. R.; Miller, F. J. (1996) Influence of tracheobronchial
47             region expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhalation Toxicol.
48             8:723-745.
49       Plopper, C. G.; Hatch, G. E.;  Wong, V.; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.;
50             Buckpitt, A. R. (1998) Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury,
51             site-specific ozone dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
52             19:387-399.
         January 2005                                    4-18          DRAFT-DO NOT QUOTE OR CITE

-------
  1       Postlethwait, E. M; Joad, J. P.; Hyde, D. M; Schelegle, E. S.; Brie, J. M; Weir, A. J.; Putney, L. F.; Wong, V. J.;
  2             Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
  3             tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22:  191-199.
  4       Pryor, W. A. (1992) How far does ozone penetrate into the pulmonary air/tissue boundary before it reacts? Free
  5             Radic. Biol. Med.  12: 83-88.
  6       Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
  7             of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52: 173-178.
  8       Rigas, M. L.; Catlin, S. N.; Ben-Jebria, A.; Ultman, J. S. (2000) Ozone uptake in the intact human respiratory tract:
  9             relationship between inhaled dose and actual dose. J. Appl. Physiol. 88: 2015-2022.
10       Santiago, L. Y.; Harm, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001) Ozone adsorption in the human nose during
11             unidirectional airflow. J. Appl. Physiol. 91: 725-732.
12       Sarangapani, R.; Gentry,  P. R.; Covington, T. R.; Teeguarden, J.  G.; Clewell, H.  J., III. (2003) Evaluation of the
13             potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors.
14             Inhalation Toxicol. 15: 987-1016.
15       Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar
16             to nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
17       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
18             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF.  3v.
19             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
20             www.epa.gov/ncea/ozone.htm.
21       Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs: intersubject
22             variability in physiologic response. Boston, MA: Health Effects Institute.
23
         January 2005                                   4-19         DRAFT-DO NOT QUOTE OR CITE

-------
 i          CHAPTER 4 ANNEX. DOSIMETRY OF OZONE IN
 2                           THE RESPIRATORY TRACT
 3
 4
 5     AX4.1  INTRODUCTION
 6           This annex serves to provide supporting material for Chapter 4 - Dosimetry, Species,
 7     Homology, Sensitivity, and Animal-to-Human Extrapolation.  It includes tables that summarize
 8     new literature published since the last O3 criteria documents (U.S. Environmental Protection
 9     Agency, 1996). In addition, it provides descriptions of those new findings, in many cases, with
10     more detail than is provided in the chapter.
11           Dosimetry refers to measuring or estimating the quantity of or rate at which a chemical is
12     absorbed by target sites within the respiratory tract (RT).  The compound most directly
13     responsible for toxic effects may be the inhaled gas O3 or one of its chemical reaction products.
14     Complete identification of the actual toxic agents and their integration into dosimetry is a
15     complex issue that has not been resolved. Thus, most dosimetry investigations are concerned
16     with the dose of the primary inhaled chemical.  In this context, a further confounding aspect can
17     be the units of dose (e.g., mass retained per breath, mass retained per breath per body weight,
18     mass retained per breath per respiratory tract surface area). That is, when comparing dose
19     between species, what is the relevant measure of dose? This question has not been  answered;
20     units are often dictated by the type of experiment or by a choice made by the investigators.
21     There is also some lack of agreement as to what constitutes "dose." Dahl's seminal paper (1990)
22     classified O3 as a reactive gas and discussed the characterization of dose measurement by
23     parameters including:  (1) inhaled O3 concentration; (2) amount of O3 inhaled as determined by
24     minute volume, vapor concentration,  and exposure duration; (3) uptake or  the amount of O3
25     retained (i.e., not exhaled); (4) O3 or its active metabolites delivered to target cells or tissues;
26     (5) O3 or its  reactive metabolites delivered to target biomolecules or organelles; and (6) O3 or its
27     metabolites participating in the ultimate toxic reactions - the effective dose. This
28     characterization goes from least complex to greatest, culminating in measurement of the fraction
29     of the inhaled O3 that participates in the effects of cellular perturbation and/or injury.
30     Understanding dosimetry as it relates to O3-induced injury is complex due  to the fact that O3
31     interacts primarily with the epithelial lining fluid (ELF) which contains surfactant and

       January 2005                             AX4-1      DRAFT-DO NOT QUOTE OR CITE

-------
 1      antioxidants. In the upper airways ELF is thick and highly protective against oxidant injury.
 2      In lower airways ELF is thinner, has lower levels of antioxidants, and thus, allows more cellular
 3      injury.  Adding to the complexity is the fact that O3 can react with molecules in the ELF to create
 4      even more reactive metabolites, which can then diffuse within the lung or be transported out of
 5      the lung to generate systemic effects. Section 5.3.1 contains further information on the cellular
 6      targets of O3 interactions and antioxidants.
 7          Experimental dosimetry studies in laboratory animals and humans, and theoretical
 8      (dosimetry modeling) studies, have been used to obtain information on O3 dose. Since the last
 9      ozone criteria document (U.S. Environmental Protection Agency, 1996), all new experiments
10      have been carried out in humans to obtain direct measurements of absorbed O3 in the RT, the
11      upper RT (URT) region proximal to the tracheal entrance, and in the tracheobronchial (TB)
12      region;  no uptake experiments have been performed using laboratory animals. Experimentally
13      obtaining dosimetry data is extremely difficult in smaller regions or locations, such as specific
14      airways or the centriacinar region  (CAR; junction of conducting airways and gas exchange
15      region), where lesions caused by O3 occur. Nevertheless, experimentation is important for
16      determining dose, making dose comparisons between subpopulations and between different
17      species, assessing hypotheses and concepts, and validating mathematical models that can be used
18      to predict dose at specific respiratory tract sites and under more general conditions.
19          Theoretical studies are based on the use of mathematical models developed for the
20      purposes of simulating the uptake and distribution of absorbed gases in the tissues and fluids of
21      the RT. Because the factors affecting the transport and absorption of gases are applicable to all
22      mammals, a model that uses appropriate species or disease-specific anatomical and ventilatory
23      parameters can be used to describe absorption in the species and in  different-sized, aged, or
24      diseased members of the same species.  More importantly, models also may be used to make
25      interspecies and intraspecies dose  comparisons, to compare and reconcile data from different
26      experiments, to predict dose in conditions not possible or feasible experimentally, and to better
27      understand the processes involved in toxicity.
28          A review (Miller, 1995) of the factors influencing RT uptake of O3 stated that structure of
29      the RT region, ventilation, and gas transport mechanisms were important. Additionally, local
30      dose is the critical link between exposure and response.  A criticism of previous models of O3
        January 2005                             AX4-2       DRAFT-DO NOT QUOTE OR CITE

-------
 1     uptake is that they have typically discounted the mucociliary layer of the URT and
 2     tracheobronchial regions.
 3          For a more detailed discussion on experimental and theoretical dosimetry studies the reader
 4     is referred to the previous O3 criteria document (Volume III, Chapter 8, U.S. Environmental
 5     Protection Agency, 1996).
 6
 7
 8     AX4.2  EXPERIMENTAL OZONE DOSIMETRY INVESTIGATIONS
 9          There have been some advances in understanding human O3 dosimetry that better enable
10     quantitative extrapolation from laboratory animal data. The next two sections review the
11     available new experimental studies on O3 dosimetry, which involve only human subjects and are
12     all from the same laboratory. Of the studies considered in the following discussion, five
13     involved the use of the bolus response method as a probe to obtain information about the
14     mechanism of O3 uptake in the URT and TB regions. Of the remaining two investigations, one
15     focused on total uptake by the RT and the other on uptake by the nasal cavities. Table AX4-1
16     provides a summary of the newer studies.
17
18     AX4.2.1  Bolus-Response Studies
19          The bolus-response method has been used as a probe to explore the effects of physiological
20     and anatomical differences or changes on the uptake of O3 by human beings.
21          Asplund et al. (1996) studied the effects of continuous O3 inhalation on O3 uptake and
22     Rigas et al. (1997) investigated the potential effects of continuous coexposure to O3, nitrogen
23     dioxide (NO2), or sulfur dioxide (SO2) on O3 absorption.  In both of these studies,  subjects were
24     exposed "continuously" to a gas for 2 h. Every 30 min, breathing at 250 mL/s, a series of bolus
25     test breaths was performed targeted at the lower conducting airways.  Differences in bolus-
26     response absorbed fraction from an established baseline indicated the degree to which the
27     "continuous" gas exposure affected the  absorption of O3. Depending on the gas and
28     concentration, changes in absorbed fraction ranged from  -3 to +7 % (see Table 5-1).
29     Continuous O3 exposure lowered the uptake of O3, whereas NO2 and SO2 increased the uptake of
30     O3. The investigators concluded that in the tested airways, NO2 and SO2 increased the capacity
31     to absorb O3 because more of the compounds oxidized by O3 were made available. On the other

       January 2005                            AX4-3      DRAFT-DO NOT QUOTE OR CITE

-------
3
to
o
                              Table AX4-1. New Experimental Human Studies on Ozone Dosimetrya
X
H

6
o


o
H

O
Purpose/Objective
Determine the effect
of continuous O3
inhalation on O3
uptake





Evaluate the
influence of VD on
intersubject variation
of O3 dose.





Investigate the effect
of continuous
exposure to O3,
nitrogen dioxide and
sulfur dioxide on O3
absorption.


Subject
Characteristics
8 male,
3 female,
22-3 1 years old,
166-186 cm,
64-93 kg




10 male,
22-30 years old,
163-186 cm,
64-92 kg;
10 female,
22-35 years old,
149-177 cm,
48-81 kg

6 male, 2 1-29 years
old, 165 185 cm,
60-92 kg;
6 female,
19-33 years old,
152-173 cm,
48-61 kg

Region of
Interest
Central
conducting
airways
(70-120 mL
from lips)




Conducting
airways







Lower
conducting
airways
(70-120 mL
from lips)



Breathing
Patterns/Exposure
2 h of continuous exposure at
rest: 0.0, 0.12, and 0.36 ppm
O3. Spontaneous breathing.
Bolus test breaths every
30 minutes using 250 mL/sec
constant flow rate.



Bolus-response test
(VT = 500ml at 250 mL/sec
constant flow rate). Fowler
single-breath N2 washout
method to determine VD.




2 h of continuous exposure
at rest: O3 (0, 0.36 ppm),
SO2(0, 0.36 ppm), orNO2
(0, 0.36, 0.75 ppm).
5-min Bolus test every
30 minutes: VT = 500 ml;
250 mL/sec constant flow
rate.
Results Reference
Averaged over all subjects and the Asplund et al.
4 measurement intervals, the absorbed fraction (1996)
(AF) changed +0.04, -0.005, and -0.03 for the
0, 0. 12, and 0.36 ppm continuous exposures,
respectively. These changes are
approximately +6, -1, and -4 % based on an
average AF value of 0.7 in the range
70 -120 ml.b Both non zero exposures were
significantly different than the air exposure.
On average, for the same VP , women had a Bush et al.
larger AF than men; women had a smaller VD ( 1 996a)
than men. However, for the same value of
VP/VD, AF for men and women were
indistinguishable. Further analysis indicated
"that previously reported gender differences
may be due to a failure in properly accounting
for tissue surface within the conducting
airways".
Averaging over all subjects or by gender, all Rigas et al.
exposures except O3 resulted in an increase of (1997)
AF. Based on an AF reference valueb, the
change in AF ranged from -3 to +7 %. Only
the O3 and the NO2 (0.36 ppm) exposures
were significantly different from the air
exposures.

o
HH
H
W

-------
                                   Table AX4-1 (cont'd).  New Experimental Human Studies on Ozone Dosimetry'
1
to
o
o
Purpose/Objective
Compare the
absorption of
chlorine and O3.
Determine how the
physical-chemical
properties of these
compounds affect
their uptake
distribution in the RT
Subject
Characteristics
5 male, 2 1-26 years
old, 168-198 cm,
64-95 kg; 5 female,
18-28 years old,
162-178 cm,
55-68 kgc
Region of
Interest
Conducting
airways.
Nasal & oral
routes
Breathing
Patterns/Exposure
Bolus-response technique;
VT = 500 ml; 3 constant flow
rates: 150, 250, and
1000 mL/sec.
Results
Ozone dose to the URT was sensitive to the
mode of breathing and to the respiratory rate.
With increased airflow rate, O3 retained by the
upper airways decreased from 95 to 50%.
TB region dose ranged from 0 to 35%.
Mass transfer theory indicated that the
diffusion resistance of the tissue phase is
important for O3. The gas phase resistances
were the same for O3 and C12.
Reference
Nodelman
and Ultman
(1999a)c
>
         To determine O3
         uptake relative to
         inhaled O3 dose.
5 male,
5 female,
18-35 years old,
175 ± 13 (SD) cm,
72 ± 13 (SD) kg
Respiratory     Breath-by-breath calculation
tract; oral       of O3 retention based on data
breathing       from fast response analyzers
               for O3 and airflow rates. Oral
               breathing:  0.2 or 0.4 ppm O3
               at VE of approximately 20
               L/min for 60 min or 40 L/min
               for 30 min.
The FA for all breaths was 0.86.
Concentration, minute volume, and time have
small but statistically significant effects on AF
when compared to intersubject variability.
The investigators concluded: for a given
subject, constant O3 exposure, a given exercise
level, and time < 2 h, inhaled dose is a
reasonable surrogate for the actual uptake of
O3. However, the actual doses may vary
considerably among individuals who are
exposed to similar inhaled doses.
Rigas et al.
(2000)
H
6
o
o
H
O
o
HH
H
W

-------
                                   Table AX4-1 (cont'd). New Experimental Human Studies on Ozone Dosimetry!
to
o
o
Purpose/Objective
Subject
Characteristics
Region of
Interest
Breathing
Patterns/Exposure
Results
Reference
         Study the effect of
         gas flow rate and O3
         concentration on O3
         uptake in the nose.
                      7 male, 3 female,
                      26 ± years,
                      170 ± 11 (SD)cm,
                      75 ± 20  (SD) kg
                     Nasal          For a given flow rate and
                     cavities         exposure concentration, the
                                    subjects inhaled through one
                                    nostril and exhale through the
                                    other.  For two 1-h sessions,
                                    a series of 9-12 measurements
                                    of AF were carried out for
                                    10  s each:  (1) O3 exposure
                                    concentration = 0.4 ppm; flow
                                    rates = 3, 5, 8, and 15 L/min.
                                    (2) O3 exposure = 0.1, 0.2, and
                                    0.4 ppm; flow rate =
                                    15 L/min. (3) O3 exposure
                                    = 0.4 ppm, flow rate
                                    = 15 L/min; AF was measured
                                    every 5 min for 1 h.
                                             (1) With the exposure concentration at
                                             0.4 ppm O3, AF decreased from 0.80 to
                                             0.33 when the flow rate was increased from
                                             3 to 15 L/min.  (2) At a flow rate of 15 L/min,
                                             the AF changed from 0.36 to 0.32 when the
                                             exposure concentration increased from 0.1 to
                                             0.4 ppm O3. (3) Statistical analysis indicated
                                             that the AF was not associated with the time at
                                             which the measurement was taken.
                                                                          Santiago
                                                                          etal. (2001)
H
6
o
o
H
O
         Evaluate intersubject
         variability in O3
         uptake; correlate
         differences in
         breathing pattern and
         lung anatomy with O3
         uptake
                      nonsmokers,
                      32 male, 22.9 ± 0.8
                      years old,178±l cm,
                      80.6 ± 2.5kg;
                      28 female,
                      22.4 ± 0.9 years old,
                      166 ± 1 cm,
                      62.1 ±2.2 kg
                     Respiratory     Continuous:  1 h exposure to
                     tract; oral       0.25 ppm, exercising at
                     breathing       30L/min.
                                    Bolus: breath-by- breath
                                    calculation of O3 retention.
                                    Timing of bolus varied to
                                    create penetration volumes of
                                    10 to 250 ml. Peak inhaled
                                    bolus of ~1 ppm.
                                             Continuous:  Fractional O3 uptake efficiency
                                             ranged from 0.70 to 0.98 (mean 0.89 ± 0.06).
                                             Inverse correlation between uptake and
                                             breathing frequency. Direct correlation
                                             between uptake and tidal volume. Intersubject
                                             differences in forced respiratory responses not
                                             due to differences in O3 uptake.

                                             Bolus:  The penetration volume at which 50%
                                             of the bolus was taken up was 90.4 ml in
                                             females and 107 ml in males.  Distribution of
                                             O3 shifts distally as the size of the airway
                                             increases.
                                                                          Ultman et al.
                                                                          (2004)
o
HH
H
W
         a See Appendix A for abbreviations and acronyms.
         b Fig. 4, Hu et al. (1994), for the 250 mL/s curve and penetration volume range of 70 - 120 ml; the average AF is approximately 0.7.
         0 Subject characteristics are from Nodelman and Ultman (1999b).

-------
 1      hand, they conjectured that continuous O3 exposure depleted these compounds, thereby reducing
 2      O3 uptake.
 3          Bush et al. (1996a) investigated the effect of lung anatomy and gender on O3 absorption in
 4      the conducting airways during oral breathing using the bolus-response technique. Absorption
 5      was measured using this technique applied to 10 men and 10 women. Anatomy was defined in
 6      terms of forced vital capacity (FVC), total lung capacity (TLC), and dead space (VD).  The
 7      absorbed fraction data were analyzed in terms of a function of penetration volume, airflow rate,
 8      and an "intrinsic mass transfer parameter (Ka)", which was determined for each subject and
 9      found to be highly correlated with VD, but not with height, weight, age, gender, FVC, or TLC.
10      That is, in all subjects, whether men or women, dosimetry differences could be explained by
11      differences in VD.  Based on Hu et al. (1994), where absorbed fraction was determined for
12      several flow rates, Bush et al. (1996a) inferred that Ka was proportional to flow rate/VD. The
13      investigators point out that the applicability of their results may be limited because of their
14      assumptions that Ka was independent of location in the RT and that there was no mucous
15      resistance. They also suggested that the dependence of Ka on flow rate and VD be restricted to
16      flow rates < 1000 mL/s until studies at higher rates have been performed.
17          With flow rates of 150, 250, and 1000 mL/s, Nodelman and Ultman (1999b) used the
18      bolus-response technique to compare the uptake distributions of O3 and chlorine gas (C12), and to
19      investigate how their uptakes were affected by their physical and chemical properties. Ozone
20      dose to the URT was found to be sensitive to the mode of breathing and to the airflow rate.  With
21      increased rate, O3 retained by the upper airways decreased from 95 to 50% and TB region dose
22      increased from 0 to 35%. At the highest flow rate only 10% of the O3 reached the pulmonary
23      region. Mass transfer theory indicated that the diffusion resistance of the tissue phase is
24      important for O3.  The gas phase resistances were found to be the same for O3 and C12 as
25      expected. These resistances were inversely related to the volumes of the oral and nasal cavities
26      during oral and nasal breathing, respectively.
27          Ultman et al. (2004) used both bolus and continuous exposures to test the hypotheses that
28      differences in O3 uptake in lungs are responsible for variation in O3-induced changes in lung
29      function parameters and that differences in O3 uptake are due to variations in breathing patterns
30      and lung anatomy. Thirty-two males and 28 female nonsmokers were  exposed to bolus
31      penetration volumes ranging from 10 to 250  ml, which was determined by the timing of the

        January 2005                             AX4-7      DRAFT-DO NOT QUOTE OR CITE

-------
 1      bolus injection.  The subjects controlled their breathing to generate a target respired flow of
 2      lOOOml/sec.  At this high minute ventilation, there was very little uptake in the upper airway and
 3      most of the O3 reached areas where gas exchange takes place.  To quantify intersubject
 4      differences in O3 bolus uptake, they measured the penetration volume at which 50% of the O3
 5      was taken up. Values for penetration volume ranged from 69 to  134 ml and were directly
 6      correlated with the subjects' values for anatomic dead space volume. A better correlation was
 7      seen when the volume of the upper airways was subtracted. The penetration volume at which
 8      50% of the bolus was taken up was 90.4 ml in females and 107 ml in males. This significant
 9      difference in uptake suggests to the authors that in females the smaller airways, and associated
10      larger surface-to-volume ratio, enhance local O3 uptake and cause reduced penetration of O3 into
11      the distal lung. Thus, these findings indicate that overall O3 uptake is not related to airway size,
12      but that the distribution of O3 shifts distally as the size of the airway in increased.
13
14      General comment on estimating mass transfer coefficients.  Bush et al. (1996b) and Nodelman
15      and Ultman (1999a) used a simple model to analyze their bolus-response data. This model
16      presented by Hu et al. (1992,  1994) assumed steady-state mass transfer by convection (but no
17      dispersion) and the mass transfer of O3 to the walls of a tube of uniform cross-sectional area.
18      These assumptions led to an analytical solution (for the absorbed fraction) which was a function
19      of an "overall mass transfer coefficient," penetration volume, and airflow rate. As the
20      investigators have shown, the model is very useful for statistical  analysis and hypothesis testing.
21      Given the absorbed fraction data, the model overall mass transfer coefficients  were estimated for
22      each flow rate. In those bolus-response studies that used this method to analyze data, there was
23      no discussion of the models' "accuracy" in representing mass transfer in the human respiratory
24      tract with respect to omitting  dispersion. In addition, the formulation of the gas phase mass
25      transfer coefficient does not take into account that it has a theoretical lower limit greater than
26      zero as the airflow rate goes to zero (Miller et al., 1985; Bush et al., 2001). As a consequence,
27      there is no way to judge the usefulness of the values of the estimated mass transfer coefficients
28      for dosimetry simulations that are based on convection-dispersion equations, or whether or not
29      the simple model's mass transfer coefficients, as well as other parameters derived using these
30      coefficients, are the same as actual physiological parameters.
31

        January 2005                              AX4-8       DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX4.2.2  General Uptake Studies
 2           Rigas et al. (2000) performed an experiment to determine the ratio of O3 uptake to the
 3      quantity of O3 inhaled (fractional absorption, FA). Five men and five women were exposed
 4      orally to 0.2 or 0.4 ppm O3 while exercising at a minute volume of approximately 20 L/min for
 5      60 minutes or 40 L/min for 30 minutes. Ozone retention was calculated from breath-by-breath
 6      data taken from fast response analyzers of O3 and airflow rates. The FA was statistically
 7      analyzed in terms of subject, exposure concentration, minute volume, and exposure time.
 8           Fractional absorption ranged from 0.56 to 0.98 with a mean ± SD of 0.85 ± 0.06 for all
 9      recorded breaths. Intersubject differences had the largest influence on FA, resulting in a
10      variation of approximately 10%.  Statistical analysis indicated that concentration, minute
11      volume, and exposure time had statistically significant effects on FA. However, relatively large
12      changes in these variables were estimated to result in relatively small changes in FA. Note: the
13      quantity of O3 retained by the RT is equal to FA times the quantity of O3 inhaled; thus, relatively
14      large changes in concentration, minute volume, or exposure time may result in relatively large
15      changes in the amount of O3 retained by the RT or absorbed locally. Also, according to Overton
16      et al. (1996), difference in PAR dose due to anatomical variability may be considerably larger
17      than corresponding small changes in FA would indicate.
18           Santiago et al. (2001) studied the effects of airflow rate and O3 concentration on O3 uptake
19      in the nasal cavities of three women and seven men.  Air was supplied at a constant flow rate to
20      one nostril and exited from the other nostril while the subject kept the velopharyngeal aperture
21      closed by raising the soft palate.  Thus, a constant unidirectional flow of air plus O3 was
22      restricted to the nasal cavities. The fraction of O3 absorbed was calculated using the inlet and
23      outlet concentrations. Inlet concentration and airflow rate were varied in order to determine their
24      effect on O3 uptake.
25           The mean FA decreased from 0.80 to 0.33  with an increase in flow rate from 3 to
26      15 L/min.  The effect of both flow rate and subject on FA was statistically significant. Further
27      analysis indicated that the overall mass transfer coefficient was highly correlated with the flow
28      rate and that the gas phase resistance contributed from  6.3% (15 L/min) to 23% (3 L/min) of the
29      total resistance to O3 transfer to the nasal cavity  surface. Concentration had a small, but
30      statistically significant effect on FA, when the inlet concentration was increased from 0.1  to
31      0.4 ppm O3, FA decreased from 0.36 to 0.32. The investigators observed that differences in FA

        January 2005                             AX4-9        DRAFT-DO NOT QUOTE OR CITE

-------
 1      among subjects were important; generally, subject variability accounted for approximately half
 2      of the total variation in FA.
 3           As mentioned above Ultman et al. (2004) tested hypotheses that differences in O3 uptake in
 4      lungs are responsible for variation in O3-induced changes in lung function parameters and that
 5      differences in O3 uptake are due to variations in breathing patterns and lung anatomy.  Thirty-
 6      two males and 28 female nonsmokers were exposed continuously for 1 h to either clean air or
 7      0.25 ppm ozone while exercising at a target minute ventilation of 30 L/min.  They first
 8      determined the forced expiratory response to clean air, then evaluated O3 uptake measuring dead
 9      space volume, cross-sectional area of peripheral lung (Ap) for CO2 diffusion, FEVb FVC, and
10      FEF25o/0.75o/0.  The fractional O3 uptake efficiency ranged from 0.70 to 0.98, with a mean of
11      0.89 ± 0.06.  They found an inverse correlation between uptake and breathing frequency and a
12      direct correlation between uptake and tidal volume.  The uptake efficiency decreased during the
13      four sequential 15  minute intervals of the 1 h exposure (0.906 ± 0.058 to 0.873 ± 0.088, first and
14      last, respectively), demonstrating a general decrease in uptake efficiency with increased
15      breathing frequency and decreasing tidal volume. Ozone uptake rate correlated with individual
16      %Ap, but did not correlate with individual "/oFEVj. Neither of these parameters correlated with
17      the penetration volume determined in the bolus studies mentioned above.  The authors concluded
18      that the intersubject differences in forced respiratory responses were not due to differences in O3
19      uptake. However, these data did partially support the second hypothesis, i.e., that the differences
20      in cross-sectional area available for gas diffusion induce differences in O3 uptake.
21
22
23      AX4.3  DOSIMETRY MODELING
24           When all of the animal and human in vivo O3 uptake efficiency data are compared, there is
25      a good degree of consistency across  data sets (U.S. Environmental Protection Agency, 1996).
26      This agreement raises the level of confidence with which these data sets can be used to support
27      dosimetric model formulations.
28           Recent data indicate that the primary site of acute cell injury occurs in the conducting
29      airways (Postlethwait et al., 2000).  These data must be considered when developing models that
30      attempt to predict site-specific locations of O3-induced injury. The early models computed
31      relationships between delivered regional dose and response with the assumption that O3 was the

        January 2005                            AX4-10      DRAFT-DO NOT  QUOTE OR CITE

-------
 1      active agent responsible for injury.  It is now known that reactive intermediates such as
 2      hydrohydroxyperoxides and aldehydes are important agents mediating the response to O3
 3      (further discussed in Section 5.3.1).  Thus, models must consider O3 reaction/diffusion in the
 4      epithelial lining fluid (ELF) and ELF-derived reactions products.
 5           Table AX4-2 presents a summary of new theoretical studies on the uptake of O3 by the RTs
 6      (or regions) of humans and laboratory animals that have become available since the 1996 review.
 7      They are discussed below.
 8           Overton and Graham (1995) described the development and simulation results of a
 9      dosimetry model that was applied to a TB region anatomical model that had branching airways,
10      but which had identical single-path pulmonary units distal to each terminal bronchiole.  The
11      anatomical model of the TB region  was based on Raabe et al. (1976), which reported lung cast
12      data for the TB region of a 330 g rat.
13           Rat effects data (from the PAR) are available that are identified with the lobe and the
14      generation in the lobe from which tissue samples were obtained (Pinkerton et al.,  1995,  1998).
15      Models, like Overton et al. (1995), can be helpful in understanding the distribution of the
16      magnitude of such effects as well as suggesting sampling sites for future experiments.
17           Using computational fluid dynamics (CFD), Cohen-Hubal  et al. (1996) explored the effect
18      of the mucus layer thickness in the nasal passage of a rat. The nasal lining was composed of
19      mucus and tissue layers in which mass transfer was by molecular diffusion with first order
20      chemical reaction.  Physicochemical parameters for O3 were obtained from the literature.  Three
21      scenarios were considered: 10 jim thick mucus layer, no mucus  layer, and two nasal passage
22      regions each with a different mucus layer thickness.  Predictions of overall uptake were  within
23      the range of measured uptake. Predicted regional O3 flux was correlated with measured cell
24      proliferation for the CFD simulation that incorporated two regions, each with a different mucus
25      thickness.
26           The reaction rate  constant used by Cohen-Hubal and co-workers may be too low.  Using
27      bolus-response data, Hu et al. (1994) and Bush et al. (2001) estimated a reaction rate constant
28      that is more than a 1000 times as large as that used by Cohen-Hubal et al. (1996). A rate
29      constant this large could result in a  conclusion different than those based on the smaller  constant.
30           With an RT dosimetry model, Overton et al. (1996) investigated the sensitivity of absorbed
31      fraction (AF), proximal alveolar region (PAR) dose, and PAR dose ratio to TB region volume

        January 2005                             AX4-11       DRAFT-DO NOT QUOTE OR CITE

-------
                                                   Table AX4-2. New Ozone Dosimetry Model Investigations"
to
o
o
Purpose/Objective
Type of mass transport
model/Anatomical modelb
Species/ RT region
of interest/Regional
anatomical models
Ventilation and
Exposure
                                                                                                         Results
Reference
to
 H
 6
 o
 o
 H
O
          To describe an RT
          dosimetry model that
          uses a branching TB
          region anatomical
          model and to
          illustrate the results
          of its application to a
          rat exposed to O3.
To incorporate into
the CFD model of
Kimbelletal. (1993)
resistance to mass
transfer in the nasal
lining and to
investigate the
effects of this lining
on O3 uptake.


To determine if the
single-path model is
able to simulate
bolus inhalation data
recorded during oral
breathing at quiet
respiratory flow.
                      One-dimensional (along axis
                      of airflow), time-dependent,
                      convection-dispersion
                      equation of mass transport
                      applied to each airway or
                      model segment. URT: single
                      path; TB: asymmetric
                      branching airways. PUL:
                      single path anatomical model
                      distal to each terminal
                      bronchiole.
Three dimensional steady-
state Navier-Stokes equations
for solving air velocity flow
field. Three dimensional
steady-state convection-
diffusion equation for O3
transport.
Three-dimensional CFD
model of the nasal passages
of a rat.

Single-path, one-dimensional
(along axis of airflow),
time-dependent, convection-
dispersion equation of mass
transport.  Single-path
anatomical model
                              Rat/ RT/URT:
                              Patraetal. (1987).
                              TB: multiple path
                              model of Raabe et al.
                              (1976). PUL:
                              Mercer etal. (1991).
                                                             Rat/nasal passages
                                                             Nasal passages:
                                                             Kimbelletal. (1993).
                                                             Human/ RT/URT
                                                             (oral): Olson et al.
                                                             (1973).
                                                             LRT: Weibel(1963).
                      f= ISObpm;
                      VT=1.5,2.0,
                      2.5 mL.
                      One constant
                      concentration.
                       Steady-state
                       unidirectional
                       ("inhalation")
                       flow rate
                       = 576mLl/min.
                       One constant
                       concentration.
                      VT = 500 mL,
                      f= 15bpm,
                      constant flow
                      rate = 250 mL/s.
                      Bolus-response
                      simulations.
                      (protocol used is
                      described by Hu
                      etal., 1992).
                    (1) For VT = 2.0 mL, f = 150 bpm:  The general shape
                    of the dose versus generation plot along any path from
                    the trachea to a sac is independent of path: generally
                    the tissue dose decreases with increasing generation
                    index.  In the TB region, the coefficient of variation for
                    dose ranges from 0 to 34 %, depending on generation.
                    The maximum ratio of the largest to smallest dose in
                    the same generation is 7; the average ratio being 3. In
                    the first PUL region model segment, the coefficient of
                    variation for the dose is 29  %.  (2) The average dose to
                    the first PUL region model segment increases with
                    increasing VT.

                    Predictions of overall uptake were within the range of
                    measured uptake.  Results suggest that mucus resistance
                    is important for describing  O3 dosimetry and this
                    thickness may play a role for determining patterns of
                    O3-induced lesions in the rat nasal passage.
                    Simulations are sensitive to conducting airway volume
                    but are relative insensitive to characteristics of the
                    respiratory airspace. Although the gas-phase resistance
                    to lateral diffusion limits O3 absorption during quiet
                    breathing, diffusion through mucus may become
                    important at the large respiratory flows that are
                    normally associated with exercise. The single-path
                    convection-diffusion model was a reasonable approach
                    /to simulate the bolus-response data.
                                                                                                                                                             Overton and
                                                                                                                                                             Graham
                                                                                                                                                             (1995)
                                                                                                                                                             Cohen-
                                                                                                                                                             Hubal et al.
                                                                                                                                                             (1996)
                                                                                                                                                             Bush et al.
                                                                                                                                                             (1996b)
o
HH
H
W

-------
                                             Table AX4-2 (cont'd). New Ozone Dosimetry Model Investigations a
c
1
to
o
o
Purpose/Objective
To assess age- and
gender-specific
differences in
regional and
systemic uptake.
Type of mass transport
model/Anatomical model b
PBPK, at ages 1, 3, 6, months
and 1,5, 10, 15,25, 50, and
75 years
Species/ RT region
of interest/Regional
anatomical models
Human/ET/TB
Ventilation and
Exposure
Pulmonary
ventilation ranged
from 34 mL/s (in
1 -month-old) to
190 mL/s (in
15-year-old). VT
varied with age
Results
Regional extraction is insensitive to age. Extraction per
unit surface area is 2- to 8-fold higher in infants
compared to adults. PU and ET regions have a large
increase in unit extraction with increasing age. Early
postnatal period is time of largest differences in PK, due
to immaturity of metabolic enzymes.
Reference
Sarangapani
et al. (2003)
 H
 6
 o
 o
 H
O
          To examine the
          impact on predictions
          due to the value used
          for the TB region
          volume at FRC and
          due to TB region
          volume change
          during respiration.
To make parameter
modifications so that
a single-path model
would simulate AF
from bolus-response
experiments
involving O3
(and C12).
                      Single-path, one-dimensional
                      (along axis of airflow),
                      time-dependent, convection-
                      dispersion equation of mass
                      transport. Single-path
                      anatomical model
                                Single-path, one-dimensional
                                (along axis of airflow), time-
                                dependent, convection-
                                dispersion equation of mass
                                transport.  Single-path
                                anatomical model
Human /RT/
URT: Nunnetal.
(1959)
LRT: Weibel(1963)

Rat/RT/
URT: Patraetal.
(1987)
TB:  Yehetal. (1979)
PUL: Mercer etal.
(1991)

Human/ RT/URT
(oral): Olson et al.
(1973).
URT (nasal):
Olson etal. (1973)
and Guilmette et al.
(1989)
LRT: Weibel(1963).
Human: VT = 500,
2250mL;f=15,
30 bpm.

Rat:  VT=1.4,
2.4 ml; f= 96,
157 bpm.

One constant
concentration.
Oral & nasal
breathing. Flow
rates = 150,250,
1000 mL/s,
VT = 500 mLl.
Bolus-response
simulations
(1) A better understanding and characterization of the
role of TB region expansion (mainly the rat) and
volume is important for an improved understanding of
respiratory-tract dosimetry modeling of reactive gases.
(2) Extrapolations based on dose in the PAR can differ
significantly from those based on exposure
concentration or total uptake.
(3) Human subjects who appear similar outwardly may
have very different PAR doses and potentially different
responses to the same exposure.(Uptake by the URT
was not considered.)

(Simulation results for O3 only) (1) Using parameter
values from the literature and assuming that absorption
was gas-phase controlled, the simulations of O3 data
were realistic  at flow rate = 250 mL/s, but not realistic
at 1000 mL/s.(2) Accurate simulations  at 250 mL/s
required modification of mass transfer coefficients
reported in the literature for the conducting airways.(3)
It was necessary to include a diffusion resistance for the
epithelial lining fluid based on an assumed O3 reaction
rate constant that was much greater than in in vitro
estimates.(4) Partial validation of the final parameters
(determined at 250 mL/s) was obtained by simulations
of bolus-response data at flow rates of 150 and 1000
mL/s. Validation was obtained also by simulating
internal measurements of O3 in subjects exposed during
quiet breathing.
Overton
etal. (1996)
Bush et al.
(2001)
o
HH
H
W
"See Appendix A for abbreviations and acronyms.
bThe anatomical models used in an investigation generally differ from those described in the references, e.g., dimensions are often scaled to dimensions appropriate to the
 dosimetry investigation; or the original structure may be simplified, keeping or scaling the original dimensions.

-------
 1      (VTB) and TB region expansion in human beings and rats. The PAR was defined as the first
 2      generation distal to terminal bronchioles and the PAR dose ratio was defined as the ratio of a
 3      rat's predicted PAR dose to a human's predicted PAR dose.  This ratio relates human and rat
 4      exposure concentrations so that both species receive the same PAR dose. In rats the PAR is a
 5      region of major damage from O3. For each species, three literature values of V^ were used:
 6      a mean value and the mean ± twice the SD. The following predictions were obtained:
 7           (1) The sensitivity of AF and PAR dose to V^ depends on species, ventilation, TB region
 8      overall mass transfer coefficient (!%,), and expansion. Depending on these  latter four
 9      parameters,  AF was predicted to be  1 to 25% smaller and 1 to 40% larger than the values
10      predicted for the mean V^, given the range of V^. However, AF can be insensitive to Vra and
11      PAR dose very sensitive. For kTB = 0.26 cm/s and quiet breathing, AF  was  predicted to vary by
12      less than 3% for the ±2 SD range of VTB; in contrast, the PAR dose predicted using the smallest
13      VXB is five times larger than the PAR dose predicted with the largest V^. The effect of VTB is
14      much less during heavy exercise: the ratio of maximum to minimum PAR dose was
15      approximately 1.5. In any case, the simulations predicted that fractional changes in AF due to
16      different V^ are not, in general, a good predictor of the fractional changes in PAR doses.
17           (2) Relative to no expansion in the TB region, expansion decreases both AF and PAR
18      dose.  The largest effect of including expansion in the human simulations was to decrease the AF
19      by «8%; in rats, the maximum decrease was -45%. The PAR doses decreased relatively more,
20      25 and 65% in human beings and rat, respectively.
21           (3) The authors attempted to obtain an understanding as to uncertainty or variability in
22      estimates of exposure concentrations (that give the same PAR dose in both  species) if the
23      literature mean value of VTB was used. For various values off, VT, kra, and expansion, the  PAR
24      dose ratios at upper and lower values of VTB deviated in absolute values from the PAR dose ratio
25      calculated at the mean values of VTB by as little as 10% to as large as 310%. The smallest
26      deviation occurred at the largest VT and smallest kra for both species; whereas, the largest
27      deviation occurred at the smallest VT and largest kra for both species.
28           Bush et al. (2001) modified the single-path model of Bush et al. (1996b) in order to be able
29      to simulate absorbed  fraction data for O3 (and C12, which is not considered) for three airflow
30      rates  and for oral and nasal breathing.  By adjusting several parameters a reasonable agreement
31      between predicted and experimental values was obtained. On the other hand, the O3 plots of the

        January 2005                            AX4-14      DRAFT-DO NOT QUOTE OR CITE

-------
 1      experimental and predicted values of absorbed fraction versus penetration volume (e.g., Figures
 2      4 and 5 of Bush et al., 2001) show sequential groups composed of only positive or only negative
 3      residuals, indicating a lack of fit. Possibly adjusting other parameters would eliminate this.
 4      To obtain an independent validation of the model, Bush et al. (2001) simulated measurements of
 5      O3 concentrations made by Gerrity et al. (1995) during both inhalation and exhalation at four
 6      locations between the mouth and the bronchus intermedius of human subjects. Simulated and
 7      experimental values obtained are in close agreement. Note, however, that Bush et al. made no
 8      quantitative assessment of how well their simulations agreed with the experimental data;
 9      assessments were made on the basis of visual inspection of experimental and simulated values
10      plotted on the same figure. Thus, evaluation of the model was, or is, subjective.
11          Recently Sarangapani et al. (2003) used physiologically based pharmacokinetic (PBPK)
12      modeling to characterize age- and  gender-specific differences in both regional and systemic
13      uptake of O3 in humans. This model indicated that regional extraction of O3 is relatively
14      insensitive to age, but extraction per unit surface area is 2- to 8-fold higher in infants compared
15      to adults, due to the region-specific mass transfer coefficient not varying with age. The PU and
16      ET regions have a large increase in unit extraction with increasing age because both regions
17      increase in surface area. Males and females in this model have similar trends in regional
18      extraction and regional unit extraction. In early childhood, dose metrics were as much as
19      12 times higher than adult levels, but these differences leveled out with age, such that inhalation
20      exposures varied little after age 5.  These data suggest that the early postnatal period is the time
21      of the  largest difference in pharmacokinetics observed, and this difference is primarily due to the
22      immaturity of the metabolic enzymes used to clear O3 from the respiratory tract.
23          Mudway and Kelly (2004) attempted to model O3 dose-inflammatory response using a
24      metaanalysis of 23 exposures in published human chamber studies.  The O3 concentrations
25      ranged from 0.08 to 0.6 ppm and the exposure durations ranged from 60 to 396 minutes. The
26      analysis showed linear  relationships between O3 dose and neutrophilia in bronchoalveolar lavage
27      fluid (BALF). Linear relationships were also observed between O3 dose and protein leakage
28      into BALF.
29
30
        January 2005                            AX4-15       DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX4.4  SPECIES HOMOLOGY, SENSITIVITY AND ANIMAL-TO-
 2              HUMAN EXTRAPOLATION
 3          Biochemical differences among species are becoming increasingly apparent and these
 4     differences may factor into a species' susceptibility to the effects of O3 exposure. Lee et al.
 5     (1998) compared SD rats and rhesus monkeys to ascertain species differences in the various
 6     isoforms of CYP moonoxygenases in response to O3 exposure (discussed in more detail  in
 7     Section 5.3.1.2). Differences in activities between rat and monkey were 2- to 10-fold, depending
 8     on the isoform and the specific lung region assayed. This study supports the view that
 9     differential expression of CYPs is a key factor in determining the toxicity of O3. As further
10     characterization of species- and region-specific CYP enzymes occurs, a greater understanding of
11     the differences in response may allow more accurate extrapolation from animal exposures to
12     human exposures and toxic effects.
13          Arsalane et al. (1995) compared guinea pig and human AM recovered in BALF and
14     subsequently exposed in vitro to 0.1 to  1 ppm for 60 minutes.  Measurement of inflammatory
15     cytokines showed a peak at 0.4 ppm in both species. Guinea pig AM had an increase in IL-6 and
16     TNF-a while human AM had increases in TNF-a, IL-lb, IL-6 and IL-8. This exposure also
17     caused an increase in mRNA expression for TNF-a, IL-lb, IL-6 and IL-8 in human cells.
18     At 0.1 ppm exposures, only TNF-a secretion was increased.  These data suggest similar cytokine
19     responses in guinea pigs and humans, both qualitatively and quantitatively.
20          Dormans et al. (1999) continuously exposed rats, mice, male guinea pigs to filtered air, 0.2,
21     or 0.4 ppm for 3, to 56 days or to 28 days with 3, 7, and 28 days PE.  Depending on the endpoint
22     studied, the species varied in sensitivity. Greater sensitivity was shown in the mouse as
23     determined by biochemical endpoints, persistence of bronchiolar epithelial hypertrophy, and
24     recovery time. Guinea pigs were more  sensitive in terms of the inflammatory response though
25     all three species had increases in the inflammatory response after three days that did not decrease
26     with exposure. In all species the longest exposure to the highest dose caused increased collagen
27     in ductal septa and large lamellar bodies in Type II cells, but that response also occurred in rats
28     and guinea pigs at 0.2 ppm.  No fibrosis was seen at the shorter exposure times and the authors
29     question whether fibrosis occurs in healthy humans after continuous exposure. The authors do
30     not rule out the possibility that some of these differences may be attributable to differences in
       January 2005                            AX4-16      DRAFT-DO NOT QUOTE OR CITE

-------
 1      total inhaled dose or dose actually reaching a target site. Overall, the authors rated mice as most
 2      susceptible, followed by guinea pigs and rats.
 3           Comparisons of airway effects in rats, monkeys and ferrets resulting from exposures of
 4      1.0 ppm O3 for 8 h (Sterner-Kock et al. 2000) demonstrated that monkeys and ferrets had a
 5      similar inflammatory responses and epithelial necrosis. The response of these two species was
 6      more severe than that seen in rats. These data suggest that ferrets are a good animal model for
 7      O3-induced airway effects due to the similarities in pulmonary structure between primates and
 8      ferrets.
 9           The rat is a key species used in O3 toxicological studies, but Watkinson and Gordon,
10      (1993) suggest that, because the rat has both behavioral and physiological mechanisms that can
11      lower core temperature in response to acute exposures, extrapolation of these exposure data to
12      humans may be limited. Another laboratory (Iwasaki et al., 1998) has demonstrated both
13      cardiovascular and thermoregulatory responses to O3 at exposure to 0.1, 0.3, and 0.5 ppm O3
14      8 h/day for 4 consecutive days.  A dose-dependent disruption of HR and Tco were seen on the
15      first and second days of exposure, which then recovered to control values. Watkinson et al.
16      (2003) exposed rats to 0.5 ppm O3 and observed this hypothermic response which included
17      lowered HR, lowered Tco, and increased inflammatory components in BALF. The authors
18      suggest that the response is an inherent reflexive pattern that can possibly attenuate O3 toxicity in
19      rodents.  They discuss the cascade of effects created by decreases in Tco, which include:
20      (1) lowered metabolic rate,  (2) altered enzyme kinetics, (3) altered membrane function,
21      (4) decreased  oxygen consumption and demand, (5) reductions in minute ventilation, which
22      would act to limit the dose of O3 delivered to the lungs. These effects are concurrent with
23      changes in HR which lead to: (1) decreased CO, (2) lowered BP, (3) decreased tissue perfusion,
24      all of which may lead to functional deficits.  The hypothermic response has not been observed in
25      humans except at very high exposures.
26
        January 2005                             AX4-17      DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Arsalane, K.; Cosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
  3             synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
  4             13:60-68.
  5       Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S. (1996) Longitudinal distribution of ozone absorption in
  6             the lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
  7       Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996a) Longitudinal distribution of O3
  8             absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
  9       Bush, M. L.; Raybold, T.; Abeles, S.; Hu, S.-C; Ben-Jebria, A.; Ultman, J. S. (1996b) Longitudinal distribution of
10             ozone absorption in the lung: simulation with a single-path model. Toxicol. Appl. Pharmacol. 140: 219-226.
11       Bush, M. L.; Zhang, W.; Ben-Jebria, A.; Ultman, J. S. (2001) Longitudinal distribution of ozone and chlorine in the
12             human respiratory tract: simulation of nasal and oral breathing with the single-path diffusion model. Toxicol.
13             Appl. Pharmacol. 173: 137-145.
14       Cohen-Hubal, E. A.; Kimbell, J. S.; Fedkiw, P. S. (1996) Incorporation of nasal-lining mass-transfer resistance into a
15             CFD model for prediction of ozone dosimetry in the upper respiratory tract. Inhalation Toxicol. 8: 831-857.
16       Dahl, A. R. (1990) Dose concepts for inhaled vapors and gases. Toxicol. Appl. Pharmacol. 103: 185-197.
17       Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
18             time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
19       Gerrity, T. R.; Biscardi, F.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995) Bronchoscopic
20             determination of ozone uptake in humans. J. Appl. Physiol. 79: 852-860.
21       Hu, S. C.; Ben-Jebria, A.; Ultman, J. S. (1992) Longitudinal distribution of ozone absorption in the lung: quiet
22             respiration in healthy subjects. J. Appl. Physiol. 73:  1655-1667.
23       Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
24             respiratory flow. J. Appl. Physiol. 77:  574-583.
25       Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
26             conscious rats. Ind. Health 36: 57-60.
27       Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
28             differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
29             and induction in rats. Drug Metab. Dispos. 26: 396-400.
3 0       Miller, F. J. (1995) Uptake and fate of ozone in the respiratory tract. Toxicol. Lett. 82/83: 277-285.
31       Miller, F. J.; Overton, J. H., Jr.; Jaskot, R. H.; Menzel, D. B. (1985) A model of the regional uptake of gaseous
3 2             pollutants in the lung: I. the sensitivity of the uptake of ozone in the human lung to lower respiratory tract
33             secretions and exercise. Toxicol. Appl. Pharmacol. 79: 11-27.
34       Mudway, I. S.; Kelly, F. J. (2004) An investigation of inhaled ozone dose and the magnitude of airway inflammation
35             in healthy adults. Am. J. Respir. Crit. Care Med. 169: 1089-1095.
36       Nodelman, V.; Ultman, J. S. (1999) Longitudinal distribution of chlorine absorption in human airways: comparison
37             of nasal and oral quiet breathing. J. Appl. Physiol. 86: 1984-1993.
38       Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat respiratory tract model with
39             an asymmetric tracheobronchial region patterned on complete conducting airway cast data. Comput. Biomed.
40             Res. 28: 171-190.
41       Overton, J. H.; Graham, R. C.; Menache, M.  G.; Mercer, R. R.; Miller, F. J. (1996) Influence of tracheobronchial
42             region expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhalation Toxicol.
43             8:723-745.
44       Pinkerton, K. E.; Menache, M. G.; Plopper, C. G. (1995) Consequences of prolonged inhalation of ozone on F344/N
45             rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary acinus, and lung
46             antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98; research report no. 65.
47             Available from: NTIS, Springfield, VA; PB95-261996.
48       Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
49             ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
50             epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
51             research report no. 65.
52       Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir, A. J.; Putney, L.  F.; Wong, V. J.;
53             Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
54             tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
         January 2005                                  AX4-18        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Raabe, O. G.; Yeh, H. C.; Schum, G. M; Phalen, R. F. (1976) Tracheobronchial geometry: human, dog, rat,
  2             hamster. Albuquerque, NM: Lovelace Foundation; report no. LF-53.
  3       Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
  4             of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52: 173-178.
  5       Rigas, M. L.; Catlin, S. N.; Ben-Jebria, A.; Ultman, J. S. (2000) Ozone uptake in the intact human respiratory tract:
  6             relationship between inhaled dose and actual dose. J. Appl. Physiol. 88: 2015-2022.
  7       Santiago, L. Y.; Harm, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001) Ozone adsorption in the human nose during
  8             unidirectional airflow. J. Appl. Physiol. 91: 725-732.
  9       Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G.; Clewell, H. J., III. (2003) Evaluation of the
10             potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors.
11             Inhalation Toxicol. 15: 987-1016.
12       Sterner-Kock, A.; Kock, M.; Braun,  R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar
13             to nonhuman primates. Am. J. Respir. Crit. Care  Med. 162: 1152-1156.
14       U.S. Environmental Protection Agency. (1996) Air quality  criteria for ozone and related photochemical oxidants.
15             Research Triangle Park, NC: Office of Research  and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
16             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
17             www.epa.gov/ncea/ozone.htm.
18       Watkinson, W. P.; Aileru, A. A.; Dowd, S. M.; Doerfler, D. L.;  Tepper, J. S.; Costa, D. L. (1993) Acute effects of
19             ozone on heart rate and body temperature in the unanesthetized, unrestrained rat maintained at different
20             ambient temperatures. Inhalation Toxicol. 5: 129-147.
21       Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.;  Costa, D. L. (2003) Cardiac and thermoregulatory
22             responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
23             environmental factors. Environ. Res. 92: 35-47.
24
25
         January 2005                                 AX4-19       DRAFT-DO NOT QUOTE OR CITE

-------
 i           5.  TOXICOLOGICAL EFFECTS OF OZONE AND
 2             RELATED PHOTOCHEMICAL OXIDANTS IN
 3                LABORATORY ANIMALS AND IN VITRO
 4                                   TEST SYSTEMS
 5
 6
 7     5.1    INTRODUCTION
 8          A wide range of effects of ozone (O3) has been demonstrated in laboratory animals.  The
 9     major research findings are that environmentally relevant levels of O3 cause lung inflammation;
10     decreases in host defenses against infectious lung disease; acute changes in lung function,
11     structure, and metabolism; chronic lung disease, some elements of which are irreversible; and
12     systemic effects on target organs (e.g., brain, heart, liver, immune system)  distant from the lung.
13     The research also has served to expand the understanding of mechanisms of O3 toxicity and the
14     relationships between concentration and duration of exposure.
15          The framework for presenting the health effects of O3 in animals begins with a presentation
16     of respiratory tract effects, followed by systemic effects, and then interactions of O3 with other
17     common co-occurring pollutants.  The information discussed in this chapter is founded on a very
18     wide body of literature on studies in laboratory animals and on in vitro test systems of animal
19     cell lines and organ systems that may mimic responses in intact animals. The direct effects of O3
20     in humans are discussed in the following chapter (Chapter 6).
21          This chapter is  not intended to be a compendium of all that is known about O3; rather, it is
22     an update of the toxicology chapter from the last O3 criteria document (U.S. Environmental
23     Protection Agency, 1996), or 1996 O3 CD, and other reviews of the earlier published literature.
24     The historical O3 literature is very briefly summarized in an opening paragraph of each section or
25     subsection. This paragraph is intended as a very concise overview of previous work, and the
26     reader is referred to the 1996 O3 CD for more detailed discussion of the literature prior to the
27     early 1990's. Each section then continues with brief discussions of the key new studies (or
28     somewhat older studies that were not included in the previous CD).  Longer discussions of new
29     studies are included where warranted. Sections are ended with comparisons of data from the
       December 2005                           5-1        DRAFT-DO NOT QUOTE OR CITE

-------
 1      previous CD with new data and basic conclusions are drawn.  More detailed descriptive
 2      summaries of new studies and results are provided in text and tables in Annex AX5.
 3           Except for nitrogen dioxide (NO2), the subject of another criteria document (U.S.
 4      Environmental Protection Agency, 1993), there is very little relevant information on other
 5      photochemical oxidants in the published literature.  What is known about the effects of these
 6      other oxidants is also summarized briefly in this chapter.
 7
 8
 9      5.2   RESPIRATORY TRACT EFFECTS OF OZONE
10      5.2.1   Biochemical Effects
11           Biochemically detected effects of O3 are integrally involved in effects on both structure and
12      function (respiratory and nonrespiratory) of the respiratory tract.  Changes in xenobiotic
13      metabolism, antioxidant metabolism and oxygen consumption, lipids and arachidonic acid
14      metabolism, and collagen metabolism are all observed with O3 exposure, though the mechanisms
15      and associations are not fully understood.
16
17      5.2.1.1   Cellular Targets of O3 Interaction
18           Ozone has the potential to interact with a wide range of different cellular components that
19      include polyunsaturated fatty acids (PUFAs); some protein amino acid residues (cysteine,
20      histidine, methionine, and tryptophan); and some low-molecular-weight compounds that include
21      glutathione (GSH), urate, vitamins C and E, and free amino acids. Early work demonstrated that
22      O3 being a highly reactive compound, does not penetrate much beyond the epithelial lining fluid
23      (ELF).  Ozone-induced cell damage most likely results from its reactions with PUFAs to form
24      stable but less reactive ozonide, aldehyde, and hydroperoxide reaction products (the reactions
25      are summarized in Figure 5-1 of Annex AX5).  These reaction products (Crigee ozonides and
26      hydroxyhydroperoxides) may act as signal transduction  molecules involved in signaling of
27      cellular responses such as inflammation, and thus mediate O3 toxicity.
28           These recent reports combined with observations reported in the previous O3 CD (US EPA,
29      1996) suggest that interactions of O3 with cellular components and ELF generate toxic ozonation
30      products and mediate toxic effects through these products.
31

        December 2005                             5-2        DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.2.1.2 Monooxygenases
 2           Both short- and long-term exposures to O3 have been shown to enhance lung xenobiotic
 3      metabolism, possibly as a result of changes in the number and function of bronchiolar epithelial
 4      Clara cells and alveolar epithelial Type 2 cells. Studies of the effects of O3 on lung
 5      monooxygenases are listed in Table 5-1.  Early studies showed that exposure to O3 increased
 6      CYP 2B1 (the major CYP isoform in rat lung) content and activity in rat lung. Ozone exposures
 7      also caused hypertrophy and hyperplasia of CYP 2B 1-immunoreactive Clara cells. Comparisons
 8      of rat and rhesus monkey CYP isoforms demonstrated species-specific and region-specific (e.g.,
 9      trachea, parenchyma, differences in the activities of P450 isoforms (Lee et al., 1998)
10           Watt et al. (1998) found that 1 ppm O3 in both short and long-term exposures in rat
11      increased CYP 2E1 in a region-specific manner. Paige et al. (2000) showed that a long term
12      exposure ( 0.8 ppm 8h/day for 90 days) increased the activity of CYP 2B in distal lung, but not
13      trachea or intrapulmonary airways.  Studies have focused on P450 gene expression to examine
14      possible genetic mechanisms that may explain differential O3-sensitivity (Mango et al., 1998).
15      Mice (129 strain) deficient in Clara cell secretory protein (CCSP-/-), which are oxidant-sensitive,
16      were exposed to 1 ppm O3 for 2 hours.  The CCSP null mice demonstrated increases in IL-6 and
17      metallothionein (Mt) mRNA that preceded decreases in Clara cell CYP2F2 mRNA (normally
18      expressed at high levels in mouse lung) levels. In  129 strain wildtype (WT) mice, RNA levels
19      changed similarly,  to a lesser degree. These data suggest a protective role against oxidant
20      damage for CSPP,  and further, that genetic susceptibility to oxidant stress may be mediated, in
21      part, by the gene coding for CSPP.
22
23      5.2.1.3 Antioxidants, Antioxidant Metabolism, and Mitochondrial Oxygen Consumption
24           Ozone also undergoes reactions with AA, GSH, and uric acid, all antioxidants present in
25      ELF. This is a protective interaction, but even with environmentally relevant exposures to O3,
26      the reactivity of O3 is not quantitatively quenched.  Antioxidants offer some protection from O3
27      exposure, but often do not maintain sufficiently high concentration to fully protect the lung.
28      Thus, O3- induced cell injury occurs in both the lower and upper respiratory tract. Early work
29      has shown that short-term exposures to < 1 ppm O3 increase antioxidant metabolism, including
30      levels of cytosolic enzymes G6PD, 6PGD, GR, and GSHPx.  Re-exposure after a recovery
        December 2005                             5-3         DRAFT-DO NOT QUOTE OR CITE

-------
 1      period causes increases equivalent to first-time exposures, thus previous exposure appears to not
 2      be protective.
 3           Increases in enzyme activity appear to increase as a function of age, suggesting that O3
 4      exposure can cause greater lung injury in the older animal.  Species differences exist in
 5      antioxidant metabolism, with guinea pigs being very sensitive due to their diminished increases
 6      in antioxidants and antioxidant enzymes. Long term exposures of rats to urban patterns of O3
 7      (daily peaks of 0.25 ppm) caused increases in GSHPx and GR, but not superoxide dismutase
 8      (SOD).  The  enzyme changes could be accounted for by changes in the steady-state cell
 9      population or in cellular antioxidant capacity.
10           Ozone induced both site- and cell-specific changes in  copper-zinc (Cu-Zn) and manganese
11      (Mn) SOD in rats exposed to 1.0 ppm O3 for up to 3 months (Weller et al., 1997).  Cu-Zn SOD
12      labeling was  decreased in epithelial cells in airways and parenchyma. Mn SOD labeling was
13      increased in both AM and epithelial type II cells of the centriacinar region (CAR), which the
14      authors suggest may allow these cells to tolerate further O3 exposure.
15           In a recent report Freed et al.  (1999) evaluated the role of antioxidants in O3- induced
16      oxidant stress in dogs (exposed to 0.2 ppm in a 6-h exposure) by inhibiting the antioxidant
17      transport using probenencid  (an ani on-transport inhibitor). Blocking antioxidant transport
18      caused heterogeneously distributed increases in peripheral airway  resistance and reactivity,
19      supporting the hypothesis that in the lung periphery, endogenous antioxidants moderate the
20      effects of O3  and that this exposure is a subthreshold stimulus for producing effects on peripheral
21      airway resistance and reactivity in dogs. The authors further found that treatment with
22      probenecid also inhibited O3-induced neutrophilic inflammation, providing evidence for a
23      dissociation between airway function and inflammation, and suggesting that O3-induced
24      inflammation and airway hyperreactivity (AHR) are independent phenomena.
25           Mudway and Kelly (1998) modeled the interactions of O3 with ELF antioxidants using a
26      continually mixed, interfacial exposure set up with O3 concentrations of 0 to 1.5 ppm.  Uric  acid
27      was ranked the most O3-reactive, AA the second most reactive, and GSH the least reactive. Thus,
28      they concluded that GSH is not an important substrate for O3, while uric acid appeared to be the
29      most important substrate which  confers protection from O3 by removing it from inhaled air and
30      limiting the amount that reaches the distal lung.  The authors acknowledge limitations in
31      extrapolating these data to in vivo O3 exposures.

        December 2005                            5-4        DRAFT-DO NOT  QUOTE OR CITE

-------
 1      5.2.1.4 Lipid Metabolism and Content of the Lung
 2           One of the major postulated molecular mechanisms of action of O3 is peroxidation of
 3      unsaturated fatty acids in the lung. In both acute and short-term studies, a variety of lung lipid
 4      changes occur, including an increase in arachidonic acid. Metabolism of arachidonic acid
 5      produces a variety of biologically active mediators that can, in turn, affect host defenses, lung
 6      function, the immune system, and other functions. The protein A component of surfactant is also
 7      a primary target of O3 interaction. During the first few days of O3 exposure, the changes in lung
 8      lipid biosynthesis can be accounted for by the alveolar epithelial proliferative repair. With
 9      longer exposures (eg 0.12 ppm for 90 days ), an increases in PUFAs and a decrease in
10      cholesterol-esters are seen, indicative of long-term alterations of surfactant lipid composition.
11           Several new studies examined the effects of O3 exposure on phospholipids in lung tissue.
12      Ozonation of PUFAs is has been shown to generate other aldehydes such as nonanal and hexanal
13      in rat (Pryor et al., 1996; Frampton et al., 1999). These aldehydes are short-lived and found to
14      not affect lung function (Frampton et al., 1999). These observations suggest that levels of these
15      aldehydes are dependent on a dynamic relationship between their production and their
16      di sappearance from the ELF.
17           Pryor et al. (1995) proposed a cascade mechanism whereby ozonation products cause
18      activation of specific lipases, which then trigger the activation of second messenger pathways
19      (e.g., phospholipase A2 or phospholipase C). This group (Kafoury et al., 1999) showed that
20      exposure of cultured human bronchial epithelial cells to the lipid ozonation product 1 -palmitoyl-
21      2-(9-oxononanoyl)-sn-glycero-3-phosphocholine elicited release of platelet-activating factor
22      (PAF) and prostaglandin E2, but not IL-6.  The lipid ozonation product 1-hydroxy-l-
23      hydroperoxynonane caused release of PAF and IL-6 in these cells, but not prostaglandin E2.
24      These results suggest to the authors that O3-induced production of lipid ozonation products
25      causes release of proinflammatory mediators that then generate an early inflammatory response.
26      Long et al. (2001) exposed hamsters to 0.12, 1.0 or 3.0 ppm O3 to evaluate lipid peroxidation and
27      antioxidant depletion.  Six hour exposures to the two higher levels resulted in increased BALF
28      neutrophil numbers and F2-isoprostanes. Exposures to 1.0 ppm O3 with 1 h of exercise caused
29      increased levels of F2-isoprostanes.
30           Postlethwait et al. (1998) utilized three biologically relevant models, isolated epithelial
31      lining fluid, intact lung, and liposome suspensions to determine the O3-induced production of

        December 2005                            5-5         DRAFT-DO NOT QUOTE OR CITE

-------
 1      heptanal, nonanal and hexanal.  Data obtained from these studies suggested that PUFAs directly
 2      react with O3 and the amount of bioactive lipids produced is inversely related to ascorbic acid
 3      (AA) availability.  The authors caution that there are limitations to the use of measurements of
 4      these reactions products in determining O3 dose-response relationships due to the heterogenous
 5      nature of O3 reactions in the ELF. Connor et al. (2004) have recently reported that interfacial
 6      phospholipids may modulate the distribution of inhaled O3 and the extent of site-specific cell
 7      injury.  They utilized interfacial films composed of dipalmitoylglycero-3-phosphocholine
 8      (DPPC) with rat lung lavage fluid and human fibroblast cell culture systems.
 9           Hamilton et al. (1998) reported increased protein adducts in human AM exposed to
10      0.4 ppm O3 for 1 h with exercise.  These adducts were found to be created due to the formation
11      of one of the most toxic ozonation products, 4-hydroxynonenal (FINE).  Using human AM in
12      vitro cultures treated with FINE they also demonstrated a potential role of FINE in acute cellular
13      toxic effect of O3.
14           Uhlson et al. (2002) reacted O3 with calf lung surfactant which resulted in the production
15      of l-palmitoyl-2-(9'-oxo-nonanoyl)-glycerophosphocholine (16:Oa/9-al-GPCho).  The biological
16      activity of this oxidized phospholipid included: (1) decreased macrophage viability,
17      (2) induction of apoptosis in pulmonary epithelial-like A549 cells, (3) and release of IL-8 from
18      A549 cells. Exposures levels of 0.125 ppm O3 in this in vivo system were capable of generating
19      biologically active phospholipids that were capable of mediating toxic effects of O3.
20           Thus, new work has attempted to elucidate the mechanisms by which reactions of O3 with
21      lipids create phospholipids that then mediate downstream toxic effects.  However, it is uncertain
22      whether these described changes in lipid content and/or metabolism lead to significant changes
23      in surface tension or compliance properties of the lung, and thus  are biologically relevant and
24      affect human health.
25
26      5.2.1.5 Protein Synthesis
27           Collagen, a structural protein involved in fibrosis, increases with O3 exposure. Some
28      studies have shown that this increase persists after exposure stops. The increased collagen has
29      been correlated with structural changes in the lung. Rats exposed to an urban pattern of O3 with
30      daily peaks of 0.25 ppm for  38 weeks displayed extracellular matrix thickening. Increased levels
31      of collagen in CAR were demonstrated in female rats exposed to 0.5 to 1.0 ppm O3 for 6 h/day

        December 2005                            5-6         DRAFT-DO NOT QUOTE OR CITE

-------
 1      for 20 months and in monkeys exposed to 0.61 ppm for 1 year. Both increased age and health
 2      status (e.g., emphysemic) were implicated in the increased collagen formation in response to O3
 3      exposure. A recent time-course study (van Bree et al., 2001) evaluating the lung injury and
 4      changes in collagen content in rats exposed acutely or subchronically to 0.4 ppm O3
 5      demonstrated CAR thickening of septa which progressed from 7 through 56 days of exposure.
 6      Though collagen content decreased with PE recovery, the structural fibrotic changes in ductular
 7      septa and respiratory bronchioles persisted,  suggesting that subchronic O3 exposures in rats
 8      creates  a progression of structural lung injury that can evolve to a more chronic form, which
 9      included fibrosis.  As with changes in lung lipids, the biological relevance and adverse health
10      effects of altered protein synthesis and collagen accumulation are uncertain.
11
12      5.2.1.6  Gene Expression
13           Gohil et al. (2003) examined differential gene expression in C57BL/6 mice exposed to
14      1 ppm O3 for three consecutive nights for 8  hours. Ozone exposure induced changes in
15      expression of 260 genes ( 80% repressed and 20% induced). These included genes involved in
16      progression of the cell cycle, several NF-KB-activated genes, and genes invoved in xenobiotic
17      and major histocompatibility complex, suggesting that O3 exposure suppresses immune function
18      and xenobiotic metabolism and enhances cellular proliferation.
19
20      5.2.1.7  Summary and Conclusions - Biochemical Effects
21           Ozone has been shown to interact with a wide range of different cellular components
22      including PUFAs, amino acid residues, and some low-molecular-weight compounds (GSH,
23      urate, vitamins C  and E). As O3 does  not penetrate much beyond the ELF, damage likely results
24      from its PUFA ozonation products (mostly hydroxyhydroperoxides) involvement in signaling of
25      cellular responses such as inflammation. New work has shown that ozonation of PUFA also
26      forms the aldehydes nonanal, heptanal, and  hexanal, the production of which is dependent on AA
27      availability.  Saturated phospholipids  are thought to reduce the local dose and limit site-specific
28      cell injury from O3 exposure. Another ozonation product FINE creates protein adducts that have
29      been linked to apoptosis and heat shock proteins in vitro.
30           Both short- and long-term exposures to O3 have been  shown to enhance lung xenobiotic
31      metabolism, possibly as a result of changes  in the number and function of bronchiolar epithelial

        December 2005                            5-7         DRAFT-DO NOT QUOTE OR CITE

-------
 1      Clara cells and alveolar epithelial Type 2 cells. This modulation is both species- and region-
 2      specific and includes the isoforms CYP 2B1, CYP 2E1.CCSP is also involved in inflammatory
 3      responses to O3 exposure.  Mice strains with differing sensitivities to O3 show that responses in
 4      protein, LDH and inflammatory cell influx are due to CCSP levels and changes in lung
 5      epithelium permability.
 6          Reactions of O3 with AA, GSH, and uric acid (all antioxidants present in ELF) are a
 7      protective mechanism.  But even with environmentally relevant exposures, the reactivity of O3 is
 8      not quantitatively quenched and cell injury occurs in both the lower and upper respiratory tract.
 9      Early work has shown that short-term exposures to < 1 ppm O3 increase antioxidant metabolism.
10      Re-exposure after a recovery period causes increases equivalent to first-time exposures,
11      suggesting that previous exposure is not protective.  Increases in enzyme activity appear to
12      increase as a function of age, suggesting that O3 exposure can cause greater lung injury in the
13      older animal. Long-term urban patterns of exposure to O3 (daily peaks of 0.25 ppm) caused
14      increases in GSHPx and GR, but not SOD. Recent work has suggested that endogenous
15      antioxidants moderate the effects of O3  and that this exposure is a subthreshold  stimulus for
16      producing effects on peripheral airway resistance and reactivity, thus indicating a dissociation
17      between airway function and inflammation.
18          In both acute and short-term studies, a variety of lung lipid changes occur with O3
19      exposure, including an increase in AA.  With longer exposures (e.g., 0.12 ppm for 90 days),
20      an increase in PUFAs and a decrease in cholesterol-esters are seen, indicative of long-term
21      alterations of surfactant lipid composition.  Whether these changes in lipid content and/or
22      metabolism lead to significant changes in surface tension or compliance properties of the lung
23      remains unknown. New studies evaluating O3-induced alterations in lipid metabolism have not
24      been completed.
25          Collagen, a structural protein involved in fibrosis, increases with O3 exposure, and some
26      studies have shown that this increase persists after exposure stops. Urban patterns of exposure
27      (daily peaks of 0.25 ppm for 38 weeks) created extracellular matrix thickening.  Increases in
28      centriacinar collagen were demonstrated in female rats exposed to 0.5 to 1.0 ppm O3 for 6 h/day
29      for 20 months and in monkeys exposed to 0.61  ppm for 1 year. New work examining the time
30      course of lung injury and changes in collagen content in rats exposed acutely  or subchronically
31      to 0.4 ppm O3 showed centriacinar thickening of septa. Collagen content decreased with PE

        December 2005                            5-8        DRAFT-DO NOT QUOTE OR CITE

-------
 1      recovery but not the structural fibrotic changes in ductular septa and respiratory bronchioles,
 2      which suggests that subchronic O3 exposures in rats creates a progression of structural lung
 3      injury that can evolve to a more chronic form, which includes fibrosis.
 4
 5      5.2.2 Lung Host Defenses
 6           Defense mechanisms, including the mucociliary clearance system, AMs, and humoral- and
 7      cell-mediated immune system, exist in the lung to protect it from infectious and neoplastic
 8      disease and inhaled particles. Summaries of key new animal studies examining the effects of O3
 9      on lung host defenses are presented in Table AX5-2 of Annex AX5. Acute human exposures to
10      O3 result in  similar effects on AMs (see Chapter 6).
11
12      5.2.2.1 Clearance
13           Early  studies of the effect of O3 on the mucociliary escalator showed morphological
14      damage to ciliated epithelial cells of the tracheobronchial tree at doses of < 1 ppm.  Functionally,
15      O3 slowed particle clearance in rats at doses of 0.8 ppm for 4 h and in rabbits at 0.6 ppm for 2 h
16      exposures.  Acute exposures at 0.5 ppm O3 in sheep caused increased basal secretion of
17      glycoproteins, while longer exposures reduced tracheal glycoprotein secretions, both of which
18      can alter the effectiveness of the mucociliary escalator. Early postnatal exposures of sheep to
19      1 ppm O3 caused retardation of normal morphologic development of the tracheal epithelium,
20      decreased epithelial mucosia density, decreased tracheal mucous velocity,  and delayed
21      development of carbohydrate composition.  Conversely , alveolar clearance in rabbits after acute
22      exposure  (0.1 ppm, 2 h/day,  for 1 to 4 days) is increased.  Longer exposures showed no effect
23      and increased O3 (1.2  ppm) slowed clearance.  This pattern of clearance occurs in rats also.
24      A study using rat tracheal explants exposed to O3 (Churg et al., 1996) showed that uptake of
25      TiO2 and  asbestos was enhanced at 0.01 and 0.1 ppm, respectively.  The authors attribute the
26      increased uptake as a direct effect of O3, suggesting mediation by H2O2 or hydroxyl radical.
27      Studies of the clearance of the radiolabled chelate 99mTc diethylenetriamine pentaacetic acid
28      (Tc-DTPA) have shown that clearance is significantly increased following a 3h exposure to
29      0.8 ppm O3  in SD rats (Pearson and Bhalla,1997). Examination of regional clearance of
30      99mTc-DTPA in dogs following a 6 h isolated sublobar exposure to 0.4 ppm O3 or air showed that
31      O3 decreased the clearance halftime by 50% at 1 day following exposure (Foster and Freed,

        December 2005                            5-9         DRAFT-DO NOT QUOTE OR CITE

-------
 1      1999).  Clearance was still elevated at 7 d PE but had recovered by 14 d. So, a single local
 2      exposure to O3 increases transepithelial clearance but without any influence on contralateral
 3      segements, i.e., only for epithelia directly exposed to O3.
 4           Alveolar clearance is slower than tracheobronchial clearance and involves particle
 5      movement through interstitial pathways to the lymphatic system or movement of particle-laden
 6      AMs to the bottom of the mucociliary escalator. Exposures of rabbits to 0.1 ppm accelerated
 7      clearance while 1.2 ppm slowed clearance. A chronic exposure has been shown to slow
 8      clearance. New evaluations of the effects of O3 on alveolar clearance have not been performed.
 9
10      5.2.2.2 Alveolar Macrophages
11           A primary function of AMs is to clear the lung of infectious and non-infectious particles by
12      phagocytosis, detoxification, and removal. Further, AMs secrete cellular mediators that recruit
13      and activate inflammatory cells in the lungs.  Ozone has been shown to inhibit phagocytosis at
14      0.1 ppm for 2 h in rabbits. This inhibition returns to control levels if exposures are repeated for
15      several days.  The production of superoxide anion radicals and the activity AM lysosomal
16      enzymes (both involved in bactericidal activity) are inhibited by 3 h exposures to 0.4 and
17      0.25 ppm O3 in rodents and rabbits, respectively. Production of IFNy was decreased in rabbit
18      AM by 1 ppm O3 for 3 h.
19           New studies have shown that O3 affects AM chemotaxis, cell adhesion, and surface
20      expression of cell adhesion molecules (Bhalla,  1996). AM from SD rats exposed to 0.8 ppm O3
21      for 3 h showed greater mobility and greater adhesion than air exposed controls. This increased
22      mobility and adhesion were attenuated by CD16b and ICAM-1 antibodies, suggesting these
23      adhesion molecules modulate O3-induced inflammation. Antibodies to TNFa and ILla also
24      mitigated AM adherence, suggesting  further that the inflammatory response to O3 is mediated by
25      these cytokines (Pearson and Bhalla,  1997). Cohen et al. (1996) showed that O3 reduces binding
26      of INF y to AM in WEHI-3 cells, and additionally reduces phagocytic activity, production of
27      reactive oxygen intermediates, and elevation of intracellular Ca++. Glutathione content in AM is
28      reduced by a 2 ppm 3 h exposure to O3, possibly due to its interaction with ozonation products
29      from O3-induced lipid peroxidation (Pendino et al.,  1996).
30           Cohen et al. (2001, 2002) exposed male F-344 rats to either 0.1  or 0.3 ppm O3 for 4 h/day,
31      5 days/week or either 1 or 3 weeks. In this study, superoxide anion production was increased  at

        December 2005                             5-10        DRAFT-DO NOT QUOTE OR CITE

-------
 1      1 week.  Hydrogen peroxide production was reduced at both exposure concentrations and
 2      durations and was further reduced with INFy stimulation, suggesting that one effect of O3 is
 3      compromised killing of bacteria by AM due to the reduction in hydrogen peroxide production.
 4           Ozone treatment (2 ppm O3, 3 h in female SD rats) caused a time-dependent increase in NO
 5      levels in both AM and type II epithelial cells that was correlated with increased expression of
 6      iNOS mRNA and protein (Laskin et al., 1998).  Inhibition of NF-KB, caused a dose-dependent
 7      inhibition of NO and iNOS production.  Additionally, O3 caused a time-dependent increase in
 8      NF-KB binding activity in the nucleus of both cell types. The authors hypothesize that O3
 9      exposure causes the cytokines TNFa and IL-1 Pa to bind to  surface receptors and initiate
10      intracellular signaling pathways in AM leading to activation of NF-KB, its entry into the nucleus,
11      and its binding to the regulatory sequences of genes  such as iNOS to allow their transcription.
12      Additional studies (Laskin et al., 2002) using AM isolated from C57B16xl29 mice with a
13      targeted  disruption of the gene for iNOS showed no  toxicity to 0.8 ppm O3 for 3h, as measured
14      by BALF protein levels and nitrotyrosine staining of the lung.  Additionally, mice
15      overexpressing human Cu, Zn superoxide dismutase (SOD) and mice with a targeted disruption
16      of p50 NF-KB were also resistant to O3 toxicity. WT mice exposed to O3 showed an increase in
17      expression of STAT-1, a protein that binds to the regulatory region of iNOS.  Taken together,
18      these results suggest to the authors that a number of proteins including NF-KB, phosphoinoside
19      3-kinase, and STAT-1 that bind to and regulate expression of iNOS are modulated by O3
20      exposure. The same iNOS knockout mice strain exposed to O3 (Fakhrzadeh et al., 2002)  showed
21      no increase in AM superoxide anion and prostaglandin.  These data provide further evidence the
22      NO and its reactive oxidative product peroxynitrite are important in O3-induced lung injury.
23      Further discussions of the role of nitric oxide synthase/reactive nitrogen and
24      cytokines/chemokines in O3-induced inflammation are provided in Section 5.2.3.
25
26      5.2.2.3 Immune System
27           The effects  of O3 on the immune system are complex and depend on the exposure
28      parameters and observation periods. T-cell-dependent functions appear to be more affected than
29      B-cell-dependent functions.  Generally, there is an early immunosuppressive effect that can, with
30      continued exposure, either return to normal or actually enhance immunity.  Changes in immune
31      cell population occur with O3 exposure including T:B-cell ratios in the MLN. Natural killer

        December 2005                            5-11        DRAFT-DO NOT QUOTE OR CITE

-------
 1      (NK) cell activity increases with 1 week exposures of 0.2 to 0.4 ppm O3 but decreases with
 2      exposures to 0.82 ppm. Ozone exposure has also shown to be responsible for enhancement of
 3      allergic sensitization at levels of 0.5 to 0.8 ppm for 3 days.  Studies of the effects of O3 on the
 4      immune system are summarized in Table AX5-2.
 5           Recent work examining immune system responses to O3, Garssen et al. (1997) have shown
 6      that BALB/c mice sensitized with PCI are hyperreactive to carbachol after a PSA challenge, but
 7      not if exposed to 0.4 to 1.6 mg/m3 O3.  The sensitized mice also demonstrated a suppressed
 8      inflammatory reaction (PMN) with 1.6 mg O3 exposure.  These results are opposite to the effect
 9      on type I (IgE-mediated) allergic reactions, which the authors suggest is due to activation of Th-
10      2 cell-dependent reactions that are possibly potentiated by O3 or to a direct effect by O3 on Th-1
11      cells or other cells that are crucial  for the tracheal hyperreactivity and inflammation seen in this
12      mouse model.
13           Kleeberger et al. (2000, 2001b) have recently demonstrated a potential interaction between
14      the innate and acquired immune system with O3 exposure.  Using O3-susceptible (C57BL/6J) and
15      O3-resistant (C3H/HeJ) mice, they identified a candidate gene on chromosome 4, Toll-like
16      receptor 4 (Tlr4). Ozone exposure (0.3ppm for 24 to 72 hours) of C3H/HeJ and C3H/HeOuJ
17      mice, the latter differing from the O3-resistant strain by a polymorphism in the coding region of
18      77r¥,  then demonstrated greater protein concentrations in the OuJ strain. The two strains
19      exhibited differential expression of Tlr4 mRNA with O3 exposure. Thus, a quantitative trait
20      locus on chromosome 4 appears to be responsible for a significant portion of the genetic variance
21      in O3 -induced lung hyperpermeability. In these mouse strains lavageable protein concentration
22      was lowered by inhibition of inducible nitric oxide synthase (iNOS) and by  targeted disruption
23      of Nos2. Comparisons of C3H/HeJ and C3H/HeOuJ O3 exposures demonstrated reduced Nos2
24      and Tlr4 mRNA levels in the O3-resistant C3H/HeJ mice. These data are consistent with the
25      hypothesis that O3 -induced lung hyperpermeability is mediated by iNOS. These studies suggest
26      a role for TLR4 in the host response to O3 similar to the role it has demonstrated in
27      lipopolysaccharide (LPS) sensitivity (Schwartz 2002; Wells et al. 2003). TLR4 signaling is
28      thought to be critical to linking the innate and acquired immune system through antigen
29      presenting cells and Thl/Th2 differentiation.
30           Ozone exposure has been shown to affect IgE responses in both in vitro and in mice.
31      Becker et al. (1991) demonstrated changes in IgG production in cultured human lymphocytes

        December 2005                            5-12        DRAFT-DO NOT QUOTE OR CITE

-------
 1     with O3 exposures of 1.0, 0.5, and 0.1 ppm for 2 h.  Subsequent to O3 exposure, cells were
 2     stimulated with pokeweed mitogen (PWM, a T-cell-dependent stimulus) or Staphylococcus
 3     aureus Cowan 1 strain (SAC, a T-cell-independent stimulus). Both B and T cells were affected
 4     by O3. T cells also demonstrated an increase in IL-6 and a decrease in IL-2, which suggested to
 5     the authors that O3 may have direct effects on IgG producing cells and concurrently an effect that
 6     is mediated by altered production of T cell immunoregulatory molecules. Responses to repeated
 7     O3 (0.08 - 0.25 ppm) and OVA (1%) exposures were compared in "IgE-high responder"
 8     (BALB/c) and "IgE-low responder" (C57BL/6) mice (Neuhaus-Steinmetz et al., 2000). Ozone
 9     appeared to shift the immune response toward a Th2-like pattern in the two mouse strains with
10     differing potentials for developing allergic reactions.
11           Surfactant protein A and D (SP-A and SP-D) were shown to create an  inflammatory
12     feedback loop with perturbations in lung immune defenses (reviewed in Hawgood and Poulain,
13     2001). Earlier studies suggested that SP-A is a target for O3 toxicity by causing inhibition of SP-
14     A self-association and SP-A-mediated lipid vesicle  aggregation.  Further, O3 reduced the ability
15     of SP-A to inhibit phospholipid secretion by  alveolar type II cells O3 and reduced the capacity of
16     SP-A to induce superoxide anion production and enhance phagocytosis of herpes simplex virus.
17     Bridges et al. (2000) reported that both SP-A and SP-D directly protect surfactant phospholipids
18     and macrophages from oxidative damage by  blocking accumulation of TEARS and conjugated
19     dienes.
20           Eight human variants of PS-A in CHO  cells exposed to O3 (Ippm for 4 hr) showed
21     decreased ability to  stimulate cytokine (TNF- and IL-8) production in THP-1 cells, a
22     macrophage-like cell line (Wang et al., 2002). Each variant had a unique time- and dose-
23     dependent pattern of stimulation of cytokine  production with O3 exposure which the authors
24     attribute to possible differences in susceptibility to O3 oxidation.  Targeted disruption of mouse
25     SP- A and SP-D (Hawgood et al, 2002) caused increases  in BAL phospholipid, macrophage, and
26     protein through 24 weeks of age. Further, the deficient mice developed patchy lung
27     inflammation and air space enlargement consistent with emphysema.  Future experiments using
28     these null mice will  help to establish the role of SP-A and SP-D in pulmonary host defense to O3
29     exposure.
30
       December 2005                            5-13        DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.2.2.4 Interactions with Infectious Microorganisms
 2           Ozone-induced dysfunction of host defense systems results in enhanced susceptibility to
 3      bacterial lung infections. Acute exposures of 0.08 ppm (3h) O3 can overcome the ability of mice
 4      to resist infection (by decreasing lung bactericidal activity) with Streptococcal bacteria, resulting
 5      in mortality. Changes in antibacterial defenses are dependent on exposure regimens, species and
 6      strain of test animal, species of bacteria, and age of animal, with young mice more susceptible to
 7      the effects of O3. The effect of O3 exposure on antibacterial host defenses appears to be
 8      concentration- and time-dependent. Early studies using the mouse "infectivity model," consisting
 9      of exposure to clean air or O3 followed by exposure to an aerosolized microorganism, showed
10      that the difference in mortality between O3-exposed groups and controls is concentration-related.
11      Chronic exposures (weeks, months) of 0.1 ppm do not cause greater effects on infectivity than
12      short exposures, due to defense parameters becoming reestablished with prolonged exposures.
13           More recent studies of O3-induced modulation of cell-mediated immune responses showed
14      effects on the  onset and persistence of infection.  Cohen et al. (2001,2002) exposed male F-344
15      rats subchronically to either 0.1 or 0.3 ppm O3. Subsequent exposure with viable Listeria
16      monocytogenes demonstrated no observed effect  on cumulative mortality, but did show a
17      concentration-related effect on morbidity onset and persistence. These data suggest that O3 may
18      cause a possible imbalance between Th-1 and Th-2 cells, which can subsequently lead to
19      suppression of the resistance to intracellular pathogens.
20           Effects of O3 on viral infections are dependent on the temporal relationship between O3
21      exposure and viral infection.  Only high concentrations (1.0 ppm O3, 3 h/day, 5 days, mice)
22      increased viral-induced mortality. No detrimental effects were seen with a 120-day exposure to
23      0.5 ppm O3 on acute lung injury from influenza virus  administered immediately before O3
24      exposure started. But there were O3-enhanced postifluenzal alveolitis and lung parenchymal
25      changes.  As O3 does not affect lung influenza viral liters, it apparently does not impact antiviral
26      clearance mechanisms.  In general, the evidence suggests that O3 can enhance both bacterial and
27      viral lung infections, but the key mechanisms have not yet been identified. New studies on the
28      interactions of O3 and viral infections have not been published.
29
        December 2005                            5-14        DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.2.2.5 Summary and Conclusions - Lung Host Defenses
 2           New data on lung host defenses support earlier work which suggests that mucociliary
 3      clearance is affected in most test species at just under Ippm, with lower levels (~0.1 ppm)
 4      increasing clearance and somewhat higher levels decreasing clearance. These data also propose
 5      mechanisms whereby O3 affects clearance, which include uptake being a direct effect of O3, but
 6      modulated by ROS and hydroxyl radicals.
 7           Alveolar macrophage function is disrupted by O3 as shown by a number of studies
 8      demonstrating inhibition of phagocytosis at concentrations ranging from 0.1 to 1.2 ppm.  This
 9      inhibition returns to control levels if exposures are repeated for several days. Two new studies
10      corroborate earlier findings of increases in AM number in that same exposure range. In this
11      environmentally relevant exposure range, new studies support older findings of decreased
12      resistance to microbial pathogens as shown by the endpoints examining superoxide radical
13      formation, altered chemotaxis/motility, decreased INFy, levels, decreased lysosomal activity,
14      increased PGE levels, and increased NO mRNA and protein.
15           New research evaluating the effects of O3 on immune function advances previous work that
16      has shown that exposures can enhance or suppress immune responsiveness depending on the
17      species studied, concentration of O3, route of exposure of allergen, and timing of exposure.
18      Continuous exposure to O3 impairs immune responses for the first several days of exposure,
19      followed by an adaptation to O3 that allows a return of normal immune responses.  Most species
20      show little effect of O3 exposures prior to immunization, but a suppression of responses to
21      antigen in O3 exposures post-immunization.  The use of mouse strains with genetically
22      determined sensitivity or resistance to O3 indicated a possible interaction between the innate and
23      acquired immune system, and further, that O3 may shift the immune response towards a Th-2-
24      like pattern. Work has also focused the  deleterious effects of O3 exposure on SP-A and SP-D
25      and their immunomodulatory function in protecting against oxidative stress.
26           Several new studies evaluating the effects of O3 exposures on infectious microorganisms
27      are in concurrence with previous studies which showed, in general, increased mortality and
28      morbidity, decreased clearance, increased bacterial growth, and increased severity of infection at
29      exposure levels of 0.1 to 1 ppm O3 for 1 week.
30
        December 2005                            5-15        DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.2.3  Inflammation and Lung Permeability Changes
 2          Lung inflammation and increased permeability, which are distinct events controlled by
 3      independent mechanisms, are two well-characterized effects of O3 exposure. Disruption of the
 4      lung barrier leads to leakage of serum proteins, influx of polymorphonuclear leukocytes (PMNs),
 5      release of bioactive mediators, and movement of compounds from the airspaces into the blood.
 6      Increases in permeability and inflammation have been observed at levels as low as 0.1 ppm O3
 7      for 2 h/day for 6 days in rabbit and 0.12 ppm in mice (24-h exposure) and rats (6-h exposure).
 8      After acute exposures, the influence of the time of exposure increases as the concentration of O3
 9      increases. The exact role of inflammation in causation of lung disease is not known, nor is the
10      relationship between inflammation and changes in lung function.  Table AX5-3 in Annex AX5
11      summarizes new key studies describing the potential for O3 exposure affect lung permeability
12      and inflammation.  Controlled human exposure studies discussed in Chapter 6 indicate that the
13      majority of acute responses in humans are similar to those observed in animals.
14
15      5.2.3.1 Time Course of Inflammation and Lung Permeability Changes
16          The maximal increase in BALF protein, albumin, and PMN occurs in 8 to 18 h after the
17      cessation of acute exposure. A recent study of OVA-sensitized male Dunkin-Hartley guinea pigs
18      exposed to 1.0 ppm O3 for 3 h showed that PMN levels were significantly increased at 3 h PE,
19      but BAL protein levels were not, suggesting a lack of correlation  between the two endpoints
20      (Sun et al., 1997).  Increased PMN without a concordant increase in BAL protein levels were
21      found when the guinea pigs were exposed to 1.0 ppm O3 for 1  h and evaluated 24-h PE. The first
22      group also had an increase in AHR, but not the second group,  which suggests a dissociation
23      between PMN levels and AHR.
24          Earlier work demonstrated that O3 exposures of 0.8 to 1  ppm transiently increase the
25      permeability from the air to the blood compartment. This permeability is greatest in trachea and
26      bronchoalveolar zone,  and may allow increased entry of antigens and other bioactive compounds
27      (e.g., bronchoconstrictors)  into lung tissues.  The time course of the influx of PMNs into the lung
28      and the BALF fluid levels of macrophage inflammatory protein-2 (MIP-2) were found to be
29      roughly similar to that for proteins (Bhalla and Gupta, 2000).  Adherence of neutrophils to
30      pulmonary vascular endothelium is maximal within 2 h after exposure and returns to control
31      levels by 12 h PE (Lavnikova et al., 1998). In an in vitro system utilizing rat alveolar type II cell

        December 2005                            5-16        DRAFT-DO NOT QUOTE OR CITE

-------
 1      monolayers, O3 produced a dose-dependent increase in permeability (Cheek et al., 1995).
 2      At higher O3 levels, neutrophils exacerbated the injury, but their presence after the exposure
 3      expedited restoration of epithelial barrier. Vesely et al. (1999) have demonstrated that
 4      neutrophils contribute to the repair process in O3 -injured airway epithelium.
 5           Subchronic exposures (3 to 7 days) have been found to cause increases in BALF protein
 6      and PMNs that typically peak after a few days (depending upon species tested and exposures)
 7      and return towards control even with continuing exposure.  Van Bree et al. (2002) observed
 8      lower BALF levels of protein, fibronectin, IL-6 and inflammatory cells in rats exposed for 5 days
 9      than in rats exposed for 1 day, suggesting adaptation to O3  exposure. Postexposure challenge
10      with single O3 exposures at different time points showed recovery of susceptibility to O3.
11      McKinney et al. (1998) observed differences in IL-6 levels due to repetitive exposures and
12      demonstarted a role of IL-6 in the adaptive response induced by repeated O3 exposures.
13
14      5.2.3.2 Concentration and Time of Exposure
15           Analysis of the relative influence of concentration and duration of exposure (i.e., CXT) of
16      O3 has shown that concentration generally dominates the response.  The impact of T was C-
17      dependent (at higher Cs, the impact of T was greater); at the lowest C and T values, this
18      dependence appeared to be lost. New studies evaluating CXT relationships in animal models
19      have not been found.
20
21      5.2.3.3  Susceptibility Factors
22           Factors that have been studied for potential impact on the effects of O3 exposure include
23      age, gender, nutritional status, exposure to co-pollutants, exercise, and genetic variability. A full
24      characterization of the  effects of age on O3 responses has not been completed. Data available
25      indicate that effects of age on O3 responses are endpoint-dependent, with young mice, rats and
26      rabbits having greater prostaglandin levels with exposure and senescent rats having greater IL-6
27      and N-acetyly-p-D-glucosaminidase levels with exposure.
28           A new study (Johnston et al., 2000b) compared gene  expression of chemokines and
29      cytokine in newborn and 8-week-old C57B1/6J mice exposed to 1.0 or 2.5 ppm for 4, 20, or 24 h.
30      The newborn mice displayed increased levels of Mt mRNA only, while the 8-week-old mice had
31      increases in MIP-1 a, MIP-2, IL-6, and Mt mRNA. Comparisons were made with mice of the

        December 2005                             5-17        DRAFT-DO NOT QUOTE OR CITE

-------
 1      same age groups with exposures to endotoxin (10 min). Both age groups displayed similar
 2      cytokine/chemokine profiles with endotoxin exposure. This suggested to the authors that the
 3      responses to endotoxin, which does not cause epithelial injury, and the responses to O3, which
 4      does, demonstrate that differences in inflammatory control between newborn and adult mice is
 5      secondary to epithelial injury.
 6           Pregnancy and lactation increased the susceptibility of rats to acute O3, but no clear effects
 7      of gender have been identified. The effects of vitamin C deficiency on O3 responses are unclear.
 8      Ascorbate-deficient guinea pigs exposed to O3 demonstrated only minimal effects on injury and
 9      inflammation (Kodavanti et al., 1995). Utilizing a diet-restricted (20% of the freely-fed diet) rat
10      model, Elsayed (2001) demonstrated higher survivability on exposure to higher O3 (0.8 ppm
11      continuously for 3 d) compared to freely-fed rats. Pre-exposure to sidestream cigarette smoke
12      had been found to cause increased lung injury (Yu et al., 2002).  In vitro studies on the
13      macrophages from smoke + O3- exposed animals responded by a greater release of TNF-a
14      following LPS stimulation when compared to macrophages exposed to air, smoke or O3 alone.
15           Recent lines of evidence illustrate that genetic background is an extremely important
16      determinant of susceptibility to O3. In earlier studies using inflammation-prone (susceptible)
17      C57BL/6J (B6) and inflammation-resistant C3H/HeJ (C3) mouse strains and high doses of O3
18      (2ppm for 3 hours) identified Inf-2 as a locus controlling susceptibility. Further studies in these
19      two strains of mice identified that the acute and subacute exposures are controlled by two
20      distinct genes, referred to as Inf-l and Inf-2, respectively (Tankersley and Kleeberger, 1994).
21      Kleeberger et al. (1997) also identified another potential susceptibility gene, tumor necrosis
22      factor (Tnf, which codes for the pro-inflammatory cytokine TNF-a ) on a qualitative trait locus
23      on mouse chromosome 17. By neutralizing the  function of TNF-a with a specific antibody, they
24      were able to confer protection against O3 injury in susceptible mice. The group then
25      demonstrated a role for TNF receptor 1 and 2 (TNFR1 and TNFR2, respectively) signaling in
26      subacute ( 0.3  ppm for 48 hrs) O3-induced pulmonary epithelial injury and inflammation (Cho
27      et al., (2001).  TNFR1 and TNFR2 knockouts were less sensitive to subacute O3 exposure than
28      WT C57BL/6J mice. Further studies  using these knockouts by Shore et al. (2001) indicated  a
29      role of TNF-a in AUR but not in O3-induced infiltration of PMN, and provided evidence for the
30      mechanistic separation of hyperresponsiveness and PMN infiltration.
        December 2005                           5-18        DRAFT-DO NOT QUOTE OR CITE

-------
 1           In studies to evaluate differences in susceptibility to death from O3 exposure, Prows et al.
 2      (1997; 1999) exposed A/J and C57BL/6J mice exposed to 10 ppm O3. The A/J strain is more
 3      sensitive to O3-induced death, while the C57BL/6J strains is resistant to O3-induced death. They
 4      identified two loci (acute lung injury-1 and -3, Ali-1 and Ali-3, respectively) on chromosome
 5      11 that appear to control susceptibility to death after O3 exposure.
 6           An integrated and more comprehensive effort to identify the genetic basis for the
 7      susceptibility to O3-induced lung injury was reported by Savov et al. (2004). In this report,
 8      summarized in Table AX5-4 of Annex AX5, acute lung injury to high dose of O3 (2 ppm for 3h)
 9      was assessed and integrated with physiological, biochemical, and genetic observations using 9
10      inbred mouse strains. This work indicated the presence of genetic loci on chromosomes 1,7, 15
11      associated with phenotypic characteristics for resistance to acute O3-induced lung injury.  They
12      identified C3H/HeJ and A/J as consistently O3-resistant,C57BL/6J and 129/SvIm as consistently
13      O3-vulnerable, and CAST/Ei, BTBR, DBA/2J, FVB/NJ, and BALB/cJ as intermediate in
14      response to O3.
15           A similar comparative study by Wattiez et al. (2003) using five inbred mouse strains to
16      characterize the molecular basis for O3-induced lung injury identified differential expression of
17      CCSP isoform CC16a in BALF of C57BL/6J (O3-sensitive) and C3H/HeJ (O3-resistant) strains.
18      Ozone-induced changes in CCSP expression were evaluated in five inbred mouse strains:
19      C57BL/6J and CBA both considered sensitive to acute O3-induced inflammation, C3H/HeJ and
20      AKR/J both considered resistant, and  SJL/J considered intermediate (Broeckaert et al., 2003).
21      Two exposure paradigms (1.8 ppm O3 for 3 h or 0.11 ppm O3, 24/h day for up to 3 days) were
22      used, and BALF and serum were assayed immediately after exposure or at 6 h.  Both exposure
23      levels caused a transient increase in CC16 in serum that correlated with BALF changes in
24      protein, LDH, and inflammatory cells. There was an inverse relationship between preexposure
25      levels of CC16 in BALF and epithelial damage based on serum CC16 levels and BALF markers
26      of inflammation. There was also an inverse relationship between preexposure levels of albumin
27      in BALF and lung epithelium damage. These results suggest to the authors that a major
28      determinant of susceptibility to O3 is basal lung epithelium permeability. I.e., the leakiness of
29      the epithelium allows CC16 to enter the blood and protein and inflammatory cells to enter the
30      lung. As all of the mouse strains had similar levels of preexposure CC16 mRNA, they conclude
31      that strain differences in the basal permeability of the airway epithelium is responsible for lung

        December 2005                           5-19        DRAFT-DO NOT QUOTE OR CITE

-------
 1      differences in basal CC16 among strains. This group also used 2-dimensional protein
 2      electrophoresis to examine differences in BALF protein between C57BL/6J and C3H/HeJ mice
 3      before and after O3 exposure.  They found the CC16 monomer, a 7kD protein, in two isoforms
 4      with differing p/values, CC16a (4.9) and CC16b (5.2).  C57BL/6J mice had lower levels of
 5      CC16a (the more acidic form) than the C3H/HeJ mice, but both strain had similar levels of
 6      CC16b.  They hypothesize that the C57BL/6J strain has greater epithelial permeability, and thus
 7      allows more leakage of CC16a. These data taken together suggest a protective role against
 8      oxidant damage for CSPP, and further that genetic susceptibility to oxidant stress may be
 9      moderated, in part, by the gene coding for CSPP.
10
11      5.2.3.4 Mediators of Inflammatory  Response and Injury
12           Ozone reacts with lipids in the ELF or epithelial cell membranes, creating ozonation
13      products that then stimulate airway epithelial cells,  AMs, and PMNs to release a host of pro-
14      inflammatory mediators which include cytokines, chemokines, reactive oxygen species,
15      eicosanoids, and platelet activating factors.  At O3 exposures of > 1 ppm, these mediators recruit
16      PMN and increase expression of MIP-2 mRNA or BALF levels of MIP-2 (Driscoll et al., 1993;
17      Haddad et al., 1995; Bhalla and Gupta, 2000).  The increased mRNA expression was associated
18      with an increased neutrophilia in the lung. Zhao et al. (1998) showed that O3 exposure in mice
19      and rats causes  an increase in monocyte chemotactic protein-1 (MCP-1).
20           Fibronectin, an extracellular matrix glycoprotein, is thought to have a role in lung
21      inflammation and inflammatory disorders, and has shown to be increased with exposures of
22      1 ppm for 14 days. Gupta et al. (1998) observed an increase in both fibronectin protein and
23      mRNA expression in the lung of rats exposed to 0.8 ppm O3.  A mechanistic role of fibronectin
24      in O3-induced inflammation and injury was suggested on the basis of comparability of temporal
25      changes in BALF protein, fibronectin  and alkaline phosphatase activity (Bhalla et al., 1999).
26      Studies have reported  an effect of O3 on other cytokines and inflammatory mediators.
27      An increase occurred for cytokine-induced neutrophil chemoattractant (CINC) and NF-icB
28      expression in vivo (Haddad et al., 1996; Koto et al., 1997), for IL-8 in vivo and in vitro (Chang
29      et al., 1998),  TNFa, fibronectin, IL-1  and CINC release by macrophages ex vivo (Pendino et al.,
30      1994; Ishii et al., 1997), and NF-KB and TNFa (Nichols et al., 2001; see 6.9.2).  An increase in
31      lung CINC mRNA occurred within 2 hrs after the end of a 3 hr exposure of rats to 1 ppm O3.

        December 2005                           5-20        DRAFT-DO NOT QUOTE OR CITE

-------
 1      The CINC mRNA expression was associated with neutrophilia at 24 hrs PE. Exposure of guinea
 2      pig AMs recovered in BALF and exposed in vitro to 0.4 ppm O3 for 1- h produced a significant
 3      increase in IL-6 and TNFa (Arsalane et al., 1995).  An exposure of human AMs to an identical
 4      O3 concentration increased TNFa, IL-lb, IL-6 and IL-8 and their mRNAs.  Ozone exposure of
 5      mice caused an increase in IL-6, MlP-la, MIP-2, eotaxin and Mt abundance (Johnston et al.,
 6      1999a). The IL-6 and MT increase was enhanced in mice deficient in CCSP, suggesting a
 7      protective role of Clara cells and their secretions (Mango et al., 1998).  CCSP deficiency also
 8      increased sensitivity of mice to O3, as determined by an increase in abundance of MlP-la and
 9      MIP-2 following a 4 hr exposure (Johnston et al., 1999b).
10           Mast cells, which are located below the epithelium, release proinflammatory mediators and
11      have been shown to contribute to O3-induced epithelial damage.  Mast cell-deficient mice
12      exposed to 2 ppm O3 showed no inflammation or epithelial injury which was observed in WT
13      mice (Longphre et al., 1996). Greater increases in lavageable macrophages, epithelial cells and
14      PMNs were observed in mast cell-sufficient mice than in mast cell-deficient mice exposed to
15      0.26 ppm (Kleeberger et., 2001a).  Increases in inflammatory cells were also observed in mast
16      cell-deficient mice repleted of mast cells, but O3-induced permeability increase was not different
17      in genotypic groups exposed to 0.26 ppm. When a mast cell line was exposed to varying O3
18      concentrations, spontaneous release of serotonin and modest generation of PGD2 occurred only
19      under conditions that caused cytotoxicity (Peden and Dailey, 1995). Additionally, O3 inhibited
20      IgE- and A23187-  induced degranulation. Mast cells recovered from O3-exposed peripheral
21      airways of ascaris sensitive dogs released significantly less histamine and PGD2 following in
22      vitro challenge with ascaris antigen or calcium ionophore (Spannhake, 1996). Ozone exposure
23      also promoted eosinophil recruitment in the nose and airways in response to instillation of OVA
24      or OVA-pulsed dendritic cells and aggravated allergy like symptoms in guinea pigs (lijima et al.,
25      2001).
26           The role of PMNs and cellular mediators in lung injury and epithelial permeability has
27      been investigated using antibodies and inhibitors of known specificity to block inflammatory cell
28      functions and cytokine activity.  Treatment of rats with cyclophosphamide prior to O3 exposure
29      resulted in a decreased recovery of PMNs in the BALF and attenuated permeability induced by
30      O3 (Bassett et al., 2001). Pretreatment of animals with antiserum against rat neutrophils
31      abrogated PMN accumulation in the lung, but did not alter permeability increase produced by

        December 2005                           5-21        DRAFT-DO NOT QUOTE  OR CITE

-------
 1      O3.  DeLorme et al. (2002) showed a relationship between neutrophilic inflammation and AHR.
 2      Treatment of rats with anti-neutrophil serum protected the animals from O3-induced AHR.
 3      Studies utilizing antibodies to selected pro- or anti-inflammatory cytokines suggest a role of
 4      TNFa, IL-10, and IL-lb in O3-induced changes in permeability, inflammation and cytokine
 5      release (Ishii et al.,  1997; Reinhart et al., 1999; Bhalla et al., 2002). An attenuation of O3-
 6      induced increase  in permeability and inflammation was also observed in mice treated, either
 7      before or after exposure, with UK-74505, a platelet-activating factor (PAF) receptor antagonist
 8      (Longphre et al.,  1999). These results were interpreted to indicate that O3-induced epithelial and
 9      inflammatory changes are mediated in part by activation of PAF receptors.
10           Ozone exposure stimulates macrophage motility towards a chemotactic gradient, and
11      macrophages from rats exposed to 0.8 ppm O3 adhered to epithelial cells (ARL-14) in culture to
12      a greater extent than macrophages from air-exposed controls (Bhalla, 1996). Both macrophage
13      motility and chemotaxis were attenuated by antibodies to cell adhesion molecules CD-I Ib and
14      ICAM-1, suggesting a role for cell adhesion molecules in O3-induced cellular interactions.  This
15      may also explain  the increased tissue localization and reduced recovery of macrophages in
16      BALF (Pearson and Bhalla, 1997) following O3 exposure. Studies investigating the mechanisms
17      of PMN recruitment in the lung have explored the role of cell adhesion molecules that mediate
18      PMN-endothelial interactions. An exposure of female rats to O3 had an attenuating effect on
19      CD-18 expression on AMs and vascular PMNs, but the expression of CD62L, a member of
20      selectin family, on vascular PMNs was not affected (Hoffer et al., 1999). In monkeys, O3-
21      induced inflammation was blocked by treatment with a monoclonal antibody to CD 18,
22      suggesting dependence of PMN recruitment on this adhesion molecule (Hyde et al.,  1999).
23      Treatment of monkeys with CD 18 antibody also reduced tracheal expression of the P6 integrin
24      (Miller et al., 2001). A single 3 hr exposure of rats to O3 caused an elevation in concentration of
25      ICAM-1, but not  CD-18, in the BALF (Bhalla and Gupta, 2000). Takahashi et al. (1995a) found
26      an increase in tissue expression of ICAM-1 in mice exposed to 2 ppm O3, noting  a temporal
27      correlation of inflammatory activity and ICAM-1 expression which varied in different regions of
28      the lung. A comparable pattern of time-related changes in total protein, fibronectin and alkaline
29      phosphatase activity in the BALF of rats exposed to 0.8 ppm O3 was also noted by Bhalla et al.
30      (1999). Together, these studies support the role of extracellular matrix protein and cell  adhesion
31      molecules in the induction of lung inflammation and  injury.

        December 2005                            5-22        DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.2.3.5 The Role of Nitric Oxide Synthase and Reactive Nitrogen in Inflammation
 2           Nitric oxide (NO) is a messenger molecule involved in many biological processes,
 3      including inflammation. Cells in the respiratory tract (including mast cells, neutrophils,
 4      epithelial cells, neurons, and macrophages) produce three differing forms of nitric oxide synthase
 5      (NOS), the enzyme that catalyzes the formation of NO. NOS-l(neuronal) and NOS-3
 6      (endothelial) are constitutively expressed, whereas NOS-2 (also referred to as iNOS) is
 7      inducible, commonly by pro-inflammatory cytokines.  An acute exposure of rats to 2 ppm O3
 8      caused an increase in BALF macrophage number and total protein, in iNOS expression, in
 9      fibronectin, and in TNFa production by AMs (Pendino et al., 1995).  All of these effects of O3
10      were reduced by pretreatment with gadolinium chloride, a macrophage inhibitor. Macrophages
11      isolated from O3-exposed mice produced increased amounts of NO, superoxide anion, and PGE2,
12      but production of these mediators by macrophages from NOS knockout mice was not elevated
13      (Fakhrzadeh  et al., 2002). Additionally, mice deficient in NOS or mice treated with
14      NG-monomethyl-L-arginine, an inhibitor of total NOS, were protected from O3-induced
15      permeability, inflammation, and injury, suggesting a role of NO in the production of O3 effects
16      (Kleeberger et al., 2001b; Fakhrzadeh et al., 2002). These results contrast with a study showing
17      that O3 exposure produced greater injury, as determined by measurement of MIP-2, matrix
18      metalloproteinases, total protein, cell content and tyrosine nitration of whole lung protein, in
19      iNOS knockout mice than in wild type mice (Kenyon et al., 2002). This group suggests that
20      protein nitration is related to inflammation and is not dependent on iNOS-derived NO.  They
21      point out the possible experimental differences, such as O3 concentration, for inconsistency
22      between their results and those of Kleeberger et al. (2001b).
23           Rats pretreated with ebselen, a potent anti-inflammatory, immunomodulator,  and
24      NO/peroxynitrite scavenger, and then exposed to 2 ppm O3 for 4 -h had decreased numbers of
25      neutrophils, lowered albumin levels, and inhibited nitration of tyrosine residues in BALF 18 h
26      PE, though macrophage iNOS expression was not changed (Ishii et al., 2000a). These results
27      suggest an iNOS-independent mechanism for O3-induced inflammation.  Inoue et al.  (2000)
28      demonstrated in human transformed bronchial  epithelial cells that NO-generating compounds
29      (TNFa, IL-1P, and INF-y) induce IL-8 production and that NOS inhibitors inhibit IL-8
30      production.  In vivo experiments in the same study using male Hartley-strain guinea pigs
31      exposed to 3  ppm O3 for 2-h showed that NOS inhibitor pretreatment attenuated-O3 induced

        December 2005                           5-23        DRAFT-DO NOT QUOTE OR CITE

-------
 1      neutrophil recruitment and AHR at 5 hours after exposure. The NOS inhibitors also blunted the
 2      increase in nitrate/nitrite levels and in IL-8 mRNA, at the 5 hours PE. The authors hypothesize
 3      that NO, or its derivatives, facilitate AHR and inflammation after O3 exposure, possibly
 4      mediated by IL-8. Jang et al. (2002) showed a dose-dependent increases in nitrate (indicative of
 5      in vivo NO generation) with O3 exposure (0.12, 0.5, 1, or 2 ppm for 3 h). Functional studies of
 6      enhanced pause (Penh) demonstrated increases with O3 which were also dose-dependent. Western
 7      blot analysis of lung tissue showed increases in NOS-1, but not in NOS -3 or iNOS isoforms.
 8      These results suggest that in mice NOS-1 may induce airway responsiveness by a neutrophilic
 9      airway inflammation. The literature regarding the effects of O3 exposure on NOS activity is
10      complex and conflicting.  Similarly, the issue of protein nitration as it relates to cell injury due to
11      O3 exposure is somewhat controversial.
12
13      5.2.3.6  Summary and Conclusions - Inflammation and Permeability Changes
14           Airway mucosa in the normal lung serves as an effective barrier that controls bidirectional
15      flow of fluids and cells between the air and blood compartments. Ozone disrupts this function,
16      resulting in an increase in serum proteins, bioactive mediators, and PMNs in the interstitium and
17      air spaces of the lung. Generally, the initiation of inflammation is an important component of
18      the defense process; however, its persistence and/or repeated occurrence can result in adverse
19      health effects.
20           The relative influence of concentration and duration of exposure (i.e., C x T) has been
21      investigated extensively in rats, using BALF protein as an endpoint.  Although, the interaction
22      between C and T is complex, C generally dominated the response. The impact of T was
23      C-dependent (at higher Cs, the impact of T was greater); at the lowest C and T values, this
24      dependence appeared to be lost.
25           In rats, a single 3hr exposure to 0.5 ppm O3 produced a significant increase in both
26      permeability and inflammation, but a comparable exposure to 0.3 or 0.15 ppm did not produce
27      an effect. In a study comparing the responses of five species exposed to several concentrations
28      of O3, ranging from 0.2 to 2.0 ppm for 4 h, BAL was performed 18 h PE. Guinea pigs were the
29      most responsive (increased BALF protein at > 0.2 ppm); rabbits were the least responsive (effect
30      at 2.0 ppm only); and rats, hamsters, and mice were intermediate (effects at > 1.0 ppm). Among
31      rat strains, an acute exposure to O3 resulted in a greater injury, inflammation BALF levels of IL-

        December 2005                            5-24        DRAFT-DO NOT QUOTE OR CITE

-------
 1      6 in Wistar than in SD or F344 rats.  As exposures continue for 3 to 7 days, the increases in
 2      BALF protein and PMNs typically peak after a few days (depending upon species tested and
 3      exposures) and return towards control even with continuing exposure.
 4           Other factors that have been studied for potential impact on the effects of O3 include age,
 5      gender, nutritional status, genetic variability, exercise and exposure to co-pollutants. The effects
 6      of age on lung inflammation are not well known. After an acute exposure to 0.8 or  1 ppm, young
 7      mice, rats, and rabbits had  greater changes in prostaglandins in BALF, but there were no age-
 8      dependent effects on BALF protein or cell number. Comparisons of male and female animals,
 9      and vitamin C or ascorbate deficiency did not reveal significant differences in the effects of O3,
10      but exercise during exposure increased susceptibility.
11           Ozone also increases the permeability from the air to the blood compartment.  Ozone
12      (0.8 ppm; 2 h) caused a 2-fold increase of the transport of labeled DTPA from the rat tracheal
13      lumen to the blood.  This coincided with a 2-fold increase in the number of endocytic vesicles in
14      epithelial cells that contained intraluminally instilled HRP as a tracer.  These studies also suggest
15      an uneven disruption of tight junctions and alternate transport through endocytotic mechanisms.
16      In studies aimed at detecting the effects of O3 exposure on regional permeability,  O3 increased
17      the transmucosal transport of DTPA  and BSA more in the trachea and bronchoalveolar zone than
18      in the nose.  These changes in barrier integrity may allow increased entry of antigens and other
19      bioactive compounds (e.g., bronchoconstrictors) into lung tissue. Data from analyses at regular
20      intervals PR indicate that maximal increases in BALF protein, albumin and number of PMNs
21      occur 8 to 18 h (depending on the study) after an acute exposure ceases.
22          Recent studies have placed a major focus on mediators released from inflammatory cells to
23      understand the mechanisms of O3-induced inflammation and injury. Cytokines and chemokines
24      have been shown to be released as a result of stimulation or injury of macrophages,  epithelial
25      cells and PMNs. Exposure of guinea pig AMs recovered in BALF and exposed in vitro to
26      0.4 ppm O3 produced a significant increase in IL-6 and TNFa. An exposure of human AMs to
27      an identical O3 concentration increased TNFa, IL-lp, IL-6 and IL-8.  The expression of MIP-2
28      mRNA or BALF levels of  MIP-2 increased in mice and rats exposed to O3 concentrations
29      > 1 ppm. An increase after O3 exposure has also been reported for other cytokines and
30      inflammatory mediators, including CINC and fibronectin. The CINC mRNA expression was
31      associated with neutrophilia at 24 hrs PE. Ozone exposure of mice also caused an increase in

        December 2005                            5-25        DRAFT-DO NOT QUOTE OR CITE

-------
 1      IL-6, MlP-la and eotaxin in mice.  Further understanding of the role of mediators has come
 2      from studies utilizing antibodies and inhibitors of known specificity. In these studies treatment
 3      of rats with an anti IL-6 receptor antibody prior to a nighttime exposure to O3 abolished
 4      O3-induced cellular adaptive response following a subsequent exposure.  Studies utilizing
 5      antibodies to selected pro- or anti-inflammatory cytokines suggest a role of TNFa, interleukin-10
 6      (IL-10) and IL-lp in O3-induced changes in permeability, inflammation and cytokine release.
 7           Studies investigating the mechanisms of PMN recruitment in the lung have explored the
 8      role of cell adhesion molecules that mediate PMN-endothelial cell interactions.  An increase in
 9      tissue expression of 1C AM-1 occurred in mice exposed to 2 ppm O3. A comparable pattern of
10      time-related changes in total protein, fibronectin and alkaline phosphatase activity in the BALF
11      was observed in rats exposed to 1 ppm O3.  In monkeys, the O3-induced inflammation was
12      blocked by treatment with a monocolonal antibody to CD 18, suggesting dependence of PMN
13      recruitment on this adhesion molecule.  Together, these studies support the role of extracellular
14      matrix protein and cell adhesion molecules in lung inflammation and injury.
15           Ozone exposure also affects macrophage functions, and consequently their role in lung
16      inflammation. Macrophages isolated from O3-exposed mice produced increased amounts of NO,
17      superoxide anion and PGE2, but production of these mediators by macrophages from NOS
18      knockout mice was not elevated. Additionally, mice deficient in NOS or mice treated with NG-
19      monomethyl-L-arginine, an inhibitor of total NOS, were protected from O3-induced
20      permeability, inflammation and injury, suggesting a role of NO in the production of O3 effects.
21
22      5.2.4  Morphological Effects
23           Most mammalian species show generally similar morphological responses to <  1 ppm O3
24      which differ only by region, cell type, exposure parameters, and length of time between exposure
25      and examination. Constant low exposures to O3  create an early bronchoalveolar exudation,
26      which declines with continued exposure and drops in the PE period. Epithelial hyperplasia also
27      starts early, increases in magnitude for several weeks, plateaus with continuing exposure,  and
28      declines slowly during PE.  Interstitial fibrosis has  a later onset, continues to increase throughout
29      the exposure,  and can continue to increase after the exposure ends. Nonhuman primates respond
30      more than rats at this concentration, due to differences in antioxidants, the CAR (predicted to
31      receive the highest dose of O3), the presence of respiratory bronchioles, acinar volume, and

        December 2005                            5-26        DRAFT-DO NOT QUOTE OR CITE

-------
 1      differences in the nasal cavity's ability to "scrub" the O3.  Ciliated epithelial cells of the airway,
 2      Type 1 epithelial cells of the gas-exchange region, and ciliated cells in the nasal cavity are the
 3      cells most affected by O3. Ciliated cells are replaced by nonciliated cells (which are unable to
 4      provide clearance function)  and Type 1 cells are replaced by Type 2 cells, which are thicker and
 5      produce more lipids.  Inflammation also occurs, especially in the CAR, wherein the tissue is
 6      thickened as collagen accumulates.  At exposures of 0.25  ppm O3 (8 h/day, 18 mo) in monkeys,
 7      the distal  airway is remodeled as bronchiolar epithelium replaces the cells present in alveolar
 8      ducts. In both rodents and monkeys, it appears that the natural seasonal patterns of O3 exposure
 9      alters morphology more than continuous exposures, thus long-term animal studies with
10      uninterrupted exposures may underestimate morphological effects.
11
12      5.2.4.1 Short Term Exposure Effects
13           Morphological effects of key exposure studies generally lasting less than 1 week are
14      summarized in Table AX5-5.  Harkema  et al. (1997a) reviewed toxicological studies of the nasal
15      epithelial response to short-term O3. New information regarding the effects of O3 in this region
16      include demonstrations that the topical anti-inflammatory corticosteriod fluticasone propionate
17      prevents inflammation and mucous  cell metaplasia in rats after cumulative O3 exposure (0.5 ppm
18      O3, 8h/day, for 3 or 5 days) (Hotchkiss et al., 1998).  Exposure to bacterial endotoxin, a common
19      ambient air toxicant, can potentiate  mucous cell metaplasia in the nasal transitional epithelium  of
20      rats caused by a previous O3 exposure (Fanucchi et al., 1998). Male F344/N Hsd rats were
21      intranasally instilled with endotoxin after exposure to filtered air (FA) or 0.5 ppm O3, (8  h/d for 3
22      d). Mucous  cell metaplasia  was not found in the air/endotoxin group, but was found in the
23      O3/saline  group and was most severe in the O3/endotoxin group. A similar synergistic effect was
24      demonstrated by Wagner et  al. (2001a,b) with exposure of Fischer rats to O3 and endotoxin.
25      Ozone alone created epithelial lesions in the nasal transitional epithelium, while endotoxin alone
26      caused lesions in the respiratory epithelium of the nose and conducting airways.  The enhanced
27      O3-induced mucous cell metaplasia  was  related to neutrophilic inflammation.
28           Pre-metaplastic responses, such as mucin mRNA upregulation, neutrophilic inflammation,
29      and epithelial proliferation, were shown to be responsible for O3-induced mucous cell metaplasia
30      in the transitional epithelium of rats (Cho et al.,  1999a, 2000). Male F344/N rats exposed to O3,
31      (0.5 ppm,8 h/d for 1, 2, or 3  d) demonstrated a rapid increase in an airway-specific mucin gene

        December 2005                            5-27        DRAFT-DO NOT QUOTE OR CITE

-------
 1      mRNA rapidly after exposure to O3, both before and during the onset of mucous cell metaplasia.
 2      Neutrophilic inflammation coincided with epithelial DNA synthesis and upregulation of rMuc-
 3      SAC, but was resolved before the development of epithelial metaplasia. The mucous cell
 4      metaplasia was neutrophil-dependent, whereas O3-induced epithelial cell proliferation and mucin
 5      gene upregulation were neutrophil-independent.
 6           Dormans et al. (1999) compared the extent and time course of fibrotic changes in mice,
 7      rats, and guinea pigs exposed to 0.2 and 0.4 ppm O3 for 3, 7, 28, and 56 days.  They found a
 8      concentration-related centriacinar inflammation in all three species, with a maximum after 3 days
 9      of exposure and total recovery within 3 days after exposure.  Repair of O3 damage by removal of
10      injured epithelial cells is enhanced  by the influx of neutrophils (Hyde et al., 1999; Veseley et al.,
11      1999b; Miller et al., 2001; see  Section 5.2.3).  Labeling indices for rat nasal transitional
12      epithelial cell DNA were greatest 20 to 24 h after O3 (0.5 ppm for 8h) exposure, but still greater
13      than control by 36 h PE (Hotchkiss et al., 1997).
14           Very few published studies have explicitly explored susceptibility factors such as species,
15      gender, age, antioxidant defense, acute  and chronic airway disease, and exercise.  Most typical
16      laboratory  species studied have qualitatively similar effects associated with O3 exposure.
17      Dormans et al. (1999) compared morphological, histological, and biochemical effects in the rat,
18      mouse, and guinea pig after O3 exposure and after recovery in clean air. Wistar RIV:Tox male
19      rats, NIH male mice, and Hartley Crl:(HA)BR male guinea pigs were continuously exposed to
20      FA, 0.2, or 0.4 ppm for 3, 7, 28, and 56 days.  Recovery from 28 days of exposure was studied at
21      intervals of 3, 7, and 28 days PE. The mouse was the most sensitive as shown by a concentration
22      and exposure-time dependent persistence of bronchiolar epithelial hypertrophy, elevated lung
23      enzymes, and slow  recovery from exposure. Exposure to the high dose for 56 d in both rats and
24      guinea pigs caused  increased amounts of collagen in ductal septa and large lamellar bodies in
25      Type II cells.  The inflammatory response was greater in the guinea pig. Overall, the authors
26      rated mice as most susceptible, followed by guinea pigs and rats.
27           Ferrets, monkeys and rats were exposed  to O3 (1.0 ppm, 8 h) to compare airway effects
28      Sterner-Kock et al.  (2000). The ferrets and monkeys had similar epithelial necrosis and
29      inflammation that was more severe than that found in rats. Because ferrets have a similar
30      pulmonary structure as humans (e.g., well-developed respiratory bronchioles and submucosal
31      glands), the authors concluded that the ferret would be a better model than rodents for O3-

        December 2005                            5-28        DRAFT-DO NOT QUOTE OR CITE

-------
 1      induced airway effects.  Age susceptibility is dependent on the endpoint examined. One new
 2      study (Dormans et al., 1996) demonstrated that O3- induced centriacinar lesions are larger in
 3      younger rats than in older rats.
 4           Some new studies have examined O3-induced morphological effects in compromised
 5      laboratory animals. Rats with endotoxin-induced rhinitis were more susceptible to mucous cell
 6      metaplasia in the nasal transitional epithelium caused by a 3-day exposure to 0.5 ppm O3 (Cho
 7      et al.,1999b).  Wagner et al. (2002) reported a similar O3-induced enhancement of inflammatory
 8      and epithelial  responses associated with allergic rhinitis. Brown Norway rats were exposed to
 9      0.5 ppm O3, 8h/day for 1 day or 3 consecutive days and then immediately challenged intranasally
10      with either saline or ovalbumin (OVA). Multiple exposures to O3 caused greater increases in
11      mucosubstances produced in the nose by allergen challenge.
12           Recent research has focused on the concept of O3 susceptible and non-susceptible sites
13      within the respiratory tract, including in situ antioxidant status and metabolic activity.  Plopper
14      et al. (1998) examined whether the variability of acute epithelial injury to short-term O3 exposure
15      within the tracheobronchial tree is related to local tissue doses of O3 or to local concentrations of
16      reduced glutathione (GSH). Adult male rhesus monkeys exposed to O3 (0.4 or 1.0 ppm for 2 h)
17      demonstrated  significant cellular injury at all sites, but the most damage, along with increased
18      inflammatory  cells, occurred in the proximal respiratory bronchiole. A significant reduction in
19      GSH was found in the proximal bronchus at 0.4 ppm O3 and in the respiratory bronchiole at
20      1.0 ppm O3. A significant decrease in the percent of macrophages, along with significant
21      increases in the percent of neutrophils and eosinophils, and a doubling of total lavage protein,
22      were found after exposure to 1.0 ppm O3 only.  The authors concluded that the variability of
23      local O3 dose in the respiratory tract was related to inhaled O3 concentration and was closely
24      associated with local GSH depletion and with the degree of epithelial injury.
25           Plopper  and colleagues (e.g., Watt et al.,1998; Paige et al., 2000)  explored the site-specific
26      relationship between epithelial effects of O3 exposure and the metabolism of bioactivated
27      compounds within the respiratory tract of rats.  The distribution of CYP2E1-dependent activity,
28      measured with a selective substrate (p-nitrocatechol), was found to be highest in the distal
29      bronchioles and minor daughter airways, and lower in the lobar bronchi and major daughter
30      airways.  Short-term O3 exposure (1 ppm for 8 h) increased CYP2E1 activity in the lobar
31      bronchi/major daughter airways only; however,  long-term O3 exposure (1 ppm for 90 days)

        December 2005                            5-29        DRAFT-DO NOT QUOTE OR CITE

-------
 1      decreased CYP2E1 activity in the major and minor airways, further complicating the
 2      interpretation of O3 effects based on concentration and duration of exposure and recovery. Rats
 3      treated i.p. with 1-nitronaphthalene, a pulmonary toxicant requiring metabolic activation, and
 4      exposed to 0.8 ppm O3, 8h/day for 90 days showed greater histopathologic and morphometric
 5      effects in the CAR of the lung (Paige et al., 2000). Despite reported tolerance to oxidant stress
 6      after long-term O3 exposure, there was increased severity of ciliated cell toxicity.
 7
 8      5.2.4.2  Summary of Short-Term Morphological Effects
 9           Short-term exposures to O3 cause similar alterations in lung structure in a variety of
10      laboratory animal  species at concentrations of 0.15 ppm in rats and lower concentrations in
11      primates. Cells in the CAR are the primary targets of O3, but ciliated epithelial cells in the nasal
12      cavity and airways and Type  1 epithelial cells in the gas exchange region  are also targets. New
13      work has shown that a topical anti-inflammatory corticosteroid can prevent these effects in nasal
14      epithelia, while exposure to bacterial endotoxin can potentiate the effects.  Ozone-induced
15      fibrotic changes in the CAR are maximal at 3 d of exposure and recover 3 d PE with exposures
16      of 0.2 ppm in rodents.  New studies of susceptibility factors demonstrated that ferrets and
17      monkeys have similar inflammatory and necrotic responses to 1 ppm O3, which differs from
18      lesser injury seen in rats. Rats with induced allergic rhinitis are more susceptible to 0.5 ppm
19      than are controls.  Important new work has demonstrated variability of local O3 dose and
20      subsequent injury  in the RT due to depletion of GSH.  The proximal respiratory bronchiole
21      receives the most acute epithelial injury from exposures < 1 ppm, while metabolic effects were
22      greatest in the distal bronchioles and minor daughter airways.
23
24      5.2.4.3  Long Term Exposure Effects
25           Summaries of new studies of morphological  effects of exposure lasting longer than 1 week
26      are listed in Table AX5-6 in Annex AX5.  In general,  as the duration of exposure lengthens,
27      there is not a concomitant linear increase in the intensity of effect of a given endpoint.  Rather, as
28      exposure proceeds past 1 week to 1 year, Type 1 cell necrosis and inflammatory responses
29      generally decrease to near control values, and hyperpalstic and fibrotic changes remain elevated.
30      After long-term exposure ended, some indicies of fibrosis persisted and in some  cases became
31      more severe during PE periods in clean air.

        December 2005                            5-30        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Effects of O3 on the upper respiratory tract of F344 rats exposed to O3 (0.12, 0.5, or
 2      1.0 ppm for 20 months) included marked mucous cell metaplasia in the rats exposed to 0.5 and
 3      1.0 pm O3, but not at 0.12 ppm O3 (Harkema et al., 1997a) . In a follow-up study, hyperplasia
 4      was found in the nasal epithelium of rats exposed to 0.25 and 0.5 ppm, 8h/day, 7 days/week, for
 5      13 weeks (Harkema et al., 1999).  The mucous cell metaplasia, and associated intraepithelial
 6      mucosubstances, induced by 0.5 ppm O3 persisted for 13 weeks after exposure.  An acute (8-h)
 7      exposure to 0.5 ppm O3 13 weeks after the chronic exposure induced an additional increase of
 8      mucosubstances in the nasal epithelium of rats but not in rats chronically exposed to 0 or
 9      0.25 ppm O3.  The persistent nature of the O3-induced mucous cell metaplasia in rats reported in
10      this study suggests that O3 exposure may have the potential to induce similar long-lasting
11      alterations in the airways of humans.
12           No significant changes in nasal tissue were seen in rats continuously exposed for 49 days
13      to the ambient air of Mexico City, Mexico (Moss et al., 2001), which is in contrast to two rat
14      studies which did demonstrate development of secretory hyperplasia in rats exposed to ambient
15      air of Sao Paulo (Saldiva et al., 1992; Lemos et al., 1994).  Because of the persistent nature of
16      these changes in the controlled studies with rats, and the fact that the upper airways of humans
17      are probably more sensitive, like the monkey, the authors suggested that long-term exposure to
18      ambient levels of O3 could induce significant nasal epithelial lesions that may compromise the
19      upper respiratory tract defense mechanisms of exposed human populations.
20           Rats exposed to 0.5 ppm O3 for 1 month  exhibited Bcl-2 in protein extracts of nasal
21      epithelium (Tesfaigzi  et al., 1998). Further, after 3 and 6 months of exposure, the number of
22      metaplastic mucous cells in the transitional epithelium was indirectly related to the percentage of
23      cells that were Bcl-2 positive . Cells from rats exposed to FA did not express any Bcl-2.  This
24      study suggests that apoptosis regulators like Bcl-2 may play a role in the development and
25      resolution of mucous cell metaplasia in the nasal airway
26           A spectrum of lesions was reported (Herbert et al., 1996) in the nasal cavity and
27      centriacinar lung of male and female mice exposed to 0.5 or 1.0 ppm of O3 for 2 years, which
28      persisted with continued exposure for 30 months. These lesions included bone loss in the
29      maxilloturbinates, mucosal inflammation, mucous cell metaplasia in the nasal transitional
30      epithelium and increased interstitial and epithelial thickening in the proximal  alveolar region.
31      In the CAR, there were increased numbers of nonciliated cells. However, changes in other

        December 2005                            5-31        DRAFT-DO NOT QUOTE OR CITE

-------
 1      endpoints including lung function and lung biochemistry were not evident. The investigators'
 2      interpretation of the entire study is that rodents exposed to the two higher O3 concentrations had
 3      some structural hallmarks of chronic airway disease in humans. This interpretation is
 4      strengthened further by a comparative pathology study of young (11 to 30 years old) accident
 5      victims from Los Angeles, CA, and Miami, FL, (Sherwin et al., 2000) showing increased scores
 6      for extent and severity of chronic inflammation in the CAR of Los Angeles residents. (See
 1      Chapter 7 for a more detailed discussion of the population-based studies on O3.)
 8           A long-term study using a simulated, seasonal O3-exposure pattern in infant monkeys was
 9      reported by Plopper and colleagues (Evans et al., 2003; Schelegle et al., 2003a; Chen et al., 2003
10      Plopper and Fanucchi, 2000; Fanucchi et al., 2000) using. Infant rhesus monkeys (30 days old)
11      were exposed to FA, house dust mite allergen aerosol (HDMA), or O3 + HDMA.  The 0.5 ppm
12      O3 exposures were 8 h/day for 5 days, every 14 days for a total of 11 O3 episodes.  Half of the
13      monkeys were sensitized to house dust mite allergen (Dermatophagoides farinae) at 14 and
14      28 days of age.  The sensitized monkeys were exposed to HDMA for 2h/day on Days 3-5 of the
15      FA or O3 exposures. The lungs were removed during the last FA exposure and the right and left
16      cranial and right middle lobes were separately inflation fixed.  Microdisection and morphometric
17      analyses were performed on the conducting airways to the level of the most proximal respiratory
18      bronchiole. Repeated exposures to O3 or O3 + HDMA over a 6-month period resulted in an
19      atypical development of the basement membrane zone of airways in nonsensitized developing
20      monkeys.  Remodeling in the distal conducting airways was found in the sensitized monkeys as a
21      result of the damage and repair processes occurring with repeated exposure (Evans et al., 2003;
22      Schelegle et al., 2003a; Fanucchi et al., 2000).  Lung function changes in these monkeys
23      (Schelegle et al., 2003b), and associated adaptation of the respiratory motor responses (Chen
24      et al., 2003), are described in Section 5.2.5.2.  Collectively, these findings provide a
25      pathophysiologic basis  for changes in airway function described in children growing up in
26      polluted metropolitan areas (e.g., Tager, 1999) (See Chapter 7).
27           Necropsy of the left caudal lobe of these infant monkeys showed accumulation of
28      eosinophils and mucous cells within the combined epithelium and interstitium compartments in
29      the conducting airways and in the terminal/respiratory bronchioles (Schelegle  et al., 2003a) .
30      House dust mite sensitization and HDMA challenge alone, or combined with O3 exposure,
31      resulted in significantly greater eosinophil accumulation in the conducting airways when

        December 2005                            5-32        DRAFT-DO NOT QUOTE OR CITE

-------
 1      compared to FA and O3 only exposures.  A significant accumulation of eosinophils was found in
 2      the terminal/respiratory bronchioles of the sensitized monkeys challenged with HDMA when
 3      compared to monkeys exposed to FA, O3, and FIDMA + O3. The mean mass of mucous cells
 4      increased in the fifth generation conducting airways of sensitized animals challenged with
 5      FIDMA alone and when combined with O3  exposure, and in the terminal bronchioles of
 6      sensitized animals exposed to FIDMA + O3. The tracheal basement membrane of HDMA-
 7      sensitized monkeys exposed to FIDMA or to HDMA + O3 was significantly increased over
 8      controls; however, there were no significant changes in the airway diameter of proximal and
 9      mid-level airways.  The authors interpreted these findings to indicate that the combination of
10      cyclic O3 exposure and HDMA challenge in HDMA-sensitized infant monkeys act
11      synergistically to produce an allergic-reactive airway phenotype characterized by significant
12      eosinophilia of midlevel conducting airways, transmigration of eosinophils into the lumen, and
13      an altered structural development of conducting airways that is associated with increased airway
14      resistance and nonspecific airway reactivity (see Section 5.2.5).  Exposures of sensitized young
15      monkeys to HDMA alone, or to O3 alone, resulted in eosinophilia of the mid-level conducting
16      airways and the terminal/respiratory bronchioles, but without alterations in airway structure or
17      function.
18           Examination of development of the tracheal basement membrane zone (BMZ) in these
19      monkeys (Evans et al., 2003) showed that with exposures to either O3 or HDMA + O3 BMZ
20      development was affected.  Abnormalities in the BMZ included: (1) irregular and thin collagen
21      throughout the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
22      immunoreactivity was reduced; (4) FGF-2 immunoreactivity was absent in perlecan-deficient
23      BMZ, but was present in  the lateral intercelluar space (LIS), in basal cells, and in attenuated
24      fibroblasts; (5) syndecan-4 immumoreactivity was increased in basal cells.  The authors interpret
25      these data to suggest that O3 targets cells associated with synthesis of epithelial BMZ perlecan.
26      The absence of FGF-2, normally stored in the BMZ, could affect downstream signaling in
27      airway epithelium and could be responsible for the abnormal development of the airway seen in
28      this study, and thus be an important mechanism modulating O3-induced injury. Midlevel bronchi
29      and bronchioles from these monkeys (Larson et al., 2004) demonstrated decrements in the
30      density of epithelial nerves in the axial path between the sixth and seventh airway generations in
31      exposures to O3.  Combined O3+HDMA  exposures exacerbated this reduction.  They attribute

        December 2005                           5-33        DRAFT-DO NOT QUOTE OR CITE

-------
 1      this loss of nerve plexuses to neural regression or stunted nerve development, the latter
 2      corroborated by the Evans et al. (2003) finding of decreased growth factors following O3
 3      exposure. Additionally, they found streaks or clusters of cells immunoreactive for protein gene
 4      product 9.5 (PGP 9.5, a pan-neuronal marker) and negative for calcitonin gene-related peptide.
 5      The functional significance of this is unknown but suggests to the authors a possible injury-
 6      repair process induced by O3.
 7           Remodeling of the distal airways and CAR is one of the most disturbing aspects of the
 8      morphological changes occurring after long-term exposure to O3. Recently, bronchiolization
 9      was reported in rats exposed to 0.4 ppm O3 for only 56 days (van Bree et al., 2001).  They also
10      found collagen formation progressively increased with increasing O3 exposure and remained
11      increased into PE recovery. In addition to centriacinar remodeling, Pinkerton et al. (1998)
12      reported thickening of tracheal, bronchial, and bronchiolar epithelium after 3 or 20 months
13      exposure to 1 ppm O3, but not to 0.12 ppm. Although some older literature had reported that
14      chronic exposures to < 1.0 ppm O3 cause emphysema, no current literature supports this
15      hypothesis.
16
17      5.2.4.4  Summary and Conclusions - Long-Term Morphological Effects
18           The progression of effects during and after a chronic exposure at a range of 0.5 to 1.0 ppm
19      is complex, with inflammation peaking over the first few days of exposure, then dropping, then
20      plateauing, and finally, largely disappearing.  Epithelial hyperplasia follows a somewhat similar
21      pattern. In contrast, fibrotic changes in the tissue increase very slowly over months of exposure,
22      and, after exposure ceases, the changes sometimes persist or increase. Pattern of exposure in this
23      same concentration range determines effects, with 18 mo of daily exposure causing less
24      morphologic damage than exposures on alternating months. This is important as environmental
25      O3 exposure is typically seasonal. Plopper and colleagues' long term study of infant rhesus
26      monkeys exposed to simulated, seasonal O3 (0.5 ppm 8h/day for 5  days, every 14 days for 11
27      episodes) demonstrated : 1) remodeling in the distal airways; 2)  abnormalities in tracheal
28      basement membrane; 3) eosinophil accumulation in conducting airways; 4) decrements in airway
29      innervation.  These findings advance earlier information regarding possible injury-repair
30      processes occurring with seasonal O3 exposures.
31

        December 2005                           5-34        DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.2.5 Effects on Pulmonary Function
 2      5.2.5.1  Acute and Short-Term Exposure Effects on Pulmonary Function
 3           Numerous pulmonary function studies of the effects of short-term O3 exposure (defined
 4      here as < 1 week of exposure) in several animal species have been conducted and generally show
 5      responses similar to those of humans (e.g., increased breathing frequency, decreased tidal
 6      volume, increased resistance, decreased forced vital capacity [FVC] and changes in the
 7      expiratory flow-volume curve).  These effects are seen at 0.25 to 0.4 ppm O3 for several h in a
 8      number of species. At concentrations  of > 1 ppm, breathing mechanics (compliance and
 9      resistance) are affected.  The breathing pattern returns to normal after O3 exposure. In rats
10      exposed to 0.35 to  1 ppm O3 for 2 h/day for 5 days, there is a pattern of attenuation of pulmonary
11      function responses similar to that observed in humans.  Concurrently, there was no attenuation of
12      biochemical indicators of lung injury or of morphological changes.
13           New work demonstrating attenuation of pulmonary functions was completed by Wiester
14      et al.  (1996) who exposed male Fischer 344 rats to 0.5 ppm O3 for either 6 or 23 h/day over 5
15      days. Ozone-induced changes in lung volume were attenuated during the 5 exposure days and
16      returned to control levels after 7 days recovery. The responses to repeated O3 exposure in rats
17      were  exacerbated by reduced ambient temperature, presumably as a result of increased metabolic
18      activity.
19           Researchers have utilized inbred mouse strains with varying ventilatory responses to O3 to
20      attempt to model susceptible populations.  As differences were seen in inflammatory responses
21      to acute O3 exposures in C57BL/6J and C3FI/HeJ mice, comparisons were made of their
22      ventilatory responses also (Tankersley et al., 1993). Following an exposure of 2 ppm O3 for 3 h,
23      breathing frequency (f), tidal volume (VT), and minute ventilation were measured 1 and 24 h in
24      both normocapnia (or air at -0% CO2) and hypercapnia (5  or 8% CO2). They demonstrated that
25      acute O3 exposures caused altered hypercapnic ventilatory control, which varied between strains.
26      This suggested to the authors that O3-induced alterations in ventilation are determined, at least in
27      part, by genetic factors.
28           Paquette et al. (1994) measured ventilatory responses in C57BL/6J  and C3H/HeJ mice
29      given repeated subacute  exposures. The two strains had differing responses to both normocapnia
30      and hypercapnia. Normocapnic VE was greater following subacute O3 exposure in C57BL/6J
31      mice  than in C3H/HeJ mice, due to increased and reduced  VT, respectively. This suggests that

        December 2005                            5-35        DRAFT-DO NOT QUOTE OR CITE

-------
 1      the increased VT in C57BL/6J mice may contribute to the increased susceptibility to lung injury
 2      due to a greater dose of O3 reaching the lower lung.  Hypercapnic ventilatory responses
 3      following subacute O3 exposures demonstrated reduced VE (due to decreased VT) in C57BL/6J
 4      only. Evaluations of O3 dosimetry were performed in these two strains using 18O3-labeled ozone
 5      (Slade et al., 1997). Immediately after exposures of 2 ppm 18O3 for 2-3 h, C3H/HeJ mice had
 6      46% less 18O in lungs and 61% less in trachea, than C57BL/6J.  Additionally, C3H/HeJ mice had
 7      a greater body temperature decrease following O3 exposure than C57BL/6J mice, suggesting that
 8      the differences in susceptibility to O3 are due to differences the ability to decrease body
 9      temperature and, consequently decrease the dose of O3 to the lung.
10           Tracheal transepithelial potential (VT) has also been shown to differ in eight mouse strains
11      6 h after exposure to 2 ppm O3 for 3 h (Takahashi et al., 1995b).  AKR/J, C3H/HeJ, and CBA/J
12      were identified as resistant strains and  129/J, A/J, C57BL/6J, C3HeB/FeJ and SJL/J were
13      identified as susceptible strains. The authors noted that strains' responses to this parameter did
14      not show concordance with inflammatory responses, suggesting to the authors that the two
15      phenotypes are not controlled by the same genetic factors.
16           Savov et al. (2004) characterized ventilatory responses in nine mouse strains exposed to O3
17      (2.0 ppm O3 for 3 h). Table AX5-4 in Annex AX5 lists the baseline Penh, the Penh following O3,
18      and the Penh response to methacholine (MCh) following O3. C57BL/6J was hyporeactive to MCh
19      prior to O3, but was very responsive to MCh following O3.  Conversely, C3H/HeJ had an
20      intermediate baseline Penh and a small response to MCh following O3 exposure. This study
21      corroborates the evidence of no consistent relationship between respiratory Penh and
22      inflammation.
23
24      5.2.5.2  Summary and Conclusions - Short- and Long-Term Effects on
25              Pulmonary Function
26           Early work has demonstrated that during acute exposure of -0.2  ppm O3 in rats, the most
27      commonly observed alterations are  increased frequency of breathing and decreased tidal volume
28      (i.e., rapid, shallow breathing). Exposures of-1.0 ppm O3  affect breathing mechanics
29      (compliance and resistance). Additionally, decreased lung volumes are observed in rats with
30      acute exposures at levels of 0.5 ppm. New work utilizing inbred mouse strains with varying
31      ventilatory responses to O3 has suggested that: (1) control of the ventilatory response is

        December 2005                            5-36        DRAFT-DO NOT QUOTE OR CITE

-------
 1      determined, at least in part, by genetic factors; (2) increased VT in some strains may contribute
 2      to lung injury due to a greater dose of O3 reaching the lower lung; (3) some strains' ability to
 3      reduce body temperature may account for their decreased O3-induce lung injury; (4) tracheal
 4      transepithelial potential is determined, in part, by genetic factors.  Importantly, the genetic loci
 5      that appear to be modulating various aspects of pulmonary responses to O3 differ from each other
 6      and from loci controlling inflammatory responses.
 7           Exposures of 2 h/day for 5 days create a pattern of attenuation of pulmonary function in
 8      both rats and humans without concurrent attenuation of lung injury and morphological changes,
 9      indicating that the attenuation did not result in protection against all the effects of O3.  Long-term
10      O3 exposure studies evaluating pulmonary function are not available. Earlier work has
11      demonstrated that repeated daily exposure of rats to an episodic profile of O3 caused small, but
12      significant decrements in lung function that were consistent with early indicators of focal
13      fibrogenesis in the proximal alveolar region, without overt fibrosis.
14
15      5.2.5.3  Ozone Effects on Airway Responsiveness
16           Effects of O3 on airway reactivity have been observed in a variety of species at an exposure
17      range of 0.5 to 1 ppm. Many of the new studies on pulmonary function in laboratory animals
18      allow a better prediction of the effects of O3 exposure on the exacerbation of asthma symptoms
19      and the risk of developing asthma in humans.  However, it is necessary to understand the factors
20      that determine airway responsiveness across different mammalian species as discussed in
21      Chapter 4.
22           The physiological characteristics of asthma include intermittent airway obstruction and
23      increased airway responsiveness to various  chemical and physical stimuli.  Methods used to
24      assess airway responsiveness in humans include airway challenge with nonspecific
25      bronchoconstrictors (e.g., inhaled methacholine or histamine) and with indirect (e.g., inhalation
26      of adenosine monophosphate, hypertonic saline, mannitol) stimuli to bronchoconstriction
27      (Anderson, 1996). Laboratory animal studies employ intravenous agonist challenges as well as
28      inhalation challenges, though inhaled agonist challenges are preferred  in humans.  Sommer et al.
29      (2001) reported  some differences in the two routes for bronchoconstrictor administration.
30           Traditional studies of airway responsiveness require sedation in both infants and laboratory
31      animals. Exercise testing is not possible with sedation unless exercise is "simulated" by

        December 2005                             5-37        DRAFT-DO NOT QUOTE OR CITE

-------
 1      increasing ventilation using elevated F;CO2 and the need for artificial ventilation in laboratory
 2      animal studies may cause breathing patterns that affect O3 deposition.  load et al. (2000) reported
 3      that when 1 ppm O3 for 90 min is administered to isolated rat lung at either 2.4 ml/40 bpm or
 4      1.2 m/80 bpm, the more rapid breathing pattern elicits less epithelial cell injury than the slower
 5      breathing pattern.  Though this study design does not really model rapid shallow breathing
 6      elicited in the intact animal, it shows greater reduction in injury in the proximal axial airway
 7      compared to its adjacent airway branch and terminal bronchiole. The rapid, shallow breathing
 8      pattern protects the large conducting airways of rats, but causes a more even distribution of
 9      epithelial cell injury to the terminal bronchioles (Schelegle et al., 2001). Postlethwait et al.
10      (2000) demonstrated that the  conducting airways are the primary site of acute cytotoxicity from
11      O3 exposure.  Three-dimensional mapping of the airway tree in SD rat isolated lung exposed to
12      0, 0.25, 0.5, or 1.0 ppm O3 showed a concentration-dependent increase in injured cells.  Injury
13      was evident in proximal and distal conduction airways,  lowest in terminal bronchioles, and
14      highest in the small side branches downstream of bifurcations.  These exposure levels did not
15      concurrently elicit changes in LDH activity or total protein in BALF, suggesting that the
16      mapping technique is a more  sensitive measure of injury and is useful in dosimetry studies.
17           Whole-body plethysmography of unanesthetized, unrestrained rodents has been used to
18      indirectly measure pulmonary resistance (Shore et al., 2001, 2002; Goldsmith et al., 2002; Jang
19      et al., 2002). However, these indices of inspiratory/expiratory pressure differences, including
20      enhanced pause (Penh) may be less sensitive than direct measurements of lung airflow resistance
21      (Murphy, 2002). Sommer et  al. (1998) demonstrated that unrestrained guinea pigs have a daily
22      variability in pulmonary resistance that is similar to that occurring in humans, indicating that
23      circadian rhythms of airway caliber must be considered when performing airway challenge tests
24      in any species. Changes in airway structure caused by viral infections also must be considered
25      when evaluating laboratory animal studies. Animals with acute viral illness have morphological
26      evidence of inflammatory cell infiltration,  bronchiolar wall edema, epithelial hyperplasia, and
27      increased airway mucous plugs that can cause airway narrowing, air trapping, and serious
28      functional changes in the lung (Folkerts et al., 1998).
29           Exercise-induced bronchoconstriction in humans appears to be mediated by changes in the
30      tonicity of the airway lining fluid (Anderson and Daviskas, 2000). Brannan et al. (1998) suggest
31      that a test in laboratory animals based on the inhalation  of mannitol  aerosol (hyperosmolar)

        December 2005                            5-38        DRAFT-DO NOT QUOTE OR CITE

-------
 1      might be feasible and provide information similar to that from exercise challenges in cooperative
 2      children and adults. Unfortunately, there have been few reports of mannitol or adenosine
 3      monophosphate challenges in laboratory animals; most studies have utilized histamine,
 4      methacholine, acetylcholine, or carbachol to determine outcome.  In active humans with asthma,
 5      adenosine monophosphate challenges appear to better reflect ongoing airway inflammation than
 6      histamine or methacholine challenges (Polosa and Holgate, 1997; Avital et al, 1995a,b), and
 7      might be useful in identifying mechanisms of asthma in laboratory animals and their
 8      responsiveness to environmental pollutants.
 9
10      Airway Responsiveness in Asthma
11           The increased responsiveness to bronchoconstrictor challenge in asthma is thought to
12      result from a combination of structural and physiological factors that include increased
13      inner-wall thickness, increased smooth-muscle responsiveness, and mucus secretion. These
14      factors  also are likely to determine a level of innate airway responsiveness that is genetically
15      influenced.  This baseline responsiveness is thought to be modulated in asthma by chronic
16      inflammation and airway remodeling (Stick, 2002). For example, about 90% of children with
17      asthma symptoms in the previous year will exhibit increased airway responsiveness to one or
18      more challenge tests (Sears et al., 1986); however, 10% of healthy children also will respond to
19      one or other of the challenge tests. Longitudinal studies in adults have shown that the
20      development of airway responsiveness is associated with persistence of symptoms (O'Conner
21      et al., 1995) suggesting airway remodeling.  This hypothesis is in agreement with the
22      inconsistent relationship reported in the  literature between airway responsiveness and markers of
23      inflammation.
24
25      Airway Responsiveness in Infants
26           The age at which nonspecific AHR is first seen in humans is unknown, but it is known that
27      infants  show increased responsiveness compared to older children, probably due to differences in
28      dose received. When correction is made for this dose effect, infants and older children appear to
29      have  a similar response to inhaled histamine (Stick et al., 1990; Stick, 2002). The importance of
30      this observation is that absolute values of airway responsiveness cannot be used to compare
31      airway  responsiveness at different ages,  and  they certainly cannot be used to compare

        December 2005                           5-39        DRAFT-DO NOT QUOTE OR CITE

-------
 1      responsiveness across different species. However, airway responsiveness can be tracked over
 2      time within given populations, or alternatively, by using non-parametric analyses based on
 3      ranking subjects at each time point.  Such analyses have been used in birth-cohort studies to
 4      investigate the role of airway responsiveness in the early genesis of asthma.  One unique
 5      birth-cohort study has shown that airway responsiveness at one month is a predictor of lung
 6      function at six years (Palmer et al., 2001). Data from this study also show that the genetic
 7      determinants of atopy and airway responsiveness are independent (Palmer et al., 2000).
 8      In another study of infants with wheeze, persistence of AHR was associated with persistence of
 9      symptoms, although airway responsiveness at one month of age was neither a sensitive nor a
10      specific predictor of outcome (Delacourt et al., 2001).
11           The human studies imply that airway responsiveness is a key factor in asthma, but it is not
12      clear if the factors that are important for airway responsiveness in early life are related to
13      inflammation, structure or physiology of the airways, or the combination of all three.
14      Furthermore, it is not clear how viruses, allergens and irritants in the environment modify innate
15      airway responses (Holt et al., 1999), but they are known to be important.  Laboratory animal
16      studies have tried to answer some of these key questions.
17
18      Airway Responsiveness in Laboratory Animals
19           A large data base of laboratory animal research has been collected on the role of O3 in
20      producing an increase in AHR. Exposure levels (> 1 ppm for > 30 min) in many of these studies
21      are not environmentally relevant, but information can be obtained regarding the mechanisms of
22      action of O3 concerning: O3 concentration and  peak response time, inhaled versus intravenous
23      challenge with nonspecific bronchoconstrictors, neurogenic mediation, neutrophilic
24      inflammation, and interactions with specific biological agents (e.g., antigens and viruses).
25           Many species of laboratory animals have been used to study the effects of O3 on airway
26      bronchoconstriction.  Ozone-induced AHR in guinea pigs has been used to model human
27      bronchospasm (Kudo et al., 1996; van Hoof et al., 1996; 1997a; Matsubara et al., 1997a,b;
28      Sun and Chung, 1997; Aizawa et al., 1999a,b;  Tsai et al., 1998; Nakano et al., 2000).  Because
29      these studies were done at 2 to 3 ppm O3, these results are not directly relevant for extrapolation
30      to potential airway responses in humans exposed to ambient levels of O3. Humans with reactive
31      airway disease (e.g., asthma) appear to be sensitive to  ambient levels of O3 (see Chapters 6

        December 2005                           5-40        DRAFT-DO NOT QUOTE OR CITE

-------
 1     and 7) and the current understanding is that O3 exacerbates airway responsiveness to specific
 2     allergens, presumably by nonspecifically increasing AHR.
 3           Shore et al. (2000, 2003) have shown that O3-induced AHR is reduced in immature rats and
 4     mice.  SD rats exposed to 2 ppm O3 at ages 2, 4, 6, .8, or 12 weeks and A/J mice exposed to
 5     0.3 to 3 ppm for 3 h at age 2, 4, 8, or 12 weeks had similar concentration-related decreases in
 6     VE except at the youngest ages. This smaller decrement in VE suggested a delivered dose that
 1     was much greater in the younger animals.  This group (Shore et al., 2003) has also recently
 8     shown that obese mice have greater ventilatory responses to O3. Exposures of 2.0 ppm O3 for 3
 9     h to lean, WT C57BL/6J and ob/ob mice (mice with a genetic defect in the  coding for leptin, the
10     satiety hormone) showed that the ob/ob mice had enhanced AHR and inflammation compared to
11     the WT mice. These data correlate with epidemiological data showing increased incidence of
12     asthma in overweight children.
13           Increased AHR to various nonspecific bronchoconstrictive agents (e.g., ACh,
14     methacholine, histamine, carbachol) given by inhalation or intravenous routes has been
15     previously shown in laboratory animals exposed to O3 concentrations < 1.0 ppm (Table 5-7).
16     Recently, Dye et al. (1999) showed hyperresponsiveness to methacholine in rats 2 h after
17     exposure to 2 ppm O3 for 2 h. AHR can be induced by specific antigens as well as O3. The most
18     commonly used laboratory animal model is the OVA sensitized guinea pig. Animals sensitized
19     with OVA have been shown to have similar responses to nonspecific bronchoconstrictors as
20     control animals; however, OVA-sensitized guinea pigs exposed to O3 showed increased AHR to
21     histamine (Vargas et al., 1994), which was further enhanced by an antigen  challenge. When
22     exposed to O3 before sensitization, repeated exposures to very high levels (5.0 ppm) decreased
23     the OVA sensitization threshold; however, in already sensitized animals, a  2-h exposure to > 1.0
24     ppm enhanced airway responsiveness to OVA, suggesting that O3 exposure does not modify the
25     development of antigen-induced AHR and, in fact, may enhance AHR at high levels of exposure.
26           OVA-sensitized guinea pigs (Sun et al., 1997) and mice (Yamauchi et al., 2002) were used
27     to determine the enhancement of antigen-induced bronchoconstriction by acute, high-level O3
28     (1.0 ppm O3 for 1 h). Male Dunkin-Hartley guinea pigs were sensitized by i.p. injection of OVA
29     and exposed to  O3 alone, OVA aerosol, or O3 + OVA. Ozone exposure alone increased
30     bronchial responsiveness to ACh at 3 h, but not 24 h, while OVA alone had no effect.  Combined
31     exposure to O3 and OVA (1 ppm for 1 h, then 3 min OVA) increased bronchial responsiveness to

       December 2005                           5-41         DRAFT-DO NOT QUOTE OR CITE

-------
 1      ACh 3 h after O3 exposure.  At 24 h following O3 exposure, AHR increased when OVA
 2      challenge was performed at 21 h, suggesting that O3 pre-exposure can potentiate OVA-induced
 3      AHR.  Neutrophil counts in the BALF increased at 3 and 24 h after O3 exposure alone but were
 4      not further increased when O3 exposure was combined with OVA airway challenge; however
 5      protein content of the BALF did increase at 3 and 24 h in the O3 and OVA groups. Thus, this
 6      study also indicates that high-ambient O3 exposure can augment antigen (OVA)-induced AHR in
 7      guinea pigs.
 8          Yamauchi et al. (2002) sensitized male C57BL/6 mice by i.p. injection of OVA and then
 9      exposed them to O3. The sensitized mice had AHR to methacholine.  Ozone exposure caused
10      significant decreases in dynamic lung compliance, minute ventilation, and PaO2 in OVA-
11      sensitized mice, but not in controls. A marker of inflammation  (soluble intercellular adhesion
12      molecule-1 [sICAM-1]) was elevated in the BAL fluid of OVA-sensitized mice, but sICAM-1
13      levels were not significantly changed by O3 exposure, indicating that the O3-induced AHR to
14      methacholine was not caused by O3-induced inflammation.
15          Ozone-induced AHR may be temporally associated with neutrophils (DeLorme et al.,
16      2002) and other inflammatory cells stimulated by leukotrienes (Stevens et al.,  1995a), cytokines
17      (Koto et al., 1997),  mast cells (Igarashi et al., 1998; Noviski et al., 1999), or by oxygen radicals
18      (Takahashi et al., 1993; Stevens et al., 1995b; Tsukagoshi et al., 1995; Kudo et al., 1996).  Two
19      new studies, however, have shown that inflammation is not a prerequisite of AHR (Stevens et al.,
20      1994; Koto et al., 1997), and some investigators have suggested that O3-induced AHR may be
21      epithelium dependent (Takata et al., 1995; Matsubara et al., 1995; McGraw et al., 2000).
22      For example, neonatal rats pretreated with capsaicin, which will permanently destroy C-fibers
23      and prevent O3-induced release of neuropeptides (Vesely et al.,  1999a), and then exposed to O3
24      when adults,  showed a marked increase in airway responsiveness to inhaled aerosolized
25      methacholine (Jimba et al., 1995). Some investigators (Matsumoto et al., 1999; DeLorme et al.,
26      2002) have shown that respective intravenous pretreatment with neutrophil elastase inhibitor or
27      PMN antiserum can block O3-induced AHR; other investigators (Koto et al., 1995; Aizawa et al.,
28      1997; Takebayashi  et al., 1998) have shown that depletion of tachykinins by capsaicin treatment,
29      or by a specific tachykinin receptor antagonist, can block the induction  of AHR by O3.  The
30      seemingly disparate responses in laboratory animals may be due to species- or strain-specific
31      differences in inherent reactivity to bronchoconstrictors, or to inherent differences in

        December 2005                           5-42        DRAFT-DO NOT QUOTE OR CITE

-------
 1      susceptibility to O3-induced inflammation (Zhang et al., 1995; Depuydt et al., 1999; Dye et al.,
 2      1999).
 3           Studies that may be potentially relevant to ambient levels of O3 were conducted in vivo, in
 4      an isolated perfused lung model, and in ex vivo lung segments using multihour and repeated
 5      multihour exposures with ambient levels of O3. A study on the relationship between O3-induced
 6      AHR and tracheal epithelial function was conducted in New Zealand white rabbits by Freed
 7      et al. (1996). Rabbits exposed to O3 (0.2 ppm for 7 h) demonstrated significantly decreased PD
 8      but no changes in lung resistance. Changes in the compartmentalized lung resistance, measured
 9      in response to ACh challenge before and after bilateral vagotomy, were not significantly
10      different in air-exposed rabbits; however, bilateral vagotomy did enhance peripheral lung
11      reactivity in O3-exposed rabbits. The ACh-induced increase in lung resistance with O3 exposure
12      (140%) was two times higher than with air exposure, indicating that ambient-level O3 exposure
13      affects tracheal epithelial function in rabbits and increases central airway reactivity, possibly
14      through vagally-mediated mechanisms.
15          Pulmonary mechanics and hemodynamics were studied in the New Zealand white rabbit
16      isolated perfused lung model that allowed partitioning of the total pressure gradient into arterial,
17      pre- and post-capillary, and venous components ( Delaunois et al., 1998). Exposures to O3 (0.4
18      ppm for 4 h) were followed by evaluation of airway responsiveness to ACh, substance P (SP), or
19      histamine immediately or 48 h later. Ozone inhibited pulmonary mechanical reactivity to all
20      three bronchoconstrictors that persisted for 48 h and modified vasoreactivity of the vascular bed,
21      but only at 48 h PE. Arterial segmental pressure, normally insensitive to ACh and SP, was
22      significantly elevated by O3; precapillary segmental pressure decreased in response to Ach,
23      suggesting that O3 can induce direct vascular constriction, but the vascular responses are variable
24      and depend on the agonist used  and on the species studied.
25          Airway responsiveness to  the same three compounds was evaluated by Segura et al. (1997)
26      in guinea pigs exposed to O3 (0.15, 0.3, 0.6, or 1.2 ppm for 4 h).  Ozone did not cause AHR to
27      ACh or histamine, except at the highest concentration (1.2 ppm O3) for histamine. However,  O3
28      did cause AHR to SP at > 0.3 ppm, suggesting that O3 destroys neutral endopeptidases
29      (responsible for SP inactivation) in airway epithelial cells. Vargas et al. (1998), in a follow-up
30      study, demonstrated that guinea pigs chronically exposed to 0.3 ppm O3 for 4 h/day became
31      adapted to SP-induced AHR.  Ozone caused increased sensitivity to SP after 1, 3, 6, 12, and

        December 2005                           5-43        DRAFT-DO NOT QUOTE OR CITE

-------
 1      24 days of exposure that was associated with airway inflammation; however, after 48 days of
 2      exposure, the increased sensitivity to SP was lost.
 3           This study is in accordance with Szarek et al. (1995) who demonstrated that AHR
 4      associated with acute O3 exposures does not persist during long-term exposure to ambient-levels
 5      of O3 (<  1 ppm). Fischer 344 rats, exposed to 0.0, 0.12, 0.5, or 1.0 ppm O3, 6h/day, 5 days/week
 6      for 20 months, demonstrated significantly reduced responses to bethanechol, ACh, and electrical
 7      field stimulation in eighth generation airway segments. This suggests that some adaption had
 8      taken place during long-term exposure, possibly increased inner wall thickness.
 9           It is well known that the changes in breathing pattern and lung function caused by O3 are
10      attenuated with repeated daily exposures for at least 3 to 5 days. But guinea pigs exposed to
11      0.5 ppm  O3, 8 h/day for 7 days showed enhancement of responsiveness of rapidly adapting
12      airway receptors (load et al., 1998). Repeated exposure increased receptor activity to SP,
13      methacholine, and hyperinflation; there were no significant effects on baseline or SP- and
14      methacholine-induced changes in lung compliance and resistance, suggesting that the
15      responsiveness  of rapidly adapting receptors was enhanced.
16           Male and female Hartley guinea pigs exposed to O3 (0.1  and 0.3 ppm, 4 h/day,  4 days/week
17      for 24 weeks) were evaluated for airway responsiveness following ACh or OVA inhalation
18      challenges (Schlesinger et al., 2002a,b). Ozone exposure did not  cause AHR in nonsensitized
19      animals but did exacerbate AHR to both ACh and OVA in sensitized animals that persisted for
20      4 weeks  after exposure.  The effects of O3 on airway responsiveness were gender independent
21      and were concentration-related for the ACh challenges.
22           Schelegle et al. (2003a) evaluated  airway responsiveness in  infant rhesus monkeys exposed
23      to a 5 day O3 episode repeated every 14  days over a 6-month period.  Half of the monkeys were
24      sensitized to house dust mite allergen (HDMA; Dermatophagoid.es farinae) at 14 and 28 days of
25      age  before exposure to a total of 11 episodes of O3 (0.5 ppm, 8 h/day for 5 days followed by
26      9 days of FA), HDMA, or O3 + HDMA. Baseline Raw was significantly elevated after 10
27      exposure episodes in the HDMA +  O3 group compared to the FA, HDMA, and O3 exposure
28      groups.  Aerosol challenge with HDMA at the end of the 10th episode did not significantly affect
29      Raw, VT,  fB, or SaO2. Aerosol challenge with histamine was not significantly different after
30      6 episodes; however, the EC150 Raw for the HDMA + O3 group was significantly reduced after
31      10 episodes when compared to the FA, HDMA, and O3 exposure groups, indicating the

        December 2005                           5-44        DRAFT-DO NOT QUOTE OR CITE

-------
 1      development of AHR in this group sometime between episodes 6 and 10. The results are
 2      consistent with altered structural development of the conducting airways.
 3           During repeated episodic exposures to O3, respiratory responses are first altered to a rapid,
 4      shallow breathing pattern, which has long been considered protective, especially to the deep
 5      lung. This dogma has been discounted recently as discussed above (Schelegle et al., 2001).
 6      Alfaro et al. (2004) examined the site-specific deposition of 18O (1 ppm 2 h) at breathing
 7      frequencies of 80, 120, 160, or 200 breaths/minute. At all frequencies, parenchymal areas had a
 8      lower content of 18O than trachea and bronchi.  As breathing frequency increased from  80 to 160
 9      bpm, the deposition showed a reduction in midlevel trachea and an increase in both mainstream
10      bronchi.  At this  frequency there was also an increase in deposition in parenchyma supplied by
11      short (cranial) airway paths, consistent with results seen by Schelegle et al., (2001).  At 200 bpm
12      18O deposition in  trachea increased, concurrent with increases in right cranial and caudal bronchi
13      regions. Right cranial parenchymal content decreased at 200 bpm, whereas right caudal
14      parenchymal levels did not change at any breathing frequency. These two studies provide
15      evidence that O3-induced rapid, shallow breathing creates  a more evenly distributed injury
16      pattern, with possibly greater protection from focal injury  to the large conducting airways
17      including the trachea and the left mainstem bronchus.
18           Another study of the adaptive phenomena in SD rats used an exposure paradigm consisting
19      of 5 days of daily 8 h 1 ppm O3 exposures followed by 9 days of recovery in FA (Schelegle
20      et al., 2003b). This O3/FA pattern was repeated for 4 cycles and demonstrated that the  O3 -
21      induced rapid shallow breathing pattern was followed by adaptation occurred with each cycle.
22      But the release of SP from the trachea , the neutrophil content, and cell proliferation became
23      attenuated after the first cycle,  suggesting a disconnect from the rapid shallow breathing
24      response. Hypercellularity of the CAR epithelium and thickening of the CAR interstitium, not
25      linked to changes in cell proliferation, were also found.  The authors suggest mechanism(s) of
26      injury from repeated O3 exposures consisting of diminished neutrophilic inflammation/and or
27      release of mitogenic neuropeptides, depressed cell proliferative response, and cumulative distal
28      airway lesion.
29           Following the initial response of a rapid, shallow breathing pattern, animals eventually
30      adapt with continued episodic exposure despite the continued presence of epithelial damage,
31      altered structural  development, and inflammation of the airways. Chen et al. (2003) used a

        December 2005                            5-45       DRAFT-DO NOT QUOTE OR CITE

-------
 1      subset of the monkeys from the Schlegele et al. (2003a) study to demonstrate that attenuation of
 2      O3-induced rapid shallow breathing and lung function changes typically seen with repeated O3
 3      exposure may be caused by the adaptation of the respiratory motor responses. This episodic O3
 4      exposure appeared to create neuroplasticity of the nucleus tractus solitarius (NTS; a region of the
 5      brainstem which controls respiration), including increased nonspecific excitability of the NTS
 6      neurons, an increased input resistance, and an increased spiking response to intracellular
 7      injections of depolarizing current.
 8
 9      5.2.5.4  Summary and Conclusions - Effects on Airway Responsiveness
10           Ozone-induced AHR has been reported in a number of laboratory species at an exposure
11      range of 0.5 to 1.0 ppm  and in human asthmatics at ambient levels. In asthmatics, O3 is thought
12      to exacerbate AHR to specific allergens by nonspecifically increasing AHR.  New studies have
13      demonstrated that AHR in asthmatics is due in part to chronic inflammation and airway
14      remodeling.  Animal studies have shown that O3 exposure can augment OVA-induced AHR.
15      Importantly,  there is a temporal relationship between inflammatory cell influx and O3-induced
16      AHR, but inflammation is not a prerequisite of AHR.  Repeated O3 exposures enhance AHR,
17      possibly by modulating  rapidly adapting airway receptors or by altering the structure of
18      conducting airways.
19           Currently reported investigations on AHR with repeated O3 exposure to nonsensitized
20      laboratory animals have shown equivocal results, especially at the most relevant ambient O3
21      concentrations of < 0.3 ppm.  The few available studies in sensitized laboratory animals are
22      consistent with the O3-induced exacerbation AHR reported in atopic humans with asthma (see
23      Chapter 6) but the results are difficult to extrapolate because of interindividual and interspecies
24      differences in responsiveness to bronchoprovocation and possible adaptation of airway
25      responsiveness with long-term, repeated O3 exposures. Therefore, further studies in laboratory
26      animals are needed to investigate responses to the different  challenges in relation to
27      measurements of airway inflammation and the other physiological and structural factors known
28      to contribute to airway responsiveness in human subjects.
29           Important new information indicates that rapid shallow breathing in response to O3 causes
30      a more evenly distributed injury pattern rather than protects from injury.  New insights into the
31      mechanisms  of O3-induced AHR suggest that: (1) exercise-induced bronchoconstriction may be

        December 2005                           5-46        DRAFT-DO NOT QUOTE OR CITE

-------
 1      mediated by changes in tonicity of the bronchial smooth muscles; (2) vagally-mediated
 2      mechanisms may affect trachela epithelial function and increase central airway reactivity;
 3      (3) O3 may induce direct vascular constriction; (4) O3 may destroy neural endopeptidases in
 4      airway epithelial cells, thus preventing the inactivation of SP; and (5) repeated O3 exposures may
 5      diminish neutrophilic inflammation, depress cell proliferation, and cause cumulative distal
 6      airway lesions.
 7
 8      5.2.6 Genotoxicity Potential of Ozone
 9           There has been an historical interest in the ability of ground-level pollution to cause cancer,
10      especially lung cancer. This interest has been amplified in recent years by a report of increased
11      risks of incident lung  cancer that were associated with elevated long-term ambient
12      concentrations of O3, PM10, and SO2 in nonsmoking California males (Beeson et al., 1998;
13      Abbey et al., 1999). The nationwide American Cancer Society study (Pope et al., 2002) showed
14      no significant effect of O3 on mortality risk, but a positive association of July-September O3
15      concentrations and cardiopulmonary mortality.   Studies on children and young adults of
16      Southwest metropolitan Mexico City, repeatedly exposed to high levels of O3, PM, NOX,
17      aldehydes, metals, and other components in a complex ambient mixture, also report DNA
18      damage in blood leukocytes and nasal epithelial  cells (Valverde et al., 1997; Calderon-
19      Garciduenas et al., 1999), and abnormal nasal biopsies (Calderon-Garciduenas et al., 2001). (See
20      Chapter 6 for a discussion of the human  studies}
21           Many experimental studies have been conducted to explore the mutagenic and
22      carcinogenic potential of O3. In vitro studies are difficult to interpret due to the high exposure
23      levels and culture systems that allowed the potential formation of artifacts.  More recently
24      published in vivo exposure  studies found increased DNA strand breaks in respiratory cells from
25      guinea pigs (Ferng et  al., 1997) and mice (Bornholdt et al., 2002) but, again, only on exposures
26      to high doses of O3 (1 ppm for 72 h and 1 or 2 ppm for 90 min, respectively). Exposing the A/J
27      mouse strain (known to have a high  spontaneous incidence pulmonary  adenomas) to 0.12, 0.50,
28      and 1.0 ppm O3 for 6 h/day, 5 days/week for up to 9 months, Witschi et al. (1999) did not find O3
29      exposure-related differences in lung tumor multiplicity or incidence.
30           Similarly, in a sub-chronic exposure study (B6C3FJ mice to 0.5 ppm O3 for 6 h/day,
31      5 days/week for 12 weeks) Kim et al. (2001) did not find statistically significant increases in the

        December 2005                           5-47        DRAFT-DO NOT QUOTE OR CITE

-------
 1      incidence of lung tumors. Significant differences in mean body weight as well as mean absolute
 2      and relative weights of several organs (e.g., liver, spleen, kidney, testes, and ovary) were
 3      observed between O3-exposed and air-exposed mice. Histopathologic examination of major
 4      organs revealed oviductal carcinomas in 3/10 O3-exposed female mice.
 5
 6      5.2.6.1  Summary and Conclusions - Genotoxicity Potential of Ozone
 7           The weight of evidence from new experimental studies does not appear to support ambient
 8      O3 as a pulmonary carcinogen in laboratory animal models. These new data are in agreement
 9      with a definitive study of carcinogenicity of O3 from the NTP study (National Toxicology
10      Program,  1994; Boorman et al., 1994), which was negative in male and female rats, ambiguous
11      in male mice, and positive only in female mice at high concentrations of O3 (i.e., 1.0 ppm).
12      However, the new animal studies are not in agreement with epidemiologic studies discussed in
13      Chapter 7, which may suggest significant species differences in this health endpoint.
14
15
16      5.3 SYSTEMIC EFFECTS OF OZONE EXPOSURE
17           Ozone indirectly affects organs beyond the respiratory system due to O3 reaction products
18      entering the bloodstream and being transported to target sites. Extra-pulmonary effects could
19      also be due to the exposure-related production of mediators, metabolic products and cell
20      trafficking.  Although systemic effects are of interest and indicate a very broad array of
21      O3 effects, they are of limited influence and difficult to interpret.  By protecting from respiratory
22      tract effects, these  systemic effects will likely be protected against also.  Systemic effects are
23      only summarized briefly here and in Table AX5-8.
24
25      5.3.1  Neurobehavioral Effects
26           Animal behavior, both motor activity and operant behavior, has been shown to be
27      suppressed by acute 3 exposures of 0.12 ppm. There is a dose dependent decrease in activity
28      with increasing exposure levels. Additionally, these lowered activity levels tend to attenuate
29      with longer exposure periods.  New studies in adult laboratory animals confirm that
30      environmentally- relevant O3 concentrations from 0.2 to 1.0 ppm can decrease motor activity and
31      affect short- and long-term memory, as tested by passive avoidance conditioning in rats (Rivas-

        December 2005                            5-48        DRAFT-DO NOT QUOTE OR CITE

-------
 1      Arancibia et al., 1998; Avila-Costa et al., 1999; Dorado-Martinez et al., 2001), or water-maze
 2      learning tasks in mice (Sorace et al., 2001).  The effects have been attributed to reactive
 3      oxygen/nitrogen species and/or ozonation products. The memory deficits could be blocked by
 4      administration of vitamin E (Guerrero et al,  1999) or taurine (Rivas-Arancibia et al., 2000).
 5      Increased freezing and decreased exploratory behaviors were accompanied by decreased
 6      serotonin levels and increased levels of NO, glutamate, dopamine and striatal lipoperoxidation in
 7      rats exposed to 1 ppm of O3 for 4 h (Rivas-Arancibia et al., 2003). The O3-exposed animals also
 8      demonstrated  neuronal cytoplasm and dendrite vacuolation and dilation of RER cisterns, which
 9      the authors interpret as a neurodegenerative process resulting from the oxidative stress of acute
10      O3 exposure.  Nino-Cabrera et al. (2002) demonstrated that a 0.7 ppm O3 exposure for 4 h can
11      induce ultrastructural alterations in the hippocampus and prefrontal cortex in aged rats.  These
12      are areas of the brain where degenerative age-related changes in learning and memory functions
13      have been reported (Bimonte et al., 2003).
14           Paz (1997) reviewed a series of studies that demonstrated significant alterations of
15      electroencephalographic (EEG) patterns during sleep in animals acutely exposed to O3 (0.35 to
16      1.0 ppm). Rats and cats both showed loss of paradoxical sleep time after 2 to 8 h of O3  exposure
17      (Paz and Bazan-Perkins, 1992; Paz and Huitron-Resendiz, 1996). Increased total wakefulness,
18      alterations in circadian rhythm, and a permanent 50% loss of paradoxical sleep time were shown
19      in rat pups born to dams exposed to 1.0 ppm O3 during gestation (Haro and Paz, 1993).  Effects
20      on sleep patterns were associated with alterations in brain neurotransmitter levels (Huitron-
21      Resendiz et al., 1994; Gonzalez-Pina and Paz, 1997) thought to be caused by O3 reaction
22      products or  prostaglandins (Koyama and Hayaishi, 1994).  The permanent effects in pups caused
23      by high O3 exposure during gestation were attributed to the diminished antioxidant capability  of
24      fetal tissue (Gunther et al., 1993).
25
26      5.3.2  Neuroendocrine  Effects
27           Early  studies suggested an interaction  of O3 with the pituitary-thyroid-adrenal axis because
28      thyroidectomy, hypophysectomy, and adrenalectomy protected against the lethal effects of high
29      concentrations of O3 Concentrations of 0.7 to 1.0 ppm  O3 caused morphological changes in the
30      parathyroid; thymic atrophy; decreased serum levels of thyroid stimulating hormone,
31      triiodothyronine (T3), thyroxine (T4), free T4, and protein binding; and increased prolactin.

        December 2005                            5-49        DRAFT-DO NOT QUOTE OR CITE

-------
 1      In more recent studies, increased toxicity to O3 was reported in hyperthyroid rats by Huffman
 2      et al. (2001) and T3 supplementation was shown to increase metabolic rate and pulmonary injury
 3      in the lungs of O3-treated animals (Sen et al., 1993).
 4           The mechanisms by which O3 affects neuroendocrine function are not well understood.
 5      Cottet-Emard et al. (1997) examined catecholamine activity in rat sympathetic efferents and
 6      brain areas of prime importance to adaptation to environmental stressors.  Exposures of 0.5  ppm
 7      O3 for 5 days caused inhibition of norepinephrine turnover in heart (-48% of the control level)
 8      but not in lungs and failed to modify the tyrosine hydroxylase activity in superior cervical
 9      ganglia, and the catecholamine content in the adrenal glands.  In the CNS, O3 inhibited tyrosine
10      hydroxylase activity in noradrenergic brainstem cell groups and decreased catecholamine
11      turnover was in the cortex (- 49%) and striatum (-18%) but not in the hypothalamus.  This
12      suggests that high ambient levels of O3 can produce marked neural disturbances in structures
13      involved in the integration of chemosensory inputs, arousal, and motor control, effects that may
14      be responsible for some of the behavioral effects seen with O3 exposure.
15           High, non-ambient levels of O3 (e.g., > 1.0 ppm) affect visual and olfactory neural
16      pathways in the rat. For example, Custodio-Ramierez and Paz (1997) reported a significant
17      delay in visual evoked potentials recorded in the visual cortex and the lateral geniculate nucleus
18      of male Wistar rats acutely exposed to high levels of O3 (1.5,  and 3.0 ppm for 4 h). Colin-
19      Barenque et al. (1999), using the same strain, reported cytological and ultrastructural changes in
20      the granule layer of the olfactory bulb after a 4-h exposure to 1 to 1.5 ppm O3. Although these
21      neural effects are thought to be caused by O3 reaction products, especially free radicals, the
22      studies do not add much to an understanding of the underlying mechanisms.
23
24      5.3.3  Cardiovascular Effects
25           Studies of the effects on hematological parameters and blood chemistry have shown that
26      erythrocytes are a target of O3. Exposures to 1.0  ppm O3 for 3 h have been found to decrease
27      HR, MAP, and core temperature and to induce arrhythmias in some exposures. These effects are
28      more pronounced in adult and awake rats than in younger or sleeping animals. Exposures of
29      0.2 ppm for 48 h have been shown to cause bradycardia, while exposures of 0.1 ppm for 24 have
30      been shown to cause bradyarrhythmia in rats only.
        December 2005                            5-50        DRAFT-DO NOT QUOTE OR CITE

-------
 1           A more recent study of rats exposed to FA for 6 h, followed 2 days later by a 5-hr exposure
 2      to 0.1 ppm O3, 5 days later by a 5-hr exposure to 0.3 ppm O3, and 10 days later by a 5-hr
 3      exposure to 0.5 ppm O3 used the the head-out plethysmograph for continuous measurements
 4      (Arito et al., 1997). Each of the O3 exposures was preceded by a 1-hr exposure to FA. Transient
 5      rapid shallow breathing with slightly increased HR appeared 1-2 min after the start of O3
 6      exposures and was attributed to an olfactory response.  Persistent rapid shallow breathing with a
 7      progressive decrease in HR occurred with a latent period of 1-2 hr.  During the last 90-min of
 8      exposure, averaged values for relative minute ventilation tended to decrease with the increase in
 9      O3 concentration for young (4-6 mo) but not old (20-22 mo) rats.
10           New studies utilizing radiotelemetry transmitters  in unanesthetized and unrestrained rats,
11      Watkinson et al. (1995; 2001) and Highfill and Watkinson (1996) demonstrated that when HR
12      was reduced during O3 exposure, the Tco and activity levels also decreased. The decreases in Tco
13      and blood pressure reported by in these studies and by Arito  et al., (1997) suggest that the
14      changes in ventilation and HR are mediated through physiological and behavioral defense
15      mechanisms in an attempt to minimize the irritant effects of O3 inhalation. Decreased activity
16      was previously reported in laboratory animals during exposure to O3 (see above).
17           Similar cardiovascular and thermoregulatory responses in rats to O3 were reported by
18      Iwasaki et al. (1998).  Repeated exposure to 0.1, 0.3, and 0.5 ppm O3 8 hrs/day for 4 consecutive
19      days caused disruption of circadian rhythms of HR and Tco on the first and second exposure days
20      that was concentration-dependent. The decreased HR and Tco recovered to control values on the
21      third and fourth days of O3 exposure.
22           The thermoregulatory response to O3 was further  characterized by Watkinson et al. (2003).
23      Male Fischer-344 rats were exposed to 0.0 ppm* 24 h/day (air), 0.5 ppm x 6h/day (intermittent)
24      or 0.5 ppm x 23  h/day (continuous) at 3 temperatures, 10°  C  (cold), 22° C (room), or 34°  C
25      (warm).  Another protocol examined the effects of O3 exposure  (0.5 ppm) and exercise described
26      as rest, moderate, heavy or CO2-stimulated ventilation.  Both intermittent and continuous  O3
27      exposure caused decreases in HR and Tco and increases in BALF inflammatory markers.
28      Exercise in FA caused increases in HR and Tco while exercise in O3 caused decreases in those
29      parameters. Carbon dioxide and O3 induced the  greatest deficits in HR and Tco. Several factors
30      were suggested that may  modulate the hypothermic response, including dose, animal mass, and
31      environmental stress).

        December 2005                             5-51        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Laboratory animals exposed to relatively high ambient O3 concentrations (> 0.5 ppm)
 2      demonstrate tissue edema in the heart and lungs. This may be due to increased circulating levels
 3      of atrial natriuretic factor (ANF), which is known to mediate capillary permeability,
 4      vasodilation, and blood pressure (Daly et al., 2002). Increased levels of ANF were reported in
 5      the heart, lungs, and circulation of rats exposed to 0.5 ppm O3 for 8 h (Vesely et al.,  1994a,b,c)
 6
 7      5.3.4 Reproductive and Developmental Effects
 8           Early studies of pre- and postnatal exposure to O3 were performed at relatively  high
 9      concentrations. Teratogenic effects were not observed with intermittent exposures of 0.44 to
10      1.97 ppm O3 during any part of gestation.  Continuous exposure during mid-gestation increased
11      the resorption of embryos while exposures during late gestation delayed some behavioral
12      developments (e.g., righting, eye opening). There were no effects on neonatal mortality up to
13      1.5 ppm  O3, whereas some transient effects on weight gain were observed at exposures of
14      0.6 ppm  O3.
15           Recent studies tend to confirm previous conclusions that prenatal exposures to  O3
16      concentrations < 1.0 ppm do not cause major or widespread somatic or neurobehavioral effects
17      in the offspring of laboratory animals.  These  studies generally add some weight toward a
18      negative interpretation of the importance of contributions of low, ambient O3 to lower birth
19      weights and gross development defects reported in neonates born to women exposed to typical
20      ambient pollution (e.g., Renner, 2002;  Chen et al., 2002;  Ritz and Yu, 1999).  Some postnatal O3
21      exposure studies continue to find a few, subtle or borderline somatic and behavioral deficits that
22      will require further research to better assess potential  risk to developing humans.
23           Recent studies of somatic and neurobehavioral development in female CD-I mice exposed
24      during pregnancy (days 7 to 17) to O3 (0, 0.4,  0.8, or 1.2  ppm) failed to show any O3  effects on
25      reproductive or behavioral performance (Bignami et al.,  1994). The study did find significant
26      decreases in body weight gain and delayed eye opening in pups in the 1.2 ppm exposure group.
27      The lack of effect on behavioral performance  contrasts with earlier findings, which may be due
28      to the use of different species, differing exposure durations, cross-fostering used in the  latter
29      study different species and exposure durations during pregnancy.  A second study using CD-I
30      mice  exposed in utero from conception through day 17 of pregnancy to 0, 0.2, 0.4, and  0.6 ppm
31      O3 found no significant deficits in reproductive performance,  postnatal somatic and

        December 2005                           5-52         DRAFT-DO NOT QUOTE OR CITE

-------
 1      neurobehavioral development, or adult motor activity (Petruzzi et al., 1995).  A third study by
 2      the same group (Petruzzi et al., 1999), using O3 exposures (0.3, 0.6, or 0.9 ppm) which continued
 3      postnatally until weaning, showed subtle changes in handedness and morphine reactivity.
 4      Exposures to 0.6 ppm O3 caused a reduced preference for the right paw in adulthood.  Exposures
 5      to 0.9 ppm O3 altered hot plate avoidance after IP treatment with morphine in adulthood.
 6          CD-I mice exposed to 0.6 ppm O3 from birth through weaning demonstrated no
 7      impairment of navigational performance during acquisition and only subtle changes during
 8      reversal (Dell'Omo et al., 1995a). Additionally, there were no O3-induced effects on
 9      reproductive performance, but offspring showed a significant reduction in body weight.  Effects
10      on neurobehavioral development with this exposure were minor, with some attenuation of
11      activity responses and impairment of passive avoidance acquisition (DeH'Omo et al. (1995b).
12      The offspring of CD-I mice continuously exposed from 30 days prior to the formation of
13      breeding pairs until PND 17 to 0.0, 0.3, or 0.6 ppm O3 showed only small and selective effects
14      on somatic and sensorimotor development (Sorace et al., 2001).
15          Morphological changes were found in the anterior cerebellar lobe of rat pups born to dams
16      exposed during the entire gestation period to very high (1.0 ppm) O3 concentrations for 12 h/day.
17      (Rivas-Manzano and Paz, 1999).  Additionally, the dams displayed significantly fewer
18      implantations, increased rate of reabsorptions, a high incidence of spontaneous abortion, and
19      offspring with low birth weight, as noted by previous investigators.
20
21      5.3.5   Effects on the Liver, Spleen, and Thymus
22          Early investigations of the effects of O3 on liver centered on xenobiotic metabolism, and
23      the prolongation of sleeping time, which was observed at 0.1  ppm O3.  In some species, only
24      adults and especially females were affected.  In rats, high (1.0 to 2.0 ppm) acute O3 exposures
25      caused increased production of NO by hepatocytes and enhanced protein synthesis (Laskin et al.,
26      1994; 1996). The O3-associated effects shown in the liver are thought to be mediated by
27      inflammatory cytokines or other cytotoxic mediators released by activated macrophages in the
28      lungs (Vincent et al., 1996; Laskin et al., 1998; Laskin and Laskin, 2001). Except for the earlier
29      work on xenobiotic metabolism, the responses occurred only  after very high acute O3 exposures.
30          Examinations of the effects of O3 on spleen and thymus have shown that O3 primarily
31      affects T-cell mediated systemic immunity. As with the O3-associated effects shown in the liver,

        December 2005                            5-53        DRAFT-DO NOT QUOTE OR CITE

-------
 1      most of the statistically significant changes occurred after acute exposures to very high O3
 2      concentrations and relate to systemic oxidative stress.  Using more relevant ambient urban O3
 3      exposure patterns, effects were not found on systemic immune function of rats.
 4
 5      5.3.6  Effects on Cutaneous and Ocular Tissues
 6          Ozone exposure not only affects various organ systems, when inhaled, but also has direct
 7      effects on the exposed skin and eyes.  The outermost layer of the skin (stratum corneum; SC)
 8      may be oxidized,  which can lead to compromise of the skin barrier and an epidermal
 9      proinflammatory  response ( Weber et al., 2001; Cotovio et al., 2001;  Thiele, 2001). These
10      effects are found  only at very high concentrations (>l-5 ppm) and have not been shown at more
11      relevant ambient levels of exposure. The skin possesses a well-developed defense system
12      against oxidative  stress, utilizing nonenzymatic (e.g., vitamin C and E, glutathione, uric acid,
13      a-tocopherol) and enzymatic (e.g., superoxide dismutase,  catalase, glutathione reductase and
14      peroxidase) antioxidants (Cross et al., 1998).  Ocular tissues have similar antioxidant protective
15      function as the skin but are not as well studied (Mucke, 1996; Rose et al., 1998). Effects of
16      ground-level smog on the eyes have been reported but generally are attributed to related
17      photochemical oxidants like peroxyacetyl nitrate (Vyskocil et al.,  1998) or possibly to
18      atmospheric O3 precursors or reaction products like aldehydes.
19          Hairless mice (SKH-1) were used to evaluate the cutaneous effects of O3 (1, 5,  and 10 ppm
20      for 2 h or 1 ppm for 2 h on six consecutive days) by Thiele et al. (1997a,b,c,d). In the upper
21      epidermis decreased antioxidant levels were observed at > 1.0 ppm O3, while in both upper and
22      lower epidermal layers increased malondialdehyde (MDA) was found with exposures > 5 ppm.
23      Dermal application of vitamin E prevented MDA accumulation.
24          The same mouse model exposed to O3 (0.8 to 10 ppm for 2 h) was used to demonstrate that
25      O3 depletes the low molecular weight antioxidants (e.g., a-tocopherol, vitamin C, glutathione,
26      uric acid) in the SC at > 1.0 ppm and causes increased MDA at > 5 ppm (Weber et al, 1999,
27      2000,2001). Valacchi et al. (2000) demonstrated that preexposure to O3 followed by low-dose
28      ultraviolet (UV) radiation (0.33 MED) caused depletion of a-tocopherol at an exposure level of
29      0.5 ppm O3. This suggests that combined low doses of UV radiation  and near-ambient levels of
30      O3 may cause oxidative stress on the SC. Additional studies demonstrated that stress-inducible
31      proteins (e.g., heme oxygenase-1) and other heat shock proteins (e.g., HSP27 and HSP70) were

        December 2005                            5-54        DRAFT-DO NOT QUOTE OR CITE

-------
 1      induced in deeper cellular layers of the epidermis following O3 exposure (8.0 ppm for 2 h)
 2      (Valacchi et al., 2002). Prolonged exposure to lower concentrations of O3 (0.8 ppm) for 6 h also
 3      induces cellular stress responses that included the formation of HNE protein adducts, HSP27,
 4      and heme-oxygenase-1 in the deeper cellular layers of the skin that continued for up to 18 h after
 5      O3 exposure, followed by repair processes (Valacchi et al., 2003).
 6           The importance of O3 and UV-induced cellular protein oxidation found in murine skin
 7      models to possibly similar environmentally-induced changes in human SC keratins was
 8      identified by Thiele et al. (1998, 1999) and Thiele (2001). Using the presence of carbonyl
 9      groups in proteins as a marker of reactive oxygen mediated protein oxidation, they reported
10      higher carbonyl levels in the upper SC from the tanned skin of humans and in the skin of healthy
11      human volunteers exposed to model chemical oxidants (e.g., hypochlorite, benzoyl peroxide)
12      that were inversely correlated with vitamin E levels. The environmentally-induced oxidative
13      damage identified in human SC represents an early pathophysiological stage in the development
14      of barrier disruption and inflammation, and possibly has implications for the process of
15      desquamation. The relevance of potentiation of environmental oxidative stress by O3 exposure
16      of human skin needs further study.
17
18      5.3.7  Summary and Conclusions - Systemic Effects  of Ozone
19           Neurobehavioral effects of O3 at concentrations of 0.2 to 1.0 ppm include decreased motor
20      activity, short- and long-term memory deficits, increased freezing behavior, decreased
21      exploratory behaviors. These effects have been associated with reactive oxygen/nitrogen
22      species, ozonation products, altered neurotransmitter levels, morphological changes in several
23      brain regions, and altered EEG patterns during sleep. Neuroendocrine effects of O3 include
24      morphological and hormonal changes in the pituitary-thyroid-adrenal axis at concentrations of
25      -0.75 ppm and alterations of visual and olfactory neural pathways at concentrations > 1 ppm.
26      Mechanisms underlying these effects are not understood at this time.  Cardiovascular effects of
27      O3 at concentrations of 0.3 to 0.5 ppm include decreased HR, Tco, and BP, which have been
28      termed a hypothermic response. Concentrations of O3 >0.5 ppm cause tissue edema (possibly
29      mediated by ANF).  These data are in accordance with O3-associated cardiac defects found in
30      neonates and fetuses delivered in southern California during 1987 to 1993,  suggesting that high
        December 2005                            5-55        DRAFT-DO NOT QUOTE OR CITE

-------
 1     urban exposures of O3 and its related co-pollutants can adversely affect the cardiovascular
 2     system.
 3          Prenatal exposures to O3 concentrations < 1.0 ppm did not cause noticeable somatic or
 4     neurobehavioral effects in offspring, while concentrations of 1.0 to 1.5 ppm caused varying
 5     effects on neonatal mortality.  Some studies have shown an effect of O3 on liver xenobiotic
 6     enzymes at concentrations as low as 0.1 ppm, while other studies have shown no  alterations in
 7     metabolic enzymes at even 1 ppm, with the effects appearing to be highly-species specific.
 8     Effects on spleen and thymus appear to only occur at high O3 concentrations (> 1.0 ppm), while
 9     relevant ambient, urban exposures have no effect on systemic immune function in rats. Effects
10     of O3 on cutaneous and ocular tissue are only seen at high, non-relevant concentrations.
11
12
13     5.4   INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
14            POLLUTANTS
15          Animal studies of the effects of O3 in combination with other air pollutants  show that
16     antagonism, addititivity, and synergism can result, depending on the animal species, exposure
17     regimen, and health endpoint.  These studies are of three types, ambient air mixtures, laboratory -
18     generated complex mixtures, and binary mixtures. Tables AX5-9 and AX5-10 list binary studies
19     of coexposures to nitrogen dioxide and PM, respectively.
20
21     5.4.1  Ozone and Nitrogen Oxides
22          The most commonly studied copollutant in binary mixtures with O3 is NO2. Both early
23     studies and more recent ones indicate that, although interaction may occur between these two
24     pollutants, in general, O3 often masked the effects of the NO2 or accounted for most of the
25     response,  due to the greater toxicity of O3. Very generally, additivity occurred after acute
26     exposure and synergism occurred with prolonged exposure. Interpreting the mixture studies is
27     challenging because laboratory exposure patterns rarely simulate real-world exposure patterns.
28     In the case of NO2 and O3, NO2 typically peaks before O3, with a mixture occurring between the
29     peaks,  but most laboratory exposures used mixtures only. Also, most studies of O3 and NO2
30     mixtures used ambient levels of O3 and levels of NO2 high above ambient.
       December 2005                            5-56        DRAFT-DO NOT QUOTE OR CITE

-------
 1          Recent work has shown that chronic exposures of rats to O3 (0.8 ppm) and NO2 (14.4 ppm)
 2     for 6 h/day caused development of respiratory insufficiency and severe weight loss. Half of
 3     these animals died after 55 to 78 days of exposure due to severe fibrosis (Farman et al., 1997).
 4     Increased total lung collagen and elastin was observed, with loss of mature collagen, suggesting
 5     breakdown and remodeling of the lung parenchyma. Morphological examination of these
 6     coexposures demonstrate a sequence of events starting with increasing inflammatory and mild
 7     fibrotic changes for the first 3 weeks of exposure and stabilized or even reduced changes after
 8     4 to 6 weeks, and severe increases over 7 to 9 weeks of exposure (Farman et al., 1999). This
 9     suggests a repair processes occurring during the middle 4 to 6 weeks of exposure become
10     overwhelmed, leading to progressive fibrosis after 7 to 8 weeks of exposure. When the
11     coexposure was extended for 90 days, lesions were shown to extended far into the acinus, but the
12     extent of tissue  involvement was the same after 7, 78, and 90 days of exposure. At the end of
13     exposure, high levels of procollagen types I and III mRNA were observed within central acini in
14     the lungs from the combined exposure group but not in lungs from the rats exposed to O3 or NO2
15     alone.
16          Sprague-Dawley rats exposed  to 0.3 ppm O3 and the combined exposure of O3 and 1.2 ppm
17     NO2 for 3 d demonstrated significant DNA single-strand breaks in AMs ( Bermudez et al., 1999).
18     No changes were caused by NO2-only exposure. The same  exposures stimulated  the activity of
19     polyADPR synthetase, suggesting a response to lung cellular DNA repair caused  by oxidant-
20     induced lung injury (Bermudez, 2001).  The laboratory animal model  of progressive pulmonary
21     fibrosis, utilizing long-term, combined O3 (0.4 to 0.8 ppm) and high-level NO2 (7 to 14 ppm)
22     exposure, causes an initial acute pulmonary inflammation, followed by adaptation and repair,
23     and eventually causing pulmonary fibrosis after 6 to 13 weeks of exposure (Ishii et al., 2000b;
24     Weller et al., 2000). Unfortunately, this model is not very useful for understanding potential
25     interactive effects of ambient concentrations of O3 and NO2.
26
27     5.4.2   Ozone and Other Copollutants
28     Ozone and Formaldehyde
29          Early studies with combined exposures to O3 and formaldehyde  (HCHO) found evidence of
30     both synergistic and non-interactive effects. New work includes studies of biochemical and
31     histopathological endpoints in rats exposed to 0.4 ppm O3 and 3.6 ppm HCHO, alone and

       December 2005                           5-57        DRAFT-DO NOT QUOTE OR CITE

-------
 1      combined, for 8 h/day for 3 days (Cassee and Feron, 1994). They demonstrated no interactive
 2      effects in the nasal respiratory epithelium, despite the high levels of HCHO when compared to
 3      typical ambient levels of 1 to 10 ppb (e.g., Rehle et al., 2001). Mautz (2003) studied changes in
 4      breathing pattern and epithelial cell proliferation using exposures of 0.6 ppm O3 and 10 ppm
 5      HCHO alone and in combination for 3 h with exercise at two times resting ventilation. Even
 6      with exercise, HCHO does not substantially penetrate to the lower respiratory tract to interact
 7      with O3, and does not alter breathing patterns to modify local O3 dose.  Parenchymal injury was,
 8      therefore, due to O3 alone. In the nasal transitional epithelium and in the trachea, however,
 9      combined exposure produced additive effects due to the increased volume of toxicants during
10      exercise.  No other combined pollutant studies have been published in the peer-reviewed
11      literature, although two studies compared the respiratory effects of O3 to HCHO. Nielson et al.,
12      (1999) compared upper airway sensory irritation caused by HCHO concentrations up to 4 ppm to
13      the lower airway irritation caused by O3. Using BALB/c mice, they continuously measured fB,
14      Vt, expiratory flow, T;, Te, and respiratory patterns during acute, 30-min exposures. The NOEL
15      for HCHO was 0.3 ppm, compared to 1.0 ppm for O3.
16           Thus, O3 and HCHO do not appear to have additive effects, except during exercise, and
17      that is due to increased volume of gas reaching the tissue.  Any possible synergism occurs in the
18      nasal epithelium.  HCHO exerts its effects primarily in the upper respiratory tract, whereas the
19      primary site of acute cell injury from O3 occurs in the conducting airways. EPA is currently
20      completing a toxicological and epidemiological review and risk characterization for
21      formaldehyde.
22
23      Ozone and Tobacco Smoke
24           Early studies of combined exposures of O3 (Ippm) and tobacco smoke demonstrated
25      altered airway responsiveness to inhaled bronchoconstrictor challenge and tracheal vascular
26      permeability in guinea pigs.  A more recent study (Wu et al., 1997) reported that inhalation of
27      cigarette smoke evokes a transient bronchoconstrictive effect in anesthetized guinea pigs. Total
28      pulmonary resistance (RL) and dynamic lung compliance (Cdyn) were compared before and after
29      acute exposure to 1.5 ppm O3 for 1 h.  Cigarette smoke alone (7 ml) at a low concentration (33%)
30      induced a mild and reproducible bronchoconstriction that slowly developed and reached its peak
31      after a delay of > 1 min. After O3, the same cigarette smoke inhalation  challenge evoked an

        December 2005                           5-58        DRAFT-DO NOT QUOTE OR CITE

-------
 1      intense bronchoconstriction that occurred more rapidly, reaching its peak within 20 s, and was
 2      sustained for > 2 min. Pretreatment with selective antagonists of neurokinin type 1 and 2
 3      receptors completely blocked the enhanced airway responsiveness suggesting that O3 exposure
 4      induced AHR to inhaled cigarette smoke, which resulted primarily from the bronchoconstrictive
 5      effect of endogenous tachykinins.
 6          The above studies were conducted with undiluted tobacco smoke and high O3
 7      concentrations. To determine the effects of aged and diluted sidestream cigarette smoke (ADSS)
 8      as a surrogate of environmental tobacco smoke (ETS) on O3-induced lung injury, Yu et al.
 9      (2002) exposed male B6C3F1 mice to (1) FA, (2) ADSS, (3) O3, or (4) ADSS followed by O3
10      (ADSS/O3). Exposure to 30 mg/m3 ADSS, 6 h/day for 3 days, followed by exposure to 0.5 ppm
11      O3 for 24 h was associated with a significant increase in the number of cells recovered by B AL
12      compared with exposure to ADSS alone or O3 alone. Neutrophils, lymphocytes, and total
13      protein levels in BAL were increased following the combined exposure when compared with all
14      other groups. Within the CAR, the percentage of proliferating cells was unchanged from control
15      following exposure to ADSS alone but was significantly elevated following exposure to O3 and
16      further augmented in a statistically significant manner in mice exposed to ADSS/O3. Following
17      exposure to O3 alone or ADSS/O3, the ability of AMs to release IL-6 under LPS stimulation was
18      significantly decreased, while exposure to ADSS alone or ADSS/O3 caused a significantly
19      increased release of TNFa from AMs under LPS stimulation. These data suggest that ADSS
20      exposure enhances the sensitivity of animals to O3-induced lung injury.
21          Acute exposure to ETS also may make a healthy person more susceptible to sequential O3
22      exposure by affecting lung barrier function or the underlying epithelium. Toxicological studies
23      with components of ETS (e.g., nicotine receptor agonists, acrolein, and oxidants) have shown
24      that the vagal bronchopulmonary C-fibers are stimulated by acute exposures that initiate both
25      central and local responses (Bonham et al., 2001; Mutoh et al.,  2000).  The central responses
26      (e.g., tachypnea, cough, bronchoconstriction, increased mucous secretion) are more protective of
27      the lungs; however,  local responses may include increased sensitization of the C-fibers to other
28      irritants,  including O3. Active tobacco smokers should not be similarly affected because they
29      already have significant chronic airway inflammation and increased mucus production. In fact,
30      chronic smokers appear to have diminished lung function responses to O3 (see Chapter 6).
31

        December 2005                            5-59       DRAFT-DO NOT QUOTE OR CITE

-------
 1      5.4.3  Complex (Multicomponent) Mixtures Containing Ozone
 2           Studies using complex (multicomponent) mixtures containing O3 are difficult to interpret
 3      because of chemical interactions between the components, as well as the resultant production of
 4      variable amounts of numerous secondary reaction products, and a lack of precise control over the
 5      ultimate composition of the exposure environment. Irradiated automobile exhaust mixtures
 6      containing total oxdant concentrations (expressed as O3) in the range of 0.2 to 1.0 ppm have been
 7      shown to cause pulmonary function changes in several species.
 8           A more recent attempt has been made to examine multicomponent mixtures resulting from
 9      the reaction of O3 with unsaturated hydrocarbons [e.g., isoprene (C5H8) and terpene (C10H16)],
10      producing HCHO, formic acid, acetone, acrolein, acetic acid, and other oxidation products, many
11      of which are strong airway irritants.  Wilkins et al. (2001) evaluated sensory irritation by
12      measuring mean fB in the mouse bioassay and found a 50% reduction after 30 min of exposure to
13      reaction products of O3 and isoprene. The mixture at this time period contained < 0.2 ppm O3, so
14      the authors attributed the observed effects to the oxidation products. Clausen et al. (2001), using
15      the same mouse model, evaluated the reaction products of O3 and limonene. A 33% reduction in
16      mean fB was produced after 30 min of exposure to the mixture containing < 0.3 ppm O3, again
17      implicating the effects of strong irritant products. Further work needs to be done with these
18      complex reaction mixtures because of their potential impact on the respiratory tract.  The results
19      would be particularly important, however, to the reaction of O3 indoors (see Chapter 3).
20           Pollutant mixtures containing acid aerosols comprise another type of commonly examined
21      exposure atmosphere (studies summarized in Table AX5-10). Earlier studies that employed
22      simultaneous single, repeated, or continuous exposures of various animal species to mixtures of
23      acid sulfates and O3 found responses for several endpoints, including tracheobronchial
24      mucociliary clearance, alveolar clearance, pulmonary mechanics, and lung morphology, to be
25      due solely to O3 Some synergism was noted for bacterial infectivity, response to antigen, and
26      effects on lung protein content and the rate of collagen synthesis.
27           More recent studies found some differences in airway responses to inhaled acid particle-O3
28      mixtures that may have been partly due to airway dosimetry.  Various physical and chemical
29      mechanisms may be responsible (see Schlesinger, 1995).  For example, physical adsorption or
30      absorption of O3 or its reaction products on a particle could result in transport to more sensitive
31      sites, or to sites where O3, by itself, would not normally be reactive. Chemical reactions on the

        December 2005                            5-60        DRAFT-DO NOT QUOTE OR CITE

-------
 1      surface of particles can form secondary products that are more lexicologically active, or
 2      chemical characteristics of the particle may change the residence time or reactivity of oxidation
 3      products at the site of deposition.  The hypothesis that synergism between O3 and sulfates is due
 4      to decreased pH changing the residence time or reactivity of reactants, such as free radicals, was
 5      tested by Chen et al.  (1995) and El-Fawal et al. (1995). Male New Zealand white rabbits were
 6      exposed for 3  h to 125 |_ig/m3 H2SO4, 0.1, 0.3, or 0.6 ppm O3, and to combinations.  Chen et al.
 7      (1995) demonstrated that decreased pH following exposure to acid aerosol was correlated with
 8      phagocytic activity and capacity of harvested macrophages and that exposure to O3/ H2SO4 the
 9      removed this relationship. El-Fawal et al. (1995) showed that responsiveness of rabbit harvested
10      bronchial rings to ACh was increased following O3 exposure, but that O3/ H2SO4 combinations
11      resulted in antagonism.
12           Using rat tracheal explant cultures, Churg et al. (1996) demonstrated increased uptake of
13      asbestos or TiO2 in response to 10 min O3 (up to 1.0 ppm) pre-exposure suggesting that low
14      concentrations of O3 may increase the penetration of some types of PM into epithelial cells.
15      Using human  epithelial cell cultures, Madden et al. (2000) demonstrated a greater potency for
16      ozonized diesel PM to induce prostaglandin E2 production.  This suggests  that O3 can modify the
17      biological activity of PM derived from diesel exhaust.
18           Effects of combined exposures of O3 and resuspended urban  particles on cell proliferation
19      in epithelial cells of the terminal bronchioles and the alveolar ducts were examined by Vincent
20      et al. (1997) and Adamson et al. (1999).  Rats exposed to 0.8 ppm  O3 in combination with 5 or
21      50 mg/m3 particles for 4 h demonstrated greatly potentiated proliferative effects compared to O3
22      exposure alone.  These findings using resuspended dusts, although at high concentrations, are
23      consistent with the studies demonstrating interaction between H2SO4 aerosols and O3. Effects of
24      acute coexposure to O3 and fine or ultrafine H2SO4 aerosols on lung morphology were examined
25      by Kimmel  et al. (1997). They demonstrated that alveolar septal volume was increased in
26      animals co-exposed to O3 and ultrafine, but not fine, H2SO4.  Interestingly, cell proliferation was
27      increased only in animals co-exposed to fine H2SO4 and O3, as compared to animals exposed to
28      O3 alone. Subchronic exposure to acid aerosols (20 to 150 |ig/m3 H2SO4) had no interactive
29      effect on the biochemical and morphometric changes produced by  either intermittent or
30      continuous exposure to 0.12 to 0.2 ppm O3, which suggests that the interactive effects of O3 and
31      acid aerosol coexposure in the lung disappeared during the long-term exposure (Last and

        December 2005                             5-61        DRAFT-DO NOT QUOTE OR CITE

-------
 1      Pinkerton, 1997).  Sindhu et al. (1998) observed an increase in rat lung putrescine levels after
 2      repeated, combined exposures to O3 and a nitric acid vapor.
 3           Other studies have examined interactions between carbon particles and O3. The
 4      interactions of intratracheally instilled carbon particles by followed by either a 7-day or 60-day
 5      exposure to 0.5 ppm O3 in rats was evaulated by Creutzenberg et al. (1995).  The carbon
 6      particles caused diminished phagocytotic capacity and chemotactic migration capability of AMs
 7      that was stimulated by the subsequent O3 exposure. Inflammatory responses following
 8      coexposure of O3 plus fine, H2SO4-coated, carbon particles (MMAD = 0.26 jim) for 1 or 5 days
 9      was examined by Kleinman et al. (1999).  The response with the O3-particle mixture was greater
10      after 5 days (4 h/day) than after Day 1.  This contrasted with O3  exposure alone (0.4 ppm), which
11      caused marked inflammation on acute exposure, but no inflammation after 5 consecutive days of
12      exposure.
13           The effects of a mixture of elemental carbon particles, O3,  and ammonium bisulfate on rat
14      lung collagen content and macrophage activity was examined by Kleinman et al. (2000).
15      Decreases in lung collagen, and increases in macrophage respiratory burst and phagocytosis
16      were observed relative to other pollutant combinations. Mautz et al. (2001) used a similar
17      mixture (i.e., elemental carbon particles, O3, ammonium bisulfate, but with NO2 also) and
18      exposure regimen as Kleinman et al. (2000).  Also observed were were decreases in pulmonary
19      macrophage Fc-receptor binding and phagocytosis and increases in acid phosphatase staining.
20      Bronchoalveolar epithelial permeability and cell proliferation were increased.  Altered breathing-
21      patterns also were observed, with some adaptations occurring.
22           Bolarin et al. (1997) exposed rats to 50 or 100 |ig/m3 carbon particles in combination with
23      ammonium bisulfate and O3. Despite 4 weeks of exposure, they observed no changes in protein
24      concentration in lavage fluid or blood prolyl 4-hydroxylase, an enzyme involved in collagen
25      metabolism. Slight decreases in plasma fibronectin were present in animals exposed to the
26      combined pollutants versus O3 alone. Thus, the potential for adverse  effects in the lungs of
27      animals challenged with a combined exposure to particles and gaseous pollutants is dependent
28      on numerous factors, including the gaseous co-pollutant, concentration,  and time.
29           In a complex series of studies, Oberdorster and colleagues examined the interaction of
30      several pulmonary oxidative stress pollutants. Elder et al. (2000a,b) reported the results of
31      combined exposure to ultrafine carbon particles (100 |ig/m3) and O3 (1 ppm) in young and old

        December 2005                           5-62       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Fischer 344 rats that were pretreated with aerosolized endotoxin. In old rats, exposure to carbon
 2      and O3 produced an interaction that resulted in a greater influx in neutrophils than that produced
 3      by either agent alone. This interaction was not seen in young rats.  Oxidant release from lavage
 4      fluid cells also was assessed and the combination of endotoxin, carbon particles, and O3
 5      produced an increase in oxidant release in old rats. This mixture produced the opposite response
 6      in the cells recovered from the lungs of the young rats, indicating that the lungs of the aged
 7      animals underwent greater oxidative stress in response to a complex pollutant mix of particles,
 8      O3, and a biogenic agent.  Johnston et al. (2000; 2002) reported the results of combined exposure
 9      to O3 (1.0 and 2.5 ppm for 4, 20, or 24 h) and low-dose endotoxin, or to O3 and endotoxin
10      separately, in newborn and adult C57BL/6J mice. In the first study, adult (8 wk  old) mice
11      showed greater sensitivity to O3 than newborn (36 h old) mice on the basis of mRNAs encoding
12      for various chemokines and  cytokines.  In contrast, adult and newborn mice responded similarly
13      2 h after endotoxin exposure (10 ng for 10 min), suggesting that age differences  in O3-generated
14      inflammation is secondary to epithelial cell injury. In the second study, 8 wk old mice exposed
15      to O3 (1 ppm for 24 h) followed by endotoxin (37.5 EU for 10 min) showed increased
16      responsiveness over either exposure alone, on the basis of increased expression of chemokine
17      and cytokine messages and increased BAL fluid levels of protein and PMNs.
18           Fanucchi et al. (1998)  and Wagner et al. (2001a,b) examined the synergistic effect of
19      coexposure to O3 and endotoxin on the nasal transitional epithelium of rats that also was
20      mediated, in part, by neutrophils.  Fisher 344 rats intranasally instilled with endotoxin and
21      exposed to 0.5 ppm O3, 8 h per day, for 3 days developed mucous cell metaplasia in the nasal
22      transitional epithelium,  an area normally devoid of mucous cells; whereas,  intratracheal
23      instillation of endotoxin (20 jig) caused mucous cell metaplasia rapidly in the respiratory
24      epithelium of the conducting airways.  A synergistic increase of intraepithelial mucosubstances
25      and morphological evidence of mucous cell metaplasia were found in rat maxilloturbinates upon
26      exposure to both O3 and endotoxin, compared to each pollutant alone.  A similar response was
27      reported in OVA-sensitized Brown Norway rats exposed to 0.5 ppm O3, 8 h/day  for 3 days
28      (Wagner et al., 2002), indicating that coexposure to O3 and inflammatory biogenic substances
29      like allergens (e.g., OVA) or bacterial  endotoxin can augment epithelial and inflammatory
30      responses in rat nasal passages.
        December 2005                            5-63         DRAFT-DO NOT QUOTE OR CITE

-------
 1           In follow-up studies, Wagner et al. (2003) reported that coexposure of rats to O3 and
 2      endotoxin also enhanced epithelial and neutrophilic inflammatory responses in the pulmonary
 3      airways. Fisher 344 rats were intranasally instilled with endotoxin and exposed to 1.0 ppm O3
 4      for 8 h, which was repeated 24 h later. Three days after the last exposure, BALF was analyzed
 5      for inflammatory cells and secreted mucosubstances (mucin SAC), and lung tissue was processed
 6      for morphometric analysis.  Endotoxin instillation alone caused a dose-dependent increase in
 7      BALF neutrophils that was further increased 2-fold in O3-exposed rats given 20 [ig endotoxin,
 8      the highest dose.  Mucin glycoprotein SAC also was increased in the BALF at this dose but not
 9      at lower endotoxin doses. Ozone exposure alone did not cause mucus hypersecretion, but it did
10      potentiate mucus secretion in rats given both 2 and 20 [ig endotoxin and increased intraepithelial
11      mucosub stances 2-fold, which was further substantiated by significant increases in mucin gene
12      (rMucSAC) mRNA levels in the conducting airways.
13           The effect of O3 modifying the biological potency of PM (diesel PM and carbon black) was
14      examined by Madden et al. (2000) in rats. Reaction of NIST Standard Reference Material
15      # 2975 diesel PM with 0.1 ppm O3 for 48 hr increased the potency (compared to unexposed or
16      air-exposed diesel PM) to induce neutrophil influx, total protein, and LDH in lung lavage fluid in
17      response to intratracheal instillation. Exposure of the diesel PM to high, non-ambient O3
18      concentration (1.0 ppm) attenuated the increased potency, suggesting destruction of the bioactive
19      reaction products. Unlike the diesel particles, carbon black particles exposed to 0.1 ppm O3 did
20      not exhibit an increase in biological potency, which suggested that the reaction of organic
21      components of the diesel PM with O3 were responsible for the increased potency.
22           Ulrich et al. (2002) investigated the effect of ambient PM from Ottawa Canada (EHC-93)
23      on O3-induced inflammation. Male Wistar rats were exposed to 0.8 ppm O3 for 8 h and allowed
24      to recover before intratracheal instillation of 0.5, 1.5, and 5 mg of EHC-93 in 0.3 ml of saline.
25      The high concentrations of PM used were sufficient to induce pulmonary inflammation, which
26      was not exacerbated by pre-exposure to O3.  Rats from the combined exposure group did have
27      higher and more persistent lung lavage protein and albumin levels, as well as increased plasma
28      fibrinogen levels when compared to PM exposure alone.
29           The interaction of PM and O3 was further examined in a murine model of OVA-induced
30      asthma. Kobzik et al. (2001) investigated whether coexposure to inhaled, concentrated ambient
31      particles (CAPs) from Boston, MA and to O3 could exacerbate asthma-like symptoms.  On days

        December 2005                            5-64        DRAFT-DO NOT QUOTE OR CITE

-------
 1      7 and 14 of life, half of the BALB/c mice used in this study were sensitized by intraperitoneal
 2      (ip) injection of OVA and then exposed to OVA aerosol on three successive days to create the
 3      asthma phenotype. The other half received the ip OVA but were exposed to a phosphate-
 4      buffered saline aerosol (controls).  The mice were further subdivided (n > 6I/group) and exposed
 5      for 5 h to CAPs, ranging from 63 to 1,569 |ig/m3, 0.3 ppm O3, CAPs + O3, or to FA. Pulmonary
 6      resistance and airway  responsiveness to an aerosolized MCh challenge were measured after
 7      exposures. A small, statistically significant increase in pulmonary resistance and airway
 8      responsiveness, respectively, was found in both normal and asthmatic mice immediately after
 9      exposure to CAPs alone and to CAPs + O3 but not to O3 alone or to FA. By 24 h after exposure,
10      the responses returned to  baseline levels.  There were no significant increases in airway
11      inflammation after any of the pollutant exposures. In this well-designed study of a small-animal
12      model of asthma, O3 and CAPs did not appear to be synergistic. In further analysis of the data
13      using specific elemental groupings of the CAPs, the acutely increased pulmonary resistance was
14      found to be associated with the AISi  fraction of PM.  Thus, some components of concentrated
15      PM2 5 may affect airway caliber in sensitized animals, but the results are difficult to extrapolate
16      to p eopl e with asthma.
17           Animal studies have examined the adverse cardiopulmonary effects of complex  mixtures in
18      urban and rural environments  of Italy (Gulisano et al., 1997), Spain (Lorz and Lopez,  1997), and
19      Mexico (Vanda et al.,  1998; Moss et al., 2001). Some of these studies have taken advantage of
20      the differences in pollutant mixtures  of urban and rural environments to report primarily
21      morphological changes in the  nasopharynx and lower respiratory tract (Gulisano et al., 1997;
22      Lorz and Lopez, 1997) of lambs and pigeons, respectively, after natural, continuous exposures to
23      ambient pollution. Each study has provided evidence that animals living in urban air pollutants
24      have greater pulmonary changes than would occur in a rural and presumably cleaner,
25      environment.  The study by Moss et al. (2001) examined the nasal and lung tissue of rats
26      exposed  (23 h/day) to Mexico City air for up to 7 weeks and compared them to controls similarly
27      exposed  to FA. No inflammatory or epithelial lesions were found using quantitative
28      morphological techniques; however,  the concentrations of pollutants were low. Extrapolation of
29      these results to humans is restricted, however, by uncontrolled exposure conditions, small
30      sample sizes, and other unknown exposure and nutritional factors in the studies in mammals and
31      birds, and the negative studies in rodents.  They also bring up the issue of which species of

        December 2005                           5-65       DRAFT-DO NOT QUOTE OR CITE

-------
 1      "sentinel" animals is more useful for predicting urban pollutant effects in humans. Thus, in these
 2      field studies, it is difficult to assign a specific role to any specific component of the mixture for
 3      the significant cardiopulmonary effects reported.
 4           Similar morphological changes (Calderon-Garciduefias et al., 2000a; 2001) and chest X-ray
 5      evidence of mild lung hyperinflation (Calderon-Garciduefias et al., 2000b) have been reported in
 6      children residing in urban and rural areas of Mexico City. (See Chapter 7 for details of these
 1      studies^) The ambient air in urban areas, particularly in Southwestern Mexico City, is a complex
 8      mixture of particles and gases, including high concentrations of O3 and aldehydes that previously
 9      have been shown to cause airway inflammation and epithelial lesions in humans (e.g., Calderon-
10      Garciduefias et al., 1992, 1994, 1996) and laboratory animals (Morgan et al., 1986; Heck et al.,
11      1990; Harkema et al., 1994, 1997a,b).  The described effects demonstrate a persistent, ongoing
12      upper and lower airway inflammatory process and chest X-ray abnormalities in children residing
13      predominantly in highly polluted areas. Again, extrapolation of these results to urban
14      populations of the United States is difficult because of the unique complex of urban air in
15      Mexico City, uncontrolled exposure conditions, and other unknown exposure and nutritional
16      factors.
17
18      5.4.4   Summary and Conclusions - Interactions of Ozone with other
19             Co-occurring Pollutants
20          It is difficult to summarize the role that O3 plays in exposure responses to binary mixtures,
21      and even harder to determine its role in responses to multicomponent, complex atmospheres.
22      Though the specific mechanisms of action of the individual pollutants within a mixture may be
23      known, the exact bases for toxic interactions have not been elucidated clearly. Certain generic
24      mechanisms that may underlie pollutant interactions: (1) physical, involving adsorption of one
25      pollutant onto another and subsequent transport to more or less sensitive sites or to sites where
26      one of the components of the mixture normally would not deposit in concentrated amounts
27      (probably not involved in O3-related interactions; (2) production of secondary products that may
28      be more lexicologically active than the primary materials, demonstrated or suggested in a
29      number of studies as a basis for interaction between O3 and NO2 and between O3  and PM;
30      (3) biological or chemical alterations at target sites that affect response to O3 or the copollutant,
31      which which has been suggested to underlie interactions with mixtures of O3 and acid sulfates; 4)

        December 2005                            5-66        DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3- or copollutant-induced physiological change, such as alteration in ventilation pattern,
 2      resulting in changes in the penetration or deposition of one pollutant when another is present.
 3      This has been implicated in enhanced responses to various O3-containing mixtures with exercise.
 4           Evaluation of interactions between O3 and copollutants is a complex procedure. Responses
 5      are dependent on a number of host and environmental factors, such that different studies using
 6      the same copollutants may show different types or magnitudes of interactions. The occurrence
 7      and nature of any interaction is dependent on the endpoint being examined and is also highly
 8      related to the specific conditions of each study, such as animal species, health status, exposure
 9      method, dose, exposure  sequence, and the physicochemical characteristics of the copollutants.
10      Because of this, it is difficult to compare studies, even those examining similar endpoints, that
11      were performed under different exposure conditions. Thus, any description of interactions is
12      really valid only for the  specific conditions of the study in question and cannot be generalized to
13      all conditions of exposure to a particular chemical mixture.  Furthermore, it is generally not
14      possible to extrapolate the effect of pollutant mixtures from studies on the effects of each
15      component when given separately. In any case, what can be concluded from the database is that
16      interactions of O3-containing mixtures are generally synergistic (antagonism has been noted in a
17      few studies), depending on the various factors noted above, and that O3 may produce more
18      significant biological responses as a component of a mixture than when inhaled alone.
19      Furthermore, although most studies have shown that interaction occurs only at higher than
20      ambient concentrations with acute exposure,  some have demonstrated interaction at more
21      environmentally relevant levels (e.g., 0.05 to 0.1 ppm O3 with NO2) and with  repeated exposures.
22
23
24      5.5   EFFECTS OF OTHER PHOTOCHEMICAL OXIDANTS
25           Peroxyacetyl nitrate (PAN) and peroxypropionyl nitrate (PPN) are the most abundant
26      non-O3 oxidants in ambient air of industrialized areas, other than the inorganic nitrogenous
27      oxidants such as NO2, and possibly HNO3. Ambient levels  of PAN and PPN were reported to be
28      decreasing over the 1990's and available air quality data (Grosjean et al., 2001; Grosjean, 2003;
29      Jakobi and Fabian, 1997) indicate that present peak concentrations of PAN and PPN in ambient
30      air from urban areas are in the low ppb range (e.g., < 1 to 10 ppb).   The levels found in nonurban
31      areas are considerably lower (Gaffney et al.,  1993).

        December 2005                           5-67         DRAFT-DO NOT QUOTE OR CITE

-------
 1           Reactions occur in the troposphere between O3 and hydrocarbons (e.g., d-limonene) to
 2      produce epoxides, hydroperoxides, and peroxides. The majority of the measured ambient
 3      hydroperoxides produced is hydrogen peroxide (H2O2), although a small amount of organic
 4      hydroperoxides (ROOH) also may be formed.  Friedlander and Yeh (1998) have estimated that
 5      atmospheric aerosols can carry as high as 1 mM of H2O2 and organic hydroperoxides (e.g.,
 6      hydroxymethylhydroperoxide) in the associated water. In vitro cell and tissue damage are
 7      induced by high concentrations of liquid phase H2O2 (50 jiM to 1 mM). Morio et al. (2001)
 8      demonstrated that 10 and 20 ppb of inhaled H2O2 vapor can penetrate the lower lung where it
 9      causes inflammation. It is likely that hygroscopic components of PM transport ambient H2O2
10      into the lower lung and induce  tissue injury as well.  Exposure of rats to a H2O2-fine particle
11      mixture (215 or 429 |ig/m3 ammonium sulfate) resulted in increased neutrophil influx, and
12      production of inflammatory mediators by AMs (Morio et al., 2001). Hygroscopic secondary
13      organic aerosols generated by the O3/hydrocarbon reactions and their co-occurrence with H2O2
14      also provides another possible mechanism, yet to be validated, whereby H2O2 can be transported
15      into the lower respiratory tract  (e.g., Friedlander and Yeh, 1998). Interaction of inhaled O3 with
16      unsaturated fatty acids on cell membranes and mucus in the airways generates epoxides,
17      hydroperoxides, and secondary ozonation products such as 4-hydroxynonenal (see Section 5.2.1)
18           Inhalation toxicological information on the effects of the non-O3 oxidants has been limited
19      to a few studies on PAN, but at concentrations much higher (approximately 100- to 1,000 fold)
20      than levels typically found in ambient air.  Such high acute levels cause changes in lung
21      morphology, behavioral modifications, weight loss, and susceptibility to pulmonary infections.
22      Therefore, acute toxicity of PAN is much lower than O3, and it is unlikely that present ambient
23      PAN levels would affect pulmonary function responses to O3 (Vyskocil et al., 1998).
24      Cytogenetic studies indicate that PAN is not a potent mutagen, clastogen, or DNA damaging
25      agent in mammalian cells in vivo or in vitro at concentrations several orders of magnitude higher
26      than the generally encountered ambient air levels in most cities (Vyskocil et al., 1998; Kligerman
27      et al., 1995; Heddle et al., 1993).  Some studies suggest that PAN may be a weak bacterial
28      mutagen at concentrations much higher than exist in present urban atmospheres (DeMarini et al.,
29      2000; Kleindienst et al., 1990).
30
        December 2005                            5-68        DRAFT-DO NOT QUOTE OR CITE

-------
1     5.5.1  Summary and Conclusions - Effects of Other Photochemical Oxidants
2          Concentrations of PAN and PPN (<1 to 10 ppb) in ambient air are unlikely to affect
3     pulmonary function or cause DNA damage. Levels of 10-20 ppm H2O2 can penetrate to the
4     lower lung directly or be transported there by PM, where inflammation can result; however,
5     ambient levels of are typically < ~5 ppb. As toxicology studies of other photochemical oxidants
6     are rare, quantitative scientific evaluations of possible health effects of environmental exposures
7     cannot be completed at this time.
      December 2005                           5-69        DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Adamson, I. Y. R.; Vincent, R.; Bjarnason, S. G. (1999) Cell injury and interstitial inflammation in rat lung after
  3             inhalation of ozone and urban particulates. Am. J. Respir. Cell Mol. Biol. 20: 1067-1072.
  4       Aizawa, H.; Koto, H.; Nakano, H.; Inoue, H.; Matsumoto, K.; Takata, S.; Shigyo, M.; Hara, N. (1997) The effects of
  5             a specific tachykinin receptor antagonist FK-224 on ozone-induced airway hyperresponsiveness and
  6             inflammation. Respirology 2: 261-265.
  7       Aizawa, H.; Shigyo, M.; Nakano, H.; Matsumoto, K.; Inoue, H.; Hara, N. (1999a) Effect of the Chinese herbal
  8             medicine, Bakumondo-to, on airway hyperresponsiveness induced by ozone exposure in guinea-pigs.
  9             Respirology 4: 349-354.
10       Aizawa, H.; Shigyo, M.; Matsumoto, K.; Inoue, H.; Koto, H.; Hara, N. (1999b) PACAP reverses airway
11             hyperresponsiveness induced by ozone exposure in guinea pigs. Respiration 66: 538-542.
12       Alfaro, M. F.; Putney, L.; Tarkington, B. K.; Hatch, G. E.; Hyde, D. M.; Schelegle, E. S. (2004) Effect of rapid
13             shallow breathing on the distribution of 18O-labeled ozone reaction product in the respiratory tract of the rat.
14             Inhalation Toxicol.:  in press.
15       Anderson, S. D. (1996) Challenge tests to assess airway hyperresponsiveness and efficacy of drugs used in the
16             treatment of asthma. J. Aerosol Med. 9: 95-109.
17       Anderson, S. D.; Daviskas, E. (2000) The mechanism of exercise-induced asthma is ... J. Allergy Clin. Immunol.
18             106:453-459.
19       Arito, H.; Takahashi, M.; Iwasaki, T.; Uchiyama, I. (1997) Age-related changes in ventilatory  and heart rate
20             responses to acute ozone exposure in the conscious rat. Ind.  Health 35: 78-86.
21       Arsalane, K.; Gosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
22             synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
23             13:60-68.
24       Avila-Costa, M. R.; Colin-Barenque, L.; Fortoul, T. L; Machado-Salas, J. P.; Espinosa-Villanueva, J.;
25             Rugerio-Vargas, C.; Rivas-Arancibia, S. (1999) Memory deterioration in an oxidative stress model and its
26             correlation with cytological changes on rat hippocampus CA1. Neurosci. Lett. 270: 107-109.
27       Avital, A.; Springer, C.; Bar-Yishay, E.; Godfrey, S. (1995a) Adenosine, methacholine, and exercise challenges in
28             children with asthma or paediatric chronic obstructive pulmonary disease. Thorax 50: 511-516.
29       Avital, A.; Picard, E.; Uwyyed, K.; Springer, C. (1995b) Comparison of adenosine 5'-monophosphate and
3 0             methacoline for the differentiation of asthma from chronic airway diseases with the use of the auscultative
31             method in very young children. J. Pediatr. 127:  438-440.
32       Bassett, D.; Elbon-Copp, C.; Otterbein, S.; Barraclough-Mitchell, H.; DeLorme, M.; Yang, H. (2001) Inflammatory
33             cell availability affects ozone-induced lung damage. J. Toxicol. Environ. Health A 64: 547-565.
34       Becker, S.; Quay, J.; Koren, H.  S. (1991) Effect of ozone on immunoglobulin production by humanB cells  in vitro.
35             J. Toxicol. Environ.  Health 34: 353-366.
36       Beeson, W. L.; Abbey, D. E.; Knutsen, S. F. (1998) Long-term concentrations of ambient air pollutants and incident
37             lung cancer in California adults: results from the AHSMOG study. Environ. Health Perspect. 106: 813-823.
3 8       Bermudez, E.  (2001) Detection of poly(ADP-ribose) synthetase activity in alveolar macrophages of rats exposed to
39             nitrogen dioxide and ozone. Inhalation Toxicol. 13: 69-84.
40       Bermudez, E.; Ferng,  S. F.; Castro, C. E.; Mustafa, M. G. (1999) DNA strand breaks caused by exposure to ozone
41             and nitrogen dioxide. Environ. Res. 81: 72-80.
42       Bhalla, D. K. (1996) Alteration of alveolar macrophage chemotaxis, cell adhesion, and cell adhesion molecules
43             following ozone exposure of rats. J. Cell. Physiol. 169: 429-438.
44       Bhalla, D. K.; Gupta, S. K. (2000) Lung injury, inflammation, and inflammatory stimuli in rats exposed to ozone.
45             J. Toxicol. Environ.  Health 59: 211-228.
46       Bhalla, D. K.; Gupta, S. K.; Reinhart, P. G. (1999) Alteration of epithelial integrity, alkaline phosphatase activity,
47             and fibronectin expression in lungs of rats exposed to ozone. J.  Toxicol. Environ. Health A 56: 329-343.
48       Bhalla, D. K.; Reinhart, P. G.; Bai, C.; Gupta,  S. K. (2002) Amelioration of ozone-induced lung injury by anti-tumor
49             necrosis factor-"alpha". Toxicol. Sci. 69: 400-408.
50       Bignami, G.; Musi, B.; DeH'Omo, G.; Laviola, G.; Alleva, E. (1994) Limited effects of ozone exposure during
51             pregnancy on physical and neurobehavioral development of CD-I mice.  Toxicol. Appl. Pharmacol.
52             129:264-271.
53       Bimonte, H. A.; Nelson, M. E.;  Granholm, A. C. (2003) Age-related deficits as working memory load increases:
54             relationships with growth factors. Neurobiol. Aging 24: 37-48.
         December 2005                                 5-70          DRAFT-DO NOT QUOTE OR CITE

-------
  1       Bolarin, D. M; Bhalla, D. K.; Kleinman, M. T. (1997) Effects of repeated exposures of geriatric rats to ozone and
  2             particle-containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation
  3             Toxicol. 9: 423-434.
  4       Bonham, A. C.; Chen, C. Y.; Mutoh, T.; Joad, J. P. (2001) Lung C-fiber CNS reflex: role in the respiratory
  5             consequences of extended environmental tobacco smoke exposure ini young guinea pigs. Environ. Health
  6             Perspect. 109(suppl. 4): 573-578.
  7       Boorman, G. A.; Hailey, R.; Grumbein, S.; Chou, B. J.; Herbert, R. A.; Goehl, T.; Mellick, P. W.; Roycroft, J. H.;
  8             Haseman, J. K.; Sills, R. (1994) Toxicology and carcinogenesis studies of ozone and ozone
  9             4-(A^-nitrosomethylamino)-l-(3-pyridyl)-l-butanone in Fischer-344/N rats. Toxicol. Pathol. 22: 545-554.
10       Bornholdt, J.; Dybdahl, M.; Vogel, U.; Hansen, M.; Loft, S.; Wallin, H.  (2002) Inhalation of ozone induces DNA
11             strand breaks and inflammation in mice. Mutat. Res. 520: 63-71.
12       Brannan, J. D.; Koskela, H.; Anderson, S. D.; Chew, N. (1998) Responsiveness to Mannitol in asthmatic subjects
13             with exercise- and hyperventiliation-induced asthma. Am. J. Respir. Crit. Care Med. 158: 1120-1126.
14       Bridges, J. P.; Davis, H. W.; Demodarasamy, M.; Kuroki, Y.; Howies, G.; Hui, D. Y.; McCormack, F. X. (2000)
15             Pulmonary surfactant proteins A and D are potent endogenous inhibitors of lipid peroxidation and oxidative
16             cellular injury. J. Biol. Chem. 275: 38848-38855.
17       Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A. (2003) Lung hyperpermeability, Clara-cell
18             secretory potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhalation Toxicol.
19             15: 1209-1230.
20       Calderon-Garciduenas, L.; Osorno-Velaquez, A.; Bravo-Alvarez, H.; Delgado-Chavez, R.; Barrios-Marquez, R.
21             (1992) Histopathologic changes of the nasal mucosa in Southwest metropolitan Mexico City inhabitants.
22             Am. J. Pathol. 140: 225-232.
23       Calderon-Garciduenas, L.; Rodriguez-Alcaraz, A.; Garcia, R.; Sanchez,  G.; Barragan, G.; Camacho,  R.; Ramirez,
24             L. (1994) Human nasal mucosal changes after exposure to urban pollution. Environ. Health Perspect.
25             102: 1074-1080.
26       Calderon-Garciduenas, L.; Osnaya-Brizuela, N.; Ramirez-Martinez, L.; Villarreal-Calderon, A. (1996) DNA strand
27             breaks in human nasal respiratory epithelium are induced upon exposure to urban pollution. Environ. Health
28             Perspect. 104: 160-168.
29       Calderon-Garciduenas, L.; Wen-Wang, L.; Zhang, Y.-J.; Rodriguez-Alcaraz, A.; Osnaya, N.; Villarreal-Calderon,
30             A.; Santella, R. M. (1999) 8-hydroxy-2'-deoxyguanosine, a major mutagenic oxidative DNA lesion, and DNA
31             strand breaks in nasal respiratory epithelium of children exposed to urban pollution. Environ. Health Perspect.
32             107:469-474.
33       Calderon-Garciduenas, L.; Devlin, R. B.; Miller, F. J. (2000a) Respiratory tract pathology and cytokine imbalance in
34             clinically healthy children chronically and sequentially exposed to air pollutants. Med. Hypotheses 55:
35             373-378.
36       Calderon-Garciduenas, L.; Mora-Tiscareno, A.; Chung, C. J.; Valencia, G.; Fordham, L. A.; Garcia, R.;  Osnaya, N.;
37             Romero, L.; Acuna, H.; Villarreal-Calderon, A. (2000b) Exposure to air pollution is associated with lung
3 8             hyperinflation in healthy children and adolescents in southwest Mexico City: a pilot study. Inhalation Toxicol.
39             12:537-561.
40       Calderon-Garciduenas, L.; Gambling, T. M.; Acuna, H.; Garcia, R.; Osnaya, N.; Monroy, S.; Villarreal-Calderon,
41             A.; Carson, J.; Koren, H. S.; Devlin, R. B. (2001a) Canines as sentinel species for assessing chronic
42             exposures to air pollutants: part 2. cardiac pathology. Toxicol. Sci. 61:  356-367.
43       Calderon-Garciduenas, L.; Rodrigues-Alcaraz, A.; Valencia-Salazar, G.; Mora-Tascareno, A.; Garcia, R.;
44             Osnaya, N.; Villarreal-Calderon, A.; Devlin, R. B.; Van Dyke, T.  L. (2001b) Nasal biopsies of children
45             exposed to air pollutants. Toxicol. Pathol. 29: 558-564.
46       Campos-Bedolla, P.; Vargas, M. H.; Montano, L. M. (2002) Effect of acute ozone exposure on pregnant rat uterus
47             contractile responses. Reprod. Toxicol. 16: 269-273.
48       Cassee, F. R.; Feron, V. J. (1994) Biochemical and histopathological changes in nasal epithelium of rats after 3-day
49             intermittent exposure to formaldehyde and ozone alone or in combination. Toxicol. Lett. 72: 257-268.
50       Chang, M. M.-J.; Wu, R.; Plopper, C. G.; Hyde, D. M. (1998) IL-8 is one of the major chemokines produced by
51             monkey airway epithelium after ozone-induced injury. Am. J. Physiol.  275: L524-L532.
52       Cheek, J. M.; McDonald, R. J.; Rapalyea, L.; Tarkington, B. K.; Hyde, D. M. (1995) Neutrophils enhance removal
53             of ozone-injured alveolar epithelial cells in vitro. Am. J. Physiol. 269: L527-L535.
54       Chen, L.  C.; Qu, Q.; Amdur, M. O.; Schlesinger, R. B. (1995) Alteration of pulmonary macrophage intracellular pH
55             following inhalation exposure to sulfuric acid/ozone mixtures. Exp. Lung Res. 21: 113-128.
         December 2005                                 5-71          DRAFT-DO NOT QUOTE OR CITE

-------
  1       Chen, L.; Yang, W.; Jennison, B. L.; Goodrich, A.; Omaye, S. T. (2002) Air pollution and birth weight in northern
  2             Nevada, 1991-1999. Inhalation Toxicol. 14: 141-157.
  3       Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
  4             contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure in infant
  5             primates. J. Appl. Physiol. 94: 819-827.
  6       Cho, H. Y.; Hotchkiss, J. A.; Harkema, J. R. (1999a) Inflammatory and epithelial responses during the development
  7             of ozone-induced mucous cell metaplasia in the  nasal epithelium of rats. Toxicol. Sci.  51: 135-145.
  8       Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (1999b) Effects of pre-existing rhinitis on
  9             ozone-induced mucous cell metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol.  158: 92-102.
10       Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (2000) Neutrophil-dependent and
11             neutrophil-independent alterations in the nasal epithelium of ozone-exposed rats. Am.  J. Respir. Crit. Care
12             Med. 162: 629-636.
13       Cho, H.-Y.; Zhang, L.-Y.; Kleeberger, S. R. (2001) Ozone-induced lung inflammation and hyperreactivity are
14             mediated via tumor necrosis factor-a receptors. Am. J. Physiol. 280: L537-L546.
15       Churg, A.; Brauer,  M; Keeling, B. (1996) Ozone enhances the uptake of mineral particles by tracheobronchial
16             epithelial cells in organ culture. Am.  J. Respir. Crit. Care Med. 153: 1230-1233.
17       Clausen, P. A.; Wilkins, C. K.; Wolkoff, P.; Nielsen, G. D. (2001) Chemical and biological evaluation of a reaction
18             mixture of R-(+)-limonene/ozone: formation of strong airway irritants. Environ. Int. 26: 511-522.
19       Cohen, M. D.; Zelikoff, J. T.; Qu, Q.; Schlesinger, R. B. (1996) Effects of ozone upon macrophage-interferon
20             interactions. Toxicology 114: 243-252.
21       Cohen, M. D.; Sisco, M.; Li, Y.; Zelikoff, J. T.; Schlesinger, R.  B. (2001) Ozone-induced modulation of
22             cell-mediated immune responses in the lungs.  Toxicol. Appl. Pharmacol. 171: 71-84.
23       Cohen, M. D.; Sisco, M.; Baker, K.; Li, Y.;  Lawrence,  D.; VanLoveren, H.; Zelikoff, J. T.; Schlesinger, R. B.
24             (2002) Effects of inhaled ozone on pulmonary immune cells critical to antibacterial responses in situ.
25             Inhalation Toxicol. 14: 599-619.
26       Colin-Barenque, L.; Avila-Costa, M. R.; Fortoul, T.; Rugerio-Vargas, C.; Machado-Salas, J. P.;
27             Espinosa-Villanueva, J.; Rivas-Arancibia, S. (1999) Morphologic alteration of the olfactory bulb after acute
28             ozone exposure in rats. Neurosci. Lett. 274: 1-4.
29       Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
30             reactive absorption-mediated cytotoxicity in vitro.  Am. J. Physiol.: 10.1152/ajplung.00397.2003.
31       Cotovio, J.; Onno, L.; Justine, P.; Lamure, S.; Catroux, P. (2001) Generation of oxidative stress in human cutaneous
32             models following in vitro ozone exposure. Toxicol. in Vitro 15: 357-362.
33       Cottet-Emard, J.-M.; Dalmaz, Y.; Pequignot, J.; Peyrin, L.; Pequignot, J.-M. (1997) Long-term exposure to ozone
34             alters peripheral and central catecholamine activity in rats. Pfluegers Arch. 433: 744-749.
35       Creutzenberg, O.; Bellmann, B.; Klingebiel, R.; Heinrich, U.; Muhle, H. (1995) Phagocytosis and chemotaxis of rat
36             alveolar macrophages after a combined or separate exposure to ozone and carbon black. Exp. Toxicol. Pathol.
37             47:202-206.
38       Cross, C. E.; Van Der Vliet, A.; Louie, S.; Thiele, J. J.; Halliwell, B. (1998) Oxidative stress  and antioxidants at
39             biosurfaces:  plants, skin, and respiratory tract surfaces. Environ. Health Perspect. 106(suppl. 5): 1241-1251.
40       Custodio-Ramirez,  V.; Paz, C. (1997) Ozone produces  functional deficits in the rat visual pathway.
41             Electroencephalogr. Clin. Neurophysiol. 104:  269-273.
42       Daly, C.; Fox, K.; Henein, M. (2002) Natriuretic peptides in the diagnosis of heart disease-first  amongst equals?
43             Int. J. Cardiol. 84: 107-113.
44       Delacourt, C.; Benoist, M. R.; Waernessyckle, S. (2001) Relationship between bronchial responsiveness and clinical
45             evolution in infants who wheeze. A four-year prospective study. Am. J. Respir. Crit. Care Med.
46             164: 1382-1386.
47       Delaunois, A.; Segura, P.; Montano, L. M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998) Comparison of
48             ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol. Appl. Pharmacol.
49             150:58-67.
50       DeH'Omo, G.; Fiore, M.; Petruzzi, S.; Alleva, E.; Bignami, G. (1995a) Neurobehavioral development of CD-I mice
51             after combined gestational and postnatal exposure  to ozone. Arch. Toxicol. 69: 608-616.
52       DeH'Omo, G.; Wolfer, D.; Alleva, E.; Lipp,  H.-P. (1995b) Developmental exposure to ozone  induces subtle changes
53             in swimming navigation of adult mice. Toxicol.  Lett. 81: 91-99.
54       DeLorme, M. P.; Yang, H.; Elbon-Copp, C.; Gao, X.; Barraclough-Mitchell, H.;  Bassett, D. J. P. (2002)
55             Hyperresponsive airways correlate with lung tissue inflammatory cell changes in ozone-exposed rats.
56             J. Toxicol. Environ. Health Part A 65: 1453-1470.


         December 2005                                  5-72         DRAFT-DO NOT QUOTE OR CITE

-------
  1       DeMarini, D. M; Shelton, M. L.; Kohan, M. J.; Hudgens, E. E.; Kleindienst, T. E.; Ball, L. M.; Walsh, D.; de Boer,
  2             J. G.; Lewis-Bevan, L.; Rabinowitz, J. R.; Claxton, L. D.; Lewtas, J. (2000) Mutagenicity in lung of Big
  3             Blue(R) mice and induction of tandem-base substitutions in Salmonella by the air pollutant peroxyacetyl
  4             nitrate (PAN): predicted formation of intrastrand cross-links. Mutat. Res. 457: 41-55.
  5       Depuydt, P.; Joos, G. F.; Pauwels, R. A. (1999) Ambient ozone concentrations induce airway hyperresponsiveness
  6             in some rat strains. Eur. Respir. J. 14: 125-131.
  7       Dorado-Martinez, C.; Parades-Carbajal, C.; Mascher, D.; Borgonio-Perez, G.; Rivas-Arancibia, S. (2001) Effects of
  8             different ozone doses on memory, motor activity and lipid peroxidation levels, in rats. Int. J. Neurosci.
  9             108: 149-161.
10       Dormans, J. A. M. A.; Boere, A. J. F.; vanLoveren, H.; Rombout, P. J. A.; Marra, M.; vanBree, L. (1996)
11             Age-related toxicity in rat lungs following acute and repeated ozone exposure. Inhalation Toxicol. 8:  903-925.
12       Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
13             time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
14       Driscoll, K. E.; Simpson, L.; Carter, J.; Hassenbein, D.; Leikauf, G. D. (1993) Ozone inhalation stimulates
15             expression of a neutrophil chemotactic protein, macrophage inflammatory protein 2. Toxicol. Appl.
16             Pharmacol. 119: 306-309.
17       Dye, J. A.;  Madden, M. C.; Richards, J. H.; Lehmann, J. R.; Devlin, R. B.; Costa, D. L.  (1999) Ozone effects on
18             airway responsiveness, lung injury, and inflammation. Comparative rat strain and in vivo/in vitro
19             investigations. Inhalation Toxicol. 11: 1015-1040.
20       El-Fawal, H. A. N.; McGovern, T.; Schlesinger, R. B. (1995) Nonspecific bronchial responsiveness assessed in vitro
21             following acute inhalation exposure to ozone and ozone/sulfuric acid mixtures. Exp. Lung Res. 21: 129-139.
22       Elder, A. C. P.; Gelein, R.; Finkelstein J. N.; Cox,  C.;  Oberdorster, G. (2000a) Endotoxin priming affects the lung
23             response to ultrafine particles and ozone in young and old rats. In: Phalen, R. F.,  ed. Inhalation toxicology:
24             proceedings of the third colloquium on paniculate air pollution and human health (first special issue); June,
25             1999; Durham, NC. Inhalation Toxicol. 12(suppl. 1): 85-98.
26       Elder, A. C. P.; Gelein, R.; Finkelstein, J. N.; Cox, C.; Oberdorster, G. (2000b) Pulmonary inflammatory response to
27             inhaled ultrafine particles is modified by age, ozone exposure, and bacterial toxin. In: Grant, L. D., ed.
28             PM2000: paniculate matter and health. Inhalation Toxicol. 12(suppl. 4): 227-246.
29       Elsayed, N. M. (2001) Diet restriction modulates lung response and survivability of rats exposed to ozone.
30             Toxicology 159:  171-182.
31       Evans, M. J.; Johnson, L. V.; Stephens, R. J.; Freeman, G.  (1976) Renewal of the terminal bronchiolar epithelium in
32             the rat following exposure to NO2 or O3. Lab. Invest. 35: 246-257.
33       Evans, M. J.; Fanucchi, M. V.; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.; Schelegle, E. S.;
34             Gershwhin, L. J.; Miller, L. A.; Hyde, D. M.; Sannes,  P. L.; Plopper, C. G. (2003) Atypical development of
3 5             the trachea! basement membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am.  J.
36             Physiol. 285:L931-L939.
37       Fakhrzadeh, L.; Laskin, J. D.; Laskin, D. L. (2002) Deficiency in inducible nitric oxide  synthase protects mice from
38             ozone-induced lung inflammation and tissue injury.  Am. J. Respir. Cell Mol. Biol. 26: 413-419.
39       Fanucchi, M. V.; Hotchkiss, J. A.; Harkema, J. R. (1998) Endotoxin potentiates ozone-induced mucous cell
40             metaplasia in rat nasal epithelium. Toxicol.  Appl. Pharmacol. 152: 1-9.
41       Fanucchi, M. V; Wong, V J.; Hinds, D.; Tarkington,  B. K.;  Van Winkle, L. S.; Evans,  M. J.; Plopper,  C. G. (2000)
42             Repeated episodes of exposure to ozone alters postnatal development of distal conducting airways in infant
43             rhesus monkeys.  Am. J. Respir. Crit. Care Med. 161: A615.
44       Farman, C. A.; Pinkerton, K. E.; Rajini, P.; Witschi, H.; Last, J. A. (1997) Evolution of  lung lesions in rats exposed
45             to  mixtures of ozone and nitrogen dioxide. Inhalation Toxicol. 9: 647-677'.
46       Farman, C. A.; Watkins, K.; Van Hoozen, B.; Last, J. A.; Witschi, H.; Pinkerton, K. E. (1999) Centriacinar
47             remodeling and sustained procollagen gene  expression after exposure to ozone and nitrogen dioxide.  Am. J.
48             Respir. Cell Mol. Biol. 20: 303-311.
49       Ferng, S.-F.; Castro, C. E.; Afifi, A. A.; Bermudez, E.; Mustafa, M. G. (1997) Ozone-induced DNA strand breaks in
50             guinea pig tracheobronchial epithelial cells. J. Toxicol. Environ. Health 51: 353-367.
51       Folkerts,  G.; Busse, W. W.; Nijkamp, F. P.; Sorkness, R.; Gern, J. E. (1998) Virus-induced airway
52             hyperresponsiveness and asthma. Am. J. Respir. Crit. Care Med. 157: 1708-1720.
53       Foster, W. M.; Freed, A. N. (1999) Regional clearance of solute from peripheral airway epithelia: recovery after
54             sublobar exposure to ozone. J. Appl.  Physiol. 86: 641-646.
         December 2005                                 5-73          DRAFT-DO NOT QUOTE OR CITE

-------
  1       Frampton, M. W.; Pryor, W. A.; Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999) Aldehydes (nonanal and
  2             hexanal) in rat and human bronchoalveolar lavage fluid after ozone exposure. Cambridge, MA: Health Effects
  3             Institute; research report no. 90. Available: www.healtheffects.org/Pubs/Frampton-C.pdf [2000, February 9].
  4       Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex modulates airways
  5             reactivity in rabbits. Respir. Physiol. 105: 95-102.
  6       Freed, A. N.; Cueto, R.; Pryor, W. A. (1999) Antioxidant transport modulates peripheral airway reactivity and
  7             inflammation during ozone exposure. J. Appl. Physiol. 87: 1595-1603.
  8       Friedlander, S. K.; Yeh, E. K. (1998) The submicron atmospheric aerosol as a carrier of reactive chemical species:
  9             case of peroxides. Appl. Occup. Environ. Hyg. 13: 416-420.
10       Gaffney, J. S.; Marley, N.  A.; Prestbo, E. W. (1993) Measurements of peroxyacetyl nitrate at a remote site in the
11             southwestern United States: tropospheric implications. Environ. Sci. Technol. 27: 1905-1910.
12       Garssen, J.; Van Bree, L.;  Van Der Vliet, H.; Van Loveren, H. (1997) Ozone-induced impairment of pulmonary type
13             IV hypersensitivity and airway hyperresponsiveness in mice. Inhalation Toxicol. 9: 581-599.
14       Gohil, K.; Cross, C. E.; Last, J. A. (2003) Ozone-induced disruptions of lung transcriptomes. Biochem. Biophys.
15             Res. Commun. 305: 719-728.
16       Goldsmith, C.-A. W.; Ning, Y.-Y.; Qin, G.; Imrich, A.; Lawrence, J.; Murthy, G. G., K.; Catalano, P. J.; Kobzik, L.
17             (2002) Combined air pollution particle and ozone exposure increases airway responsiveness in mice.
18             Inhalation Toxicol.  14: 325-347.
19       Grosjean, D. (2003) Ambient PAN and PPN in southern California from 1960 to the SCOS97-NARSTO. Atmos.
20             Environ. 37(suppl. 2): S221-S238.
21       Grosjean, E.; Grosjean, D.; Woodhouse, L. F. (2001) Peroxyacetyl nitrate and peroxypropionyl nitrate during SCOS
22             97-NARSTO. Environ. Sci. Technol. 35: 4007-4014.
23       Guerrero, A. L.; Dorado-Martinez, C.; Rodriguez, A.; Pedroza-Rios, K.; Borgonio-Perez, G.; Rivas-Arancibia, S.
24             (1999) Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. NeuroReport 10:
25             1689-1692.
26       Gulisano, M.; Marceddu, S.; Barbara, A.; Pacini, A.; Buiatti, E.; Martini, A.; Pacini, P. (1997) Damage to the
27             nasopharyngeal mucosa induced by current levels of urban air pollution: a field study in lambs. Eur. Respir. J.
28             10:567-572.
29       Gtinther, T.; Hollriegl, V.; Vormann, J. (1993) Perinatal development of iron and antioxidant defence systems.
30             J. Trace Elem. Electrolytes Health Dis. 7: 47-52.
31       Gupta, S. K.; Reinhart, P. G.; Bhalla, D. K. (1998) Enhancement of fibronectin expression in rat lung by ozone  and
32             an inflammatory stimulus. Am. J. Physiol. 275: L330-L335.
33       Haddad, E.-B.; Liu, S. F.;  Salmon, M.; Robichaud, A.; Barnes, P. J.; Chung, K. F. (1995) Expression of inducible
34             nitric oxide synthase mRNA in Brown Norway rats exposed to ozone: effect of dexamethasone. Eur. J.
35             Pharmacol. Environ. Toxicol. Pharmacol. Sect. 293: 287-290.
36       Haddad, E.-B.; Salmon, M.; Koto, H.; Barnes,  P. J.; Adcock, I.; Chung, K. F. (1996) Ozone induction of
3 7             cytokine-induced neutrophil chemoattractant (CINC) and nuclear factor-kappa b in rat lung: inhibition by
38             corticosteroids. FEES Lett. 379: 265-268.
39       Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential  involvement of
40             4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274:  L8-L16.
41       Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W. C. (1994) Consequences of prolonged
42             inhalation of ozone on F344/N rats: collaborative studies. Part VII: effects on the nasal mucociliary apparatus.
43             Cambridge, MA: Health Effects Institute; research report no. 65.
44       Harkema, J. R.; Catalano, P. J.; Hotchkiss, J. A. (1997a) Consequences of prolonged inhalation of ozone on F344/N
45             rats: collaborative studies. Part XII. Atrophy of bone in nasal turbinates. Cambridge, MA: Health Effects
46             Institute; research report no. 65.
47       Harkema, J. R.; Hotchkiss, J. A.; Griffith, W. C. (1997b) Mucous cell metaplasia in rat nasal epithelium after a
48             20-month exposure to ozone: a morphometric study of epithelial differentiation. Am. J. Respir. Cell Mol.
49             Biol. 16: 521-530.
50       Harkema, J. R.; Hotchkiss, J. A.; Barr, E. B.; Bennett, C. B.; Gallup, M.; Lee, J. K.; Basbaum, C. (1999)
51             Long-lasting effects of chronic ozone exposure on rat nasal epithelium. Am. J. Respir. Cell Mol. Biol.
52             20:517-529.
53       Kara, R.; Paz, C. (1993) Effects of ozone exposure  during pregnancy on ontogeny of sleep in rats. Neurosci. Lett.
54             164:67-70.
55       Hawgood, S.; Poulain, F. R. (2001) The pulmonary collectins and surfactant metabolism. Annu. Rev. Physiol.
56             63:495-519.


         December 2005                                 5-74         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Hawgood, S.; Ochs, M.; Jung, A.; Akiyama, J.; Allen, L.; Brown, C.; Edmondson, J.; Levitt, S.; Carlson, E.;
  2             Gillespie, A. M.; Villar, A.; Epstein, C. J.; Poulain, F. R. (2002) Sequential targeted deficiency of SP-A and
  3             -D leads to progressive alveolar lipoproteinosis and emphysema. Am. J. Physiol. 283: L1002-L1010.
  4       Heck, H. d'A.; Casanova, M.; Starr, T. B. (1990) Formaldehyde toxicity-new understanding. Crit. Rev. Toxicol.
  5             20:397-426.
  6       Heddle, J.  A.; Shepson, P. B.; Gingerich, J. D.; So, K. W. (1993) Mutagenicity of peroxyacetyl nitrate (PAN) in
  7             vivo: tests for somatic mutations and chromosomal aberrations. Environ. Mol. Mutagen. 21: 58-66.
  8       Herbert, R. A.; Hailey, J. R.; Grumbein, S.; Chou, B. J.; Sills, R. C.; Haseman, J. K.; Goehl, T.; Miller, R. A.;
  9             Roycroft, J. H.; Boorman, G. A. (1996) Two-year and lifetime toxicity and carcinogenicity studies of ozone in
10             B6C3F1 mice. Toxicol. Pathol. 24: 539-548.
11       Highfill, J. W.; Watkinson, W. P. (1996) Ozone toxicity in the rat. II. Modeling changes due to ambient temperatures
12             and duration. J. Appl. Physiol. 80: 1811-1818.
13       Hoffer, E.; Baum, Y.; Tabak, A.; Frevert, C. (1999) Adhesion molecules of blood polymorphonuclear leukocytes
14             and alveolar macrophages in rats: modulation by exposure to ozone. Hum.  Exp. Toxicol. 18: 547-551.
15       Holt, P. G.; Macaubas, C.; Stumbles,  P. A.; Sly, P. D. (1999) The role of allergy in the development of asthma.
16             Nature (London, U. K.) 402(suppl. 25): B12-B16.
17       Hotchkiss, J. A.; Harkema, J. R.; Johnson, N. F. (1997) Kinetics of nasal epithelial cell loss and proliferation in F344
18             rats following a single exposure to 0.5 ppm ozone. Toxicol. Appl. Pharmacol. 143: 75-82.
19       Hotchkiss, J. A.; Hilaski, R.; Cho, H.; Regan, K.;  Spencer, P.; Slack, K.; Harkema, J. R.  (1998) Fluticasone
20             propionate attenuates ozone-induced rhinitis and mucous cell metaplasia in rat nasal airway epithelium.
21             Am. J. Respir.  Cell Mol. Biol.  18: 91-99.
22       Huffman, L.  J.; Judy, D. J.; Brumbaugh, K.; Frazer, D. G.; Reynolds, J. S.; McKinney, W. G.; Goldsmith, W. T.
23             (2001) Hyperthyroidism increases the risk of ozone-induced lung toxicity in rats.  Toxicol. Appl. Pharmacol.
24             173: 18-26.
25       Huitron-Resendiz, S.; Custodio-Ramirez, V.; Escalante-Membrillo, C.; Gonzalez-Pina, R.; Paz, C. (1994) Sleep
26             alterations and brain regional changes of serotonin and its metabolite in rats exposed to ozone. Neurosci. Lett.
27             177: 119-122.
28       Hyde, D. M.; Miller, L. A.; McDonald, R. J.; Stovall, M. Y.; Wong, V.; Pinkerton, K. E.; Wegner, C. D.; Rothlein,
29             R.;  Plopper, C. G. (1999) Neutrophils enhance clearance of necrotic epithelial cells in ozone-induced lung
30             injury in rhesus monkeys. Am. J. Physiol. 277: LI 190-L1198.
31       Igarashi, A.;  lijima, H.; Tamura,  G.; Shirato, K. (1998) Tazanolast inhibits ozone-induced airway
32             hyperresponsiveness in guinea pigs. Am. J. Respir. Crit. Care Med.  157: 1531-1535.
33       lijima, M.  K.; Kobayashi, T.; Kamada, H.; Shimojo, N. (2001) Exposure to ozone aggravates nasal allergy-like
34             symptoms in guinea pigs.  Toxicol. Lett. 123: 77-85.
35       Inoue, H.;  Aizawa, H.; Nakano, H.; Matsumoto, K.; Kuwano, K.; Nadel, J. A.; Kara, N.  (2000) Nitric oxide synthase
36             inhibitors attenuate ozone-induced airway inflammation in guinea pigs: possible role of interleukin-8. Am. J.
37             Respir. Crit.  Care Med. 161: 249-256.
38       Ishii, Y.; Yang, H.; Sakamoto, T.; Nomura, A.; Hasegawa, S.; Hirata, F.; Bassett,  D. J. P. (1997) Rat alveolar
3 9             macrophage cytokine production and regulation of neutrophil recruitment following acute ozone exposure.
40             Toxicol. Appl. Pharmacol. 147: 214-223.
41       Ishii, Y.; Hirano, K.; Morishima, Y.; Masuyama, K.;  Goto, Y.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Sagai, M.;
42             Sekizawa, K. (2000) Early molecular and cellular events of oxidant-induced pulmonary fibrosis in rats.
43             Toxicol. Appl. Pharmacol. 167: 173-181.
44       Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
45             conscious rats. Ind. Health 36: 57-60.
46       Jakobi, G.; Fabian, P. (1997) Indoor/outdoor concentrations of ozone and peroxyacetyl nitrate (PAN). Int. J.
47             Biometeorol. 40: 162-165.
48       Jang, A.-S.; Choi, I.-S.; Koh, Y.-I.; Park, C.-S.; Lee, J.-S. (2002) The relationship between alveolar epithelial
49             proliferation and airway obstruction after ozone exposure. Allergy 57: 737-740.
50       Jimba, M.; Skornik, W. A.; Killingsworth, C. R.; Long, N. C.; Brain, J. D.; Shore, S. A.  (1995) Role of C fibers in
51             physiological responses to ozone in rats. J.  Appl. Physiol. 78: 1757-1763.
52       Joad, J. P.; Kott, K. S.; Bonham, A. C. (1998) Exposing guinea pigs to ozone for 1 wk enhances responsiveness of
53             rapidly adapting receptors. J. Appl. Physiol. 84: 1190-1197.
54       Joad, J. P.; Brie, J. M.; Weir, A. J.; Putney, L.; Hyde, D.  M.; Postlewait, E. M.; Plopper, C.  G. (2000) Effect of
55             respiratory pattern on ozone injury to the airways of isolated rat lungs. Toxicol. Appl. Pharmacol. 169: 26-32.
         December 2005                                 5-75          DRAFT-DO NOT QUOTE OR CITE

-------
  1       Johnston, C. I; Stripp, B. R.; Reynolds, S. D.; Avissar, N. E.; Reed, C. K.; Finkelstein, J. N. (1999) Inflammatory
  2             and antioxidant gene expression in C57BL/6J mice after lethal and sublethal ozone exposures. Exp. Lung Res.
  3             25:81-97.
  4       Johnston, C. J.; Reed, C. K.; Avissar, N. E.; Gelein, R.; Finkelstein, J. N. (2000a) Antioxidant and inflammatory
  5             response after acute nitrogen dioxide and ozone exposures in C57B1/6 mice. Inhalation Toxicol. 12: 187-203.
  6       Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2000b) Newborn mice differ from adult mice in
  7             chemokine and cytokine expression to ozone, but not to endotoxin. Inhalation Toxicol.  12: 205-224.
  8       Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2002) Endotoxin potentiates ozone-induced
  9             pulmonary chemokine and inflammatory responses. Exp. Lung Res. 28: 419-433.
10       Kafoury, R. M; Pryor, W. A.;  Squadrito, G. L.; Salgo, M. G.; Zou, X.; Friedman, M. (1999) Induction of
11             inflammatory mediators in human airway epithelial cells by lipid ozonation products. Am. J. Respir. Crit.
12             Care Med. 160: 1934-1942.
13       Kenyon, N. J.; Van Der Vliet, A.; Schock, B.C.; Okamoto, T.; McGrew, G. M.; Last, J. A. (2002) Susceptibility to
14             ozone-induced acute lung injury in iNOS-deficient mice. Am. J. Physiol. 282: L540-L545.
15       Kim, M. Y.; Son, J. W.; Cho, M. H.; Choi,  C. S.; Chae, C. H.; Lee, M. H. (2001) Oviductural carcinoma in B6C3F1
16             female mice exposed to 0.5 ppm ozone. Vet. Hum. Toxicol. 43: 370-372.
17       Kimmel, T. A.; Chen, L. C.; Bosland, M. C.; Nadziejko, C. (1997) Influence of acid aerosol droplet size on
18             structural changes in the rat lung caused by acute exposure to sulfuric acid and ozone. Toxicol. Appl.
19             Pharmacol. 144: 348-355.
20       Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y.; Longphre, M.; Harkema, J.; Jedlicka, A.; Eleff, S. M.; DiSilvestre, D.;
21             Holroyd, K. J. (1997) Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
22             Nat. Genet. 17: 475-478.
23       Kleeberger, S. R.; Reddy, S.; Zhang, L.-Y.; Jedlicka, A. E. (2000) Genetic susceptibility to ozone-induced lung
24             hyperpermeability: role of toll-like receptor 4. Am. J. Respir. Cell Mol. Biol. 22: 620-627.
25       Kleeberger, S. R.; Reddy, S. P.; Zhang, L.-Y.; Cho, H.-Y.; Jedlicka, A. E. (2001) Toll-like receptor 4 mediates
26             ozone-induced murine lung hyperpermeability via inducible nitric oxide synthase. Am.  J. Physiol.
27             280: L326-L333.
28       Kleindienst,  T. E.; Shepson, P. B.; Smith, D. F.; Hudgens, E. E.; Nero, C. M.; Cupitt, L. T.; Bufalini, J. J.; Claxton,
29             L. D.  (1990) Comparison of mutagenic activities of several peroxyacyl nitrates. Environ. Mol. Mutagen.
30             16:70-80.
31       Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
32             alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
33       Kleinman, M. T.; Bufalino, C.; Rasmussen, R.; Hyde, D.; Bhalla, D. K.; Mautz, W. J. (2000) Toxicity of chemical
34             components of ambient fine paniculate matter (PM2 5) inhaled by aged rats. J. Appl. Toxicol.  20: 357-364.
35       Kligerman, A. D.; Mottus, K.; Erexson, G. L. (1995) Cytogenetic analyses of the in vitro and in vivo responses of
36             murine cells to peroxyacetyl nitrate (PAN). Mutat. Res. 341: 199-206.
37       Kobzik, L.; Goldsmith, C.-A. W.; Ning, Y. Y.; Qin, G.; Morgan, B.; Imrich, A.; Lawrence J.; Murthy, G.  G. K.;
3 8             Catalano, P.  J. (2001) Effects of combined ozone and air pollution particle exposure in mice.  Boston, MA:
39             Health Effects Institute; research report no. 106. Available: http://www.healtheffects.org/Pubs/Kobzik.pdf
40             [27 January 2003].
41       Kodavanti, U. P.; Hatch, G. E.; Starcher, B.; Giri, S. N.; Winsett, D.;  Costa, D. L. (1995) Ozone-induced pulmonary
42             functional, pathological, and biochemical changes in normal and vitamin C-deficient guinea pigs. Fundam.
43             Appl. Toxicol.  24: 154-164.
44       Koto, H.; Aizawa, H.; Takata, S.; Inoue, H.; Kara, N. (1995) An important role of tachykinins in ozone-induced
45             airway hyperresponsiveness. Am.  J.  Respir. Crit. Care Med. 151: 1763-1769.
46       Koto, H.; Salmon, M.; Haddad el-B.; Huang,  T.-J.; Zagorski, J.; Chung, K. F. (1997) Role of cytokine-induced
47             neutrophil chemoattractant (CINC) in ozone-induced airway inflammation and hyperresponsiveness. Am. J.
48             Respir. Crit.  Care Med.  156: 234-239.
49       Koyama, Y.; Hayaishi, O. (1994) Modulation by prostaglandins of activity of sleep-related neurons in the
50             preoptic/anterior hypothalamic areas in rats. Brain Res. Bull. 33: 367-372.
51       Kudo, M.; Nishikawa, M.; Ikeda, H.; Okubo, T. (1996) Involvement of superoxide anions in ozone-induced airway
52             hyperresponsiveness in unanesthetized guinea pigs. Environ. Toxicol. Pharmacol. 2: 25-30.
53       Larson,  S. D.; Schelegle, E. S.; Walby, W. F.; Gershwin, L.  J.; Fanuccihi, M. V; Evans, M. J.; Joad, J. P.;
54             Tarkington, B.  K.; Hyde, D. M.; Plopper, C. G. (2004) Postnatal remodeling of the neural components of the
5 5             epithelial-mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and
56             allergen. Toxicol. Appl. Pharmacol.  194: 211-220.


         December 2005                                 5-76         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Laskin, D. L.; Laskin, J. D. (2001) Role of macrophages and inflammatory mediators in chemically induced toxicity.
  2             Toxicology (Ireland) 160: 111-118.
  3       Laskin, D. L.; Pendino, K. J.; Punjabi, C. J.; del Valle, M. R.; Laskin, J. D. (1994) Pulmonary and hepatic effects of
  4             inhaled ozone in rats. Environ. Health Perspect. 102(suppl. 10): 61-64.
  5       Laskin, J. D.; Heck, D. E.; Laskin, D. L. (1996) Nitric oxide production in the lung and liver following inhalation of
  6             the pulmonary irritant ozone. In: Snyder, R.; Kocsis, J. J.; Sipes, I. G.; Kalf, G. F.; Jollow, D. J.; Greim, H.;
  7             Monks, T. J.; Witmer, C. M., eds. Biological Reactive Intermediates V: Basic Mechanistic Research in
  8             Toxicology and Human Risk Assessment: proceedings of the Fifth International Symposium; January 1995;
  9             Munich, Germany. Adv. Exp. Med. Biol. 387: 141-146.
10       Laskin, D. L.; Sunil, V.; Guo, Y.; Heck, D. E.; Laskin, J. D. (1998) Increased nitric oxide synthase in the lung after
11             ozone inhalation is associated with activation of NF-KB. Environ. Health Perspect. 106(suppl. 5): 1175-1178.
12       Laskin, D. L.; Fakhrzadeh, L.; Heck, D. E.; Gerecke, D.; Laskin, J. D. (2002) Upregulation of phosphoinositide
13             3 -kinase and protein kinase B in alveolar macrophages following ozone inhalation. Role of NF-"kappa"B and
14             STAT-1 in ozone-induced nitric oxide production and toxicity. Mol. Cell. Biochem. 234-235: 91-98.
15       Last, J. A.; Pinkerton, K. E. (1997) Chronic exposure of rats to ozone and sulfuric acid aerosol: biochemical and
16             structural responses. Toxicology 116: 133-146.
17       Lavnikova, N.; Prokhorova, S.; Lakhotia, A. V.; Gordon, R.; Laskin, D. L. (1998) Distinct inflammatory responses
18             of adherent vascular lung neutrophils to pulmonary irritants. J. Inflammation 48: 56-66
19       Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
20             differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
21             and induction in rats. Drug Metab. Dispos. 26: 396-400.
22       Lemos, M.; Lichtenfels, A. J. F. C.; Amaro, E., Jr.; Macchione, M.; Martins, M. A.; King, M.; Bohm, G. M.;
23             Saldiva, P. H. N. (1994) Quantitative pathology of nasal passages in rats exposed to urban levels of air
24             pollution. Environ. Res. 66: 87-95.
25       Long, N. C.; Suh, J.; Morrow, J. D.; Schiestl, R. H.; Krishna Murthy, G. G.; Brain, J. D.; Frei, B. (2001) Ozone
26             causes lipid peroxidation but little antioxidant depletion in exercising and nonexercising hamsters. J. Appl.
27             Physiol. 91:  1694-1700.
28       Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1996) Mast cells contribute to O3-induced epithelial
29             damage and  proliferation in nasal and bronchial airways of mice. J. Appl. Physiol. 80: 1322-1330.
30       Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1999) Ozone-induced pulmonary  inflammation and
31             epithelial proliferation are partially mediated by PAF. J.  Appl. Physiol. 86: 341-349.
32       Lorz, C.; Lopez, J.  (1997) Incidence of air pollution in the pulmonary surfactant system of the pigeon (Columbia
33             livid). Anat.  Rec. 249: 206-212.
34       Madden, M.  C.; Richards, J. H.; Dailey, L. A.; Hatch, G. E.; Ohio, A. J. (2000) Effect of ozone on diesel exhaust
35             particle toxicity in rat lung. Toxicol. Appl. Pharmacol. 168: 140-148.
36       Mango,  G. W.; Johnston, C. J.; Reynolds, S. D.; Finkelstein, J.  N.; Plopper, C. G.; Stripp, B. R. (1998) Clara cell
37             secretory protein deficiency increases oxidant stress response in conducting airways. Am. J. Physiol.
38             275: L348-L356.
39       Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Shimazu, N.; Iwasaki, H.; Ikezawa, K. (1995) Importance
40             of impairment of the airway epithelium for ozone-induced airway hyperresponsiveness in guinea pigs. Jpn. J.
41             Pharmacol. 67: 375-382.
42       Matsubara, S.; Kikkawa, H.; Kaminuma, O.; Ikezawa, K. (1997a) Angiotensin-converting enzyme inhibitors can
43             potentiate  ozone-induced airway hyperresponsiveness. Eur. J. Pharmacol. 337: 259-265.
44       Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Ikezawa, K.; Naito, K.  (1997b) Prevention of
45             ozone-induced airway hyperresponsiveness and epithelial injury by phosphodiesterase inhibitors in guinea
46             pigs. Environ. Toxicol. Pharmacol. 3: 201-209.
47       Matsumoto, K.; Aizawa, H.; Inoue, H.; Koto, H.; Nakano, H.; Kara, N. (1999) Role of neutrophil elastase in
48             ozone-induced airway responses in guinea-pigs. Eur. Respir. J. 14: 1088-1094.
49       Mautz, W. J. (2003) Exercising animal models in inhalation toxicology: interactions with ozone and formaldehyde.
50             Environ. Res. 92: 14-26.
51       Mautz, W. J.; Kleinman, M. T.; Bhalla, D. K.; Phalen, R. F. (2001) Respiratory tract responses to repeated inhalation
52             of an oxidant and acid gas-particle air pollutant mixture. Toxicol. Sci. 61: 331-341.
53       McGraw, D. W.; Forbes, S. L.; Mak, J. C. W.; Witte, D. P.; Carrigan, P. E.; Leikauf, G. D.; Liggett, S. B. (2000)
54             Transgenic overexpression of beta(2)-adrenergic receptors in airway epithelial cells decreases
55             bronchoconstriction. Am. J. Physiol. 279: L379-L389.
         December 2005                                 5-77          DRAFT-DO NOT QUOTE OR CITE

-------
  1       McKinney, W. I; Jaskot, R. H.; Richards, J. H.; Costa, D. L.; Dreher, K. L. (1998) Cytokine mediation of
  2             ozone-induced pulmonary adaptation. Am. J. Respir. Cell. Mol. Biol. 18: 696-705.
  3       Miller, L. A.; Barnett, N. L.; Sheppard, D.; Hyde, D. M. (2001) Expression of the P6 integrin subunit is associated
  4             with sites of neutrophil influx in lung epithelium. J. Histochem. Cytochem. 49: 41-48.
  5       Morgan, M. S.; Meyer, P.; Holub, R.; Frank, R. (1986) Overall and regional lung function in dogs exposed acutely to
  6             ozone. Environ. Res. 41: 546-557.
  7       Morio, L. A.; Hooper, K. A.; Brittingham, J.; Li, T.-H.; Gordon, R. E.; Turpin, B. J.; Laskin, D. L. (2001) Tissue
  8             injury following inhalation of fine paniculate matter and hydrogen perioxide is associated with altered
  9             production of inflammatory mediators and antioxidants by alveolar macrophages. Toxicol. Appl. Pharmacol.
10             177: 188-199.
11       Moss, O. R.; Gross, E. A.; James, R. A.; Janszen, D. B.; Ross, P. W.; Roberts, K. C.; Howard, A. M.; Harkema,
12             J. R.; Calderon-Garciduenas, L.; Morgan, K. T. (2001) Respiratory tract toxicity in rats exposed to Mexico
13             City air. Cambridge, MA: Health Effects Institute; research report no. 100. Available:
14             http://www.healtheffects.org/pubs-research.htm [15 May, 2003].
15       Mucke, W. (1996) The environment and the eye. Topics of ophthalmic toxicology. Leban. Med. J. 44: 146-150.
16       Murphy, D. J. (2002) Assessment of respiratory function in safety pharmacology. Fundam. Clin. Pharmacol.
17             16:  183-196.
18       Mutoh, T.; Joad, J. P.; Bonham, A. C. (2000) Chronic passive cigarette smoke exposure augments
19             bronchopulmonary C-fibre inputs to nucleus tractus solitarii neurones and reflex output in young guinea-pigs.
20             J. Physiol. (London) 523: 223-23 3.
21       Nakano, H.; Aizawa, H.; Matsumoto, K.; Fukuyama, S.; Inoue, H.; Kara, N. (2000) Cyclooxygenase-2 participates
22             in the late phase of airway hyperresponsiveness after ozone exposure in guinea pigs. Eur. J. Pharmacol.
23             403:267-275.
24       National Toxicology Program. (1994) NTP technical report on the toxicology and carcinogenesis studies of ozone
25             (CAS no. 10028-15-6) and ozone/NNK (CAS no.  10028-15-6/64091-91-4) in F344/N rats and B6C3FJ mice
26             (inhalation studies). Research Triangle Park, NC: U.S. Department of Health and Human Services, National
27             Institutes of Health; publication no. 95-3371. (National Toxicology Program technical report series: no. 440).
28       Neuhaus-Steinmetz, U.; Uffhausen, F.; Herz, U.; Renz, H. (2000) Priming of allergic immune responses by repeated
29             ozone exposure in mice. Am.  J. Respir.  Cell Mol. Biol. 23: 228-233.
30       Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
31             factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
32             60:356-362.
33       Nielsen, G. D.; Hougaard, K. S.; Larsen,  S. T.; Hammer, M.; Wolkoff, P.; Clausen, P. A.; Wilkins, C. K.; Alarie, Y.
34             (1999) Acute airway effects of formaldehyde and ozone inBALB/c mice. Hum. Exp. Toxicol. 18: 400-409.
35       Nino-Cabrera, H. G.; Colin-Barenque, L.; Avila-Costa, M. R.; Espinosa-Villanueva, J.; Fortoul, T. I.;
36             Rivas-Arancibia, S. (2002) Differences between hippocampus and cerebral cortex in aged rats in an oxidative
37             stress model. Int. J. Neurosci. 112: 373-381.
38       Noviski, N.; Brewer, J. P.; Skornik, W. A.; Galli, S. J.; Drazen, J. M.; Martin, T. R. (1999) Mast cell activation is
39             not required for induction of airway hyperresponsiveness by ozone in mice. J. Appl. Physiol. 86: 202-210.
40       O'Connor, G. T.;  Sparrow, D.; Weiss, S. T. (1995) A prospective longitudinal study of methacholine airway
41             responsiveness as a predictor of pulmonary-function decline: the Normative Aging Study. Am. J. Respir. Crit.
42             Care Med. 152: 87-92.
43       Paige, R. C.; Wong, V.; Plopper, C. G. (2000) Long-term exposure to ozone increases acute pulmonary centriacinar
44             injury by 1-nitronaphthalene:  II. Quantitative histopathology. J. Pharmacol. Exp. Ther. 295: 942-950.
45       Palmer, L. J.; Burton,  P. R.; Faux, J. A. (2000) Independent inheritance of serum immunoglobulin E concentrations
46             and airway responsiveness. Am. J. Respir. Crit. Care Med. 161: 1836-1842.
47       Palmer, L. J.; Rye, P. J.; Gibson, N. A. (2001) Airway responsiveness in early infancy predicts asthma, lung
48             function, and respiratory symptoms by school age. Am. J. Respir. Crit. Care Med. 163: 37-42.
49       Paquette, N. C.; Tankersley, C. G.; Zhang, L.-Y.; Kleeberger, S. R. (1994) Repeated subacute ozone exposure of
50             inbred mice: airway inflammation and ventilation. Exp. Lung Res. 20: 579-594.
51       Paz, C.; Bazan-Perkins, B. (1992) Sleep-wake  disorganization in cats exposed to ozone. Neurosci. Lett.
52             140:270-272.
53       Paz, C.; Huitron-Resendiz, S. (1996) The effects of ozone exposure on the  sleep-wake cycle and serotonin contents
54             in the pons of the rat. Neurosci. Lett. 204: 49-52.
55       Paz, C. (1997) Some consequences of ozone exposure on health. Arch. Med. Res. 28: 163-170.
         December 2005                                5-78         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Pearson, A. C.; Bhalla, D. K. (1997) Effects of ozone onmacrophage adhesion in vitro and epithelial and
  2             inflammatory responses in vivo: the role of cytokines. J. Toxicol. Environ. Health 50: 143-157.
  3       Peden, D. B.; Dailey, L. (1995) Modulation of mast cell functions by in vitro ozone exposure. Am. J. Physiol.
  4             268: L902-L910.
  5       Pendino, K. J.; Shuler, R. L.; Laskin, J. D.; Laskin, D. L. (1994) Enhanced production of interleukin-1, tumor
  6             necrosis factor-"alpha", and fibronectin by rat lung phagocytes following inhalation of a pulmonary irritant.
  7             Am. J. Respir. Cell Mol. Biol. 11: 279-286.
  8       Pendino, K. J.; Meidhof, T. M.; Heck, D. E.; Laskin, J. D.; Laskin, D. L. (1995) Inhibition of macrophages with
  9             gadolinium chloride abrogates ozone-induced pulmonary injury and inflammatory mediator production. Am.
10             J. Respir. Cell Mol. Biol. 13: 125-132.
11       Pendino, K. J.; Gardner, C. R.; Shuler, R. L.; Laskin, J. D.; Durham, S. K.; Barton, D. S.; Ohnishi, S. T.; Ohnishi, T.;
12             Laskin, D. L. (1996) Inhibition of ozone-induced nitric oxide  synthase expression in the lung by endotoxin.
13             Am. J. Respir. Cell Mol. Biol. 14: 516-525.
14       Petruzzi, S.; Fiore, M.; Dell'Omo, G.; Bignami, G.; Alleva, E. (1995) Medium and long-term behavioral effects in
15             mice of extended gestational exposure to ozone. Neurotoxicol. Teratol. 17: 463-470.
16       Petruzzi, S.; De Acetis, L.; Chiarotti, F.; Sorace, A.; Alleva, E. (1999) Limited changes in handedness and morphine
17             reactivity in CD-I mice after pre- and postnatal ozone exposure. Acta Neurobiol. Exp. 59: 115-122.
18       Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
19             ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
20             epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
21             research report no. 65.
22       Plopper, C. G.; Fanucchi, M. V. (2000) Do urban environmental pollutants exacerbate childhood lung diseases?
23             Environ. Health Perspect. 108: A252-A253.
24       Plopper, C. G.; Hatch, G. E.; Wong, V.; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.;
25             Buckpitt, A. R. (1998) Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury,
26             site-specific ozone dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
27             19:387-399.
28       Polosa, R.; Holgate, S. T. (1997) Adenosine bronchoprovocation: a promising marker of allergic inflammation in
29             asthma? Thorax 52: 919-923.
30       Postlethwait, E. M.; Cueto, R.; Velsor, L. W.;  Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
31             estimated surface concentrations and lining layer effects. Am. J. Physiol.  274: L1006-L1016.
32       Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie,  J. M.; Weir, A. J.; Putney, L. F.; Wong, V. J.;
33             Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
34             tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
35       Prows, D. R.; Shertzer, H. G.; Daly, M. J.;  Sidman, C. L.; Leikauf, G. D. (1997) Genetic analysis of ozone-induced
36             acute lung injury in sensitive and resistant strains of mice. Nat. Genet. 17: 471-474.
37       Prows, D. R.; Daly, M. J.; Shertzer, H. G.;  Leikauf,  G. D. (1999) Ozone-induced acute lung injury: genetic analysis
38             of F2 mice generated from A/J and C57BL/6J strains. Am. J. Physiol. 277: L372-L380.
39       Pryor, W. A.; Squadrito, G. L.; Friedman, M. (1995) A new mechanism for the toxicity of ozone. Toxicol. Lett.
40             82/83:287-293.
41       Pryor, W. A.; Bermudez, E.; Cueto, R.; Squadrito, G. L. (1996) Detection of aldehydes in bronchoalveolar lavage of
42             rats exposed to ozone. Fundam. Appl. Toxicol. 34: 148-156.
43       Rehle, D.; Leleux, D.; Erdelyi, M.; Tittel, F.; Fraser, M.; Friedfeld, S.; et al. (2001) Ambient formaldehyde detection
44             with a laser spectrometer based on difference-frequency generation in PPLN.  Appl. Phys. B: Lasers Opt.
45             72: 947-952.
46       Reinhart, P. G.; Gupta, S. K.; Bhalla, D. K. (1999) Attenuation of ozone-induced lung injury by interleukin-10.
47             Toxicol. Lett. 110: 35-42.
48       Renner, R. (2002) Bad air and birth defects. Environ. Health Perspect. 110: A291.
49       Ritz, B.; Yu, F. (1999) The effect of ambient carbon monoxide on low birth weight among children born in southern
50             California between 1989 and 1993. Environ. Health Perspect. 107:  17-25.
51       Ritz, B.; Yu, F.; Fruin, S.; Chapa, G.; Shaw, G. M.; Harris, J. A. (2002) Ambient air pollution and risk of birth
52             defects in Southern California. Am.  J. Epidemiol. 155:  17-25.
53       Rivas-Arancibia, S.; Vazquez-Sandoval, R.; Gonzalez-Kladiano, D.; Schneider-Rivas, S.; Lechuga-Guerrero, A.
54             (1998) Effects of ozone exposure in rats on memory and levels of brain and pulmonary superoxide dismutase.
55             Environ. Res. 76: 33-39.
         December 2005                                 5-79         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Rivas-Arancibia, S.; Dorado-Martinez, C.; Borgonio-Perez, G.; Hiriart-Urdanivia, M.; Verdugo-Diaz, L.;
  2             Duran-Vazquez, A.; Colin-Baranque, L.; Avila-Costa, M. R. (2000) Effects of taurine on ozone-induced
  3             memory deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ. Res. 82: 7-17.
  4       Rivas-Arancibia, S.; Dorado-Martinez, C.; Colin-Barenque, L.; Kendrick, K. M.;  De la Riva, C.; Guevara-Guzman,
  5             R. (2003) Effect of acute ozone exposure on locomotor behavior and striatal function. Pharmacol. Biochem.
  6             Behav. 74: 891-900.
  7       Rivas-Manzano, P.; Paz, C. (1999) Cerebellar morphological alterations in rats induced by prenatal ozone exposure.
  8             Neurosci. Lett. 276: 37-40.
  9       Rose, R. C.; Richer, S. P.; Bode, A. M. (1998) Ocular oxidants and antioxidant protection. Proc. Soc. Exp. Biol.
10             Med. 217: 397-407.
11       Saldiva, P. H. N.; King, M.; Delmonte, V. L. C.; Macchione, M.; Parada, M. A. C.; Daliberto, M. L.; Sakae, R. S.;
12             Criado, P. M. P.; Silveira, P. L. P.; Zin, W. A.; Bohm, G. M. (1992) Respiratory alterations due to urban air
13             pollution: an experimental study in rats. Environ. Res. 57: 19-33.
14       Savov, J. D.; Whitehead, G. S.; Wang, I; Liao, G.; Usuka, I; Peltz, G.; Foster, W. M.; Schwartz, D. A. (2004)
15             Ozone-induced acute pulmonary injury in inbred mouse strains. Am. J. Respir. Cell Mol. Biol. 31: 69-77'.
16       Schelegle, E. S.; Alfaro, M. F.; Putney, L.; Stovall, M.; Tyler, N.; Hyde, D. M. (2001) Effect of C-fiber-mediated,
17             ozone-induced rapid shallow breathing on airway epithelial injury in rats. J. Appl. Physiol. 91: 1611-1618.
18       Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V.; Van Winkle, L.  S.; Gerriets, J. E.; Walby, W. F.;
19             Mitchell, V; Tarkington, B. K.; Wong, V J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu,
20             R.; Evans, M. J.; Hyde, D. M.; Plopper, C.  G. (2003a) Repeated episodes of ozone inhalation amplifies the
21             effects of allergen sensitization and inhalation on airway immune and structural development in Rhesus
22             monkeys. Toxicol. Appl. Pharmacol.  191: 74-85.
23       Schelegle, E. S.; Walby, W. F.; Alfaro, M. F.; Wong, V. J.; Putney, L.; Stovall, M. Y.;  Sterner-Kock, A.; Hyde,
24             D. M.; Plopper, C. G. (2003b) Repeated episodes of ozone inhalation attenuates airway injury/repair and
25             release of substance P, but not adaptation. Toxicol. Appl. Pharmacol. 186:  127-142.
26       Schlesinger, R. B. (1995) Interaction of gaseous and paniculate pollutants in the respiratory tract: mechanisms and
27             modulators. Toxicology 105: 315-325.
28       Schlesinger, R. B.; Cohen, M. D.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
29             (2002a) Ozone differentially modulates airway responsiveness in atopic versus nonatopic guinea pigs.
30             Inhalation Toxicol. 14:431-457.
31       Schlesinger, R. B.; Cohen, M.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
32             (2002b) Ozone-induced modulation of airway hyperresponsiveness in guinea pigs. Boston, MA: Health
33             Effects Institute; research report no. 109.
34       Schwartz, D. A. (2002) TLR4 and LPS hyporesponsiveness in humans. Int. J. Hyg. Environ. Health 205:  221-227.
35       Sears, M. R.; Jones, D. T.; Holdaway, M. D.; Hewitt, C. J.; Flannery, E. M.; Herbison, G. P.; Silva, P. A. (1986)
36             Prevalence of bronchial reactivity to inhaled methacholine in New Zealand children. Thorax 41: 283-289.
37       Segura, P.; Montano, L. M.; Bazan-Perkins,  B.; Gustin, P.; Vargas, M. H. (1997)  Ozone at high-pollution urban
3 8             levels causes airway hyperresponsiveness to substance P but not to other agonists. Environ. Toxicol.
39             Pharmacol. 3: 91-95.
40       Sen, S.; Dulchavsky, S. A.; Dutta, S. (1993)  Effects of triiodothyronine (T3) supplementation upon ozone-induced
41             lung injury. Free Radic. Res. Commun. 18: 299-308.
42       Sherwin, R. P.; Richters, V.; Kraft, P.; Richters, A. (2000) Centriacinar region inflammatory disease in young
43             individuals: a comparative study of Miami and Los Angeles residents. Virchows Arch. 437: 422-428.
44       Shore, S. A.; Abraham, J. H.; Schwartzman, I. N.; Murthy, G. G.; Laporte, J. D. (2000) Ventilatory responses to
45             ozone are reduced in immature rats. J. Appl. Physiol. 88: 2023-2030.
46       Shore, S. A.; Schwartzman, I. N.; Le Blanc,  B.; Krishna Murthy, G. G.; Doerschuk, C. M. (2001) Tumor necrosis
47             factor receptor 2 contributes to ozone-induced airway hyperresponsiveness in mice. Am. J. Respir. Crit. Care
48             Med. 164: 602-607.
49       Shore, S. A.; Johnston, R. A.; Schwartzman, I. N.; Chism, D.; Krishna Murthy, G. G. (2002) Ozone-induced airway
50             hyperresponsiveness is reduced in immature mice.  J. Appl. Physiol. 92: 1019-1028.
51       Shore, S. A.; Rivera-Sanchez, Y. M.; Schwartzman, I. N.; Johnston, R. A. (2003) Responses to ozone are increased
52             in obese mice. J. Appl. Physiol. 95: 938-945.
53       Sindhu, R. K.; Mautz, W. J.; Kikkawa, Y. (1998) Chronic exposure to ozone and  nitric acid vapor results in
54             increased levels of rat pulmonary putrescine. Arch. Toxicol. 72: 445-449.
55       Slade, R.; Watkinson, W. P.; Hatch, G. E. (1997) Mouse strain differences in ozone dosimetry and body temperature
56             changes.  Am. J. Physiol. 272: L73-L77.


         December 2005                                 5-80         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Sommer, B.; Montano, L. M; Chavez, J.; Gustin, P.; Vargas, M. H. (1998) Guinea pig lung resistance shows
  2             circadian rhythmicity not influenced by ozone. Respir. Physiol. 113: 223-229.
  3       Sommer, B.; Vargas, M. H.; Chavez, J.; Carbajal, V; Segura, P.; Montano, L. M. (2001) Differences between
  4             inhaled and intravenous bronchial challenge to detect O3-induced hyperresponsiveness. J. Appl. Physiol.
  5             91:2595-2601.
  6       Sorace, A.; De Acetis, L.; Alleva, E.; Santucci, D. (2001) Prolonged exposure to low doses of ozone: short- and
  7             long-term changes to behavioral performance in mice. Environ. Res. 85: 122-134.
  8       Spannhake, E. W. (1996) Down-regulation of canine airway mast cell function following exposure to ozone in vivo.
  9             Exp. Lung Res. 22: 163-178.
10       Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar to
11             nonhuman primates. Am. J. Respir. Crit. Care Med.  162: 1152-1156.
12       Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M. J.; Ellis, R.; Dahlback, M.; O'Byrne, P. M. (1994) Effect of
13             inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs.
14             J. Appl. Physiol.  77: 2578-2583.
15       Stevens, W. H. M.; VanderHeyden, C.; Wattie, J.; Lane, C. G.; Smith, W.; O'Byrne, P. M. (1995a) Effect of a
16             leukotriene B4 receptor antagonist SC-53228 on ozone-induced airway hyperresponsiveness and inflammation
17             in dogs. Am. J. Respir. Crit.  Care Med. 152: 1443-1448.
18       Stevens, W. H. M.; Conlon, P. D.; O'Byrne, P. M. (1995b)  Ozone-induced oxygen radical release from
19             bronchoalveolar lavage cells and airway hyper-responsiveness in dogs. J. Physiol. (London) 486: 257-265.
20       Stick, S. M. (2002) Pulmonary physiology, airway responsiveness and asthma. Med. J. Aust. 177: S55-S56.
21       Stick, S. M.; Turnbull, S.; Chua, H. L.; Landau, L. L; LeSouef, P. N. (1990) Bronchial responsiveness to histamine
22             in infants and older children. Am. Rev. Respir. Dis.  142: 1143-1146.
23       Sun, J.; Chung, K. F. (1997) Interaction of ozone exposure with airway hyperresponsiveness and inflammation
24             induced by trimellitic anhydride in sensitized guinea pigs. J. Toxicol. Environ. Health 51: 77-87.
25       Sun, J.; Koto, H.; Chung, K. F. (1997) Interaction of ozone and allergen challenges  on bronchial responsiveness and
26             inflammation in sensitised guinea pigs. Int. Arch. Allergy Immunol. 112: 191-195.
27       Szarek, J. L.;  Stewart, N. L.; Zhang, J. Z.; Webb, J. A.; Valentovic, M. A.; Catalano, P. (1995) Contractile responses
28             and structure of small bronchi isolated from rats after 20 months' exposure to ozone. Fundam. Appl. Toxicol.
29             28: 199-208.
30       Tager, I. B. (1999) Air pollution and lung function growth. Is it ozone? [editorial]. Am. J. Respir. Crit. Care Med.
31             160:387-389.
32       Takahashi, T.; Miura, M.; Katsumata, U.; Ichinose, M.; Kimura, K.; Inoue, H.; Takishima, T.; Shirato, K. (1993)
33             Involvement of superoxide in ozone-induced airway hyperresponsiveness in anesthetized cats. Am. Rev.
34             Respir. Dis.  148: 103-106.
35       Takahashi, N.; Yu, X.-Y.; Schofield, B. H.; Kleeberger,  S.  R.; Scott, A. L.; Hasegawa, S.;  Spannhake, E. W. (1995a)
3 6             Expression of ICAM-1 in airway epithelium after acute ozone exposure in the mouse. J. Appl. Physiol.
37             79: 1753-1761.
38       Takahashi, M.; Kleeberger, S. R.; Croxton, T. L. (1995b) Genetic control of susceptibility  to ozone-induced changes
39             in mouse trachea! electrophysiology. Am. J. Physiol. 269: L6-L10.
40       Takata, S.; Aizawa, H.;  Inoue, H.; Koto, H.; Kara, N. (1995) Ozone exposure suppresses epithelium-dependent
41             relaxation in feline airway. Lung 173: 47-56.
42       Takebayashi,  T.; Abraham, J.; Murthy, G. G.  K.; Lilly, C.;  Rodger, I.; Shore, S. A. (1998)  Role of tachykinins in
43             airway responses to ozone in rats. J. Appl. Physiol. 85: 442-450.
44       Tankersley, C. G.; Kleeberger, S. R. (1994) Ozone-induced inflammation and altered ventilation in genetically
45             susceptible mice: a comparison of acute and subacute exposures. Toxicol Lett. 72: 279-289.
46       Tankersley, C. G.; Fitzgerald, R.  S.; Mitzner, W. A.; Kleeberger, S. R.  (1993) Hypercapnic ventilatory responses in
47             mice differentially susceptible to acute ozone exposure. J. Appl. Physiol. 75:  2613-2619.
48       Tesfaigzi, J.; Hotchkiss, J. A.; Harkema, J. R. (1998) Expression of the Bcl-2 protein in nasal epithelia of F344/N
49             rats during mucous cell metaplasia and remodeling.  Am. J. Respir. Cell Mol.  Biol. 18: 794-799.
50       Thiele, J. J. (2001) Oxidative targets in the stratum corneum. A new basis for antioxidative strategies. Skin
51             Pharmacol. Appl. Skin Physiol. 14(suppl.  1): 87-91.
52       Thiele, J. J.; Traber, M.  G.; Podda,  M.; Tsang, K.; Cross, C. E.; Packer, L. (1997a) Ozone depletes tocopherols and
53             tocotrienols topically applied to murine skin. FEES  Lett. 401: 167-170.
54       Thiele, J. J.; Podda, M.; Packer, L.  (1997b) Tropospheric ozone: an emerging environmental stress to skin. Biol.
55             Chem. 378: 1299-1305.
         December 2005                                 5-81          DRAFT-DO NOT QUOTE OR CITE

-------
  1       Thiele, J. I; Traber, M. G.; Polefka, T. G.; Cross, C. E.; Packer, L. (1997c) Ozone-exposure depletes vitamin E and
  2             induces lipid peroxidation in murine stratum corneum. J. Invest. Dermatol. 108: 753-757.
  3       Thiele, J. J.; Traber, M. G.; Tsang, K.; Cross, C. E.; Packer, L. (1997d) In vivo exposure to ozone depletes vitamins
  4             C and E and induces lipid peroxidation in epidermal layers of murine skin. Free Radical Biol. Med.
  5             23:385-391.
  6       Thiele, J. J.; Traber, M. G.; Packer, L. (1998) Depletion of human stratum corneum vitamin E: an early and sensitive
  7             in vivo marker of UV induced photo-oxidation. J. Invest. Dermatol. 110: 756-761.
  8       Thiele, J. J.; Hsieh, S. N.; Briviba, K.; Sies, H. (1999) Protein oxidation in human stratum corneum: susceptibility of
  9             keratins to oxidation in vitro and presence of a keratin oxidation gradient in vivo. J. Invest. Dermatol.
10             113:335-339.
11       Tsai, J.-J.; Lin, Y.-C.; Kwan, Z.-H.; Kao, H.-L. (1998) Effects of ozone on ovalbumin sensitization in guinea pigs.
12             J. Microbiol. Immunol. Infect.  31: 225-232.
13       Tsukagoshi, H.; Haddad, E.-B.; Sun, J.; Barnes, P. J.; Chung, K. F. (1995) Ozone-induced airway
14             hyperresponsiveness: role of superoxide anions, NEP,  and BK receptors. J. Appl. Physiol. 78:  1015-1022.
15       U.S. Environmental Protection Agency. (1993) Air quality criteria for oxides of nitrogen. Research Triangle Park,
16             NC:  Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office; report
17             nos.  EPA/600/8-9!/049aF-cF. 3v. Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and
18             PB95-124517.
19       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
20             Research Triangle Park, NC: Office of Research and Development; report nos. EP A/600/AP-93/004aF-cF. 3v.
21             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
22             www.epa.gov/ncea/ozone.htm.
23       Uhlson, C.; Harrison, K.; Allen, C. B.; Ahmad, S.; White, C.  W.; Murphy, R. C.  (2002) Oxidized phospholipids
24             derived from ozone-treated lung surfactant extract reduce macrophage and epithelial cell viability. Chem.
25             Res.  Toxicol. 15: 896-906.
26       Ulrich, M. M. W.; Alink, G. M.; Kumarathasan, P.; Vincent, R.; Boere, A. J.; Cassee, F.  R. (2002) Health effects
27             and time course of paniculate matter on the cardiopulmonary system in rats with lung inflammation.
28             J. Toxicol. Environ. Health Part A 65:  1571-1595.
29       Valacchi, G.; Van der Vliet, A.; Schock, B.C.; Okamoto, T.; Obermuller-Jevic, U.; Cross, C. E.; Packer, L.  (2002)
30             Ozone exposure activates oxidative stress responses in murine skin. Toxicology 179: 163-170.
31       Valacchi, G.; Pagnin, E.; Okamoto, T.; Corbacho, A. M.; Olano, E.; Davis, P. A.; Van der Vliet, A.; Packer, L.;
32             Cross, C. E. (2003) Induction of stress proteins and MMP-9 by 0.8 ppm of ozone in murine skin. Biochem.
33             Biophys. Res. Commun. 305: 741-746.
34       Valverde, M.; del Carmen Lopez, M.; Lopez, I.; Sanchez, I.; Fortoul, T. I.; Ostrosky-Wegman, P.; Rojas, E.  (1997)
3 5             DNA damage in leukocytes and buccal and nasal epithelial cells of individuals exposed to air pollution in
36             Mexico City. Environ. Mol. Mutagen. 30: 147-152.
37       Van Bree, L.; Dormans, J. A. M. A.; Boere, A. J. F.; Rombout, P. J. A. (2001) Time study on development and
38             repair of lung injury following ozone exposure in rats. Inhalation Toxicol. 13: 703-717.
39       Van Bree, L.; Dormans, J. A. M. A.; Koren, H. S.; Devlin, R. B.; Rombout, P. J. A. (2002) Attenuation and recovery
40             of pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhalation Toxicol.
41             14:883-900.
42       Van Hoof, I. H. J. M.; Van Bree, L.; Bast, A.  (1996) Changes in receptor function by oxidative  stress in guinea pig
43             trachea! smooth muscle. Cent. Eur. J. Public Health 4(suppl.): 3-5.
44       Van Hoof, H. J. M.; Van Acker, F. A. A.; Voss, H.-P.; Van Bree, L.; Bast, A. (1997) Acute exposure to ozone does
45             not influence neuroreceptor density and sensitivity in guinea pig lung. Toxicol. Lett. 90:  53-60.
46       Vanda, B.; de Buen, N.; Jasso, R.; Valero, G.; Vargas, M. H.; Olmos, R.; Arreola, J. L.; Santillan, P.; Alonso, P.
47             (1998) Inflammatory cells and ferruginous bodies  inbronchoalveolar lavage in urban dogs. Acta Cytol.
48             42: 939-944.
49       Vargas, M.  H.; Segura, P.; Campos, M. G.; Hong, E.; Montano, L. M. (1994) Effect of ozone exposure on
50             antigen-induced airway hyperresponsiveness in guinea pigs. J. Toxicol. Environ. Health 42: 435-442.
51       Vargas, M.  H.; Romero, L.; Sommer, B.; Zamudio, P.; Gustin, P.; Montano, L. M. (1998) Chronic exposure to ozone
52             causes tolerance to airway hyperresponsiveness in guiea pigs: lack of SOD role. J. Appl. Physiol.
53             84: 1749-1755.
54       Vesely, D. L. (1999) Atrial natriuretic peptides in the diagnosis and treatment of congestive heart failure. Congest.
55             Heart Fail. 5: 171-179.
         December 2005                                5-82         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Vesely, K. R.; Schelegle, E. S.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Hyde, D. M. (1999a) Breathing pattern
  2             response and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am. J. Respir.
  3             Cell Mol. Biol. 20:699-709.
  4       Vesely, K. R.; Hyde, D. M.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Schelegle, E. S. (1999b)
  5             Capsaicin-sensitive C-fiber-mediated protective responses in ozone inhalation in rats. J. Appl. Physiol.
  6             86:951-962.
  7       Vincent, R.; Janzen, E. G.; Chen, G.; Kumarathasan, P.; Haire, D. L.; Guenette, J.; Chen, J. Z.; Bray, T. M. (1996)
  8             Spin trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radical Res. 25: 475-488.
  9       Vincent, R.; Bjarnason, S. G.; Adamson, I. Y. R.; Hedgecock, C.; Kumarathasan, P.; Guenette, J.; Potvin, M.;
10             Goegan, P.; Bouthillier, L. (1997) Acute pulmonary toxicity of urban paniculate matter and ozone. Am. J.
11             Pathol. 151:  1563-1570.
12       Vyskocil, A.; Viau, C.; Lamy,  S. (1998) Peroxyacetyl nitrate: review of toxicity. Hum. Exp. Toxicol. 17: 212-220.
13       Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (200la) Effects of ozone and endotoxin coexposure on rat airway
14             epithelium: potentiation of toxicant-induced alterations. Environ. Health Perspect. 109(suppl. 4): 591-598.
15       Wagner, J. G.; VanDyken, S. J.; Hotchkiss, J. A.; Harkema, J. R. (200Ib) Endotoxin enhancement of ozone-induced
16             mucous cell metaplasia is neutrophil-dependent in rat nasal epithelium. Toxicol. Sci. 60: 338-347.
17       Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (2002) Enhancement of nasal inflammatory and epithelial responses
18             after ozone and allergen coexposure in brown Norway rats. Toxicol. Sci.  67: 284-294.
19       Wagner, J. G.; VanDyken, S. J.; Wierenga, J. R.; Hotchkiss, J. A.; Harkema, J.  R. (2003) Ozone exposure enhances
20             endotoxin-induced mucous cell metaplasia in rat pulmonary airways. Toxicol. Sci. 74: 437-446.
21       Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
22             ability of human surfactant protein A variants to stimulate  cytokine production. Environ. Health Perspect.
23             110:79-84.
24       Watkinson, W. P.; Wiester, M. J.; Highfill, J. W. (1995) Ozone toxicity in the rat. I. Effect of changes in ambient
25             temperature on extrapulmonary physiological parameters. J. Appl. Physiol. 78: 1108-1120.
26       Watkinson, W. P.; Campen, M. J.; Nolan, J. P.; Costa, D. L. (2001) Cardiovascular and systemic responses to
27             inhaled pollutants in rodents: effects of ozone and paniculate matter. Environ. Health Perspect.
28             109(suppl. 4): 539-546.
29       Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
3 0             responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
31             environmental factors. Environ. Res. 92: 35-47.
32       Watt, K. C.; Plopper, C. G.; Weir, A. J.; Tarkington, B.; Buckpitt, A. R. (1998)  Cytochrome P450 2E1 in rat
33             tracheobronchial airways: response to ozone  exposure. Toxicol. Appl. Pharmacol. 149: 195-202.
34       Wattiez, R.; Noel-Georis, L; Cruyt, C.; Broeckaert,  F.; Bernard, A.; Falmagne,  P. (2003) Susceptibility to oxidative
3 5             stress: proteomic analysis of bronchoalveolar lavage from  ozone-sensitive and  ozone-resistant strains of mice.
36             Proteomics 3: 658-665.
37       Weber, S. U.; Thiele, J. J.; Cross, C. E.; Packer, L. (1999) Vitamin C, uric acid, and glutathione gradients in murine
3 8             stratum corneum and their susceptibility to ozone exposure. J. Invest. Dermatol. 113:1128-1132.
39       Weber, S. U.; Jothi, S.; Thiele, J. J. (2000) High-pressure liquid chromatography analysis of ozone-induced
40             depletion of hydrophilic and lipophilic antioxidants in murine skin.  Methods Enzymol. 319: 536-546.
41       Weber, S. U.; Han, N.; Packer, L. (2001) Ozone: an emerging oxidative stressorto skin. Curr. Probl. in Dermatol.
42             29:52-61.
43       Weller, B. L.; Crapo, J. D.; Slot, J.; Posthuma, G.; Plopper, C. G.; Pinkerton, K. E. (1997) Site- and cell-specific
44             alteration of lung copper/zinc and manganese superoxide dismutases by chronic ozone exposure. Am. J.
45             Respir. Cell Mol. Biol. 17: 552-560.
46       Weller, B. L.; Witschi, H.; Pinkerton, K. E.  (2000) Quantitation and localization of pulmonary manganese
47             superoxide dismutase and tumor necrosis factor a following exposure to ozone and nitrogen dioxide. Toxicol.
48             Sci. 54: 452-461.
49       Wells, C. A.; Ravasi, T.; Faulkner, G. J.; Carninci, P.; Okazaki, Y.; Hayashizaki, Y.; Sweet, M.; Wainwright, B. J.;
50             Hume, D. A. (2003) Genetic control of the innate immune  response. BMC Immunol. 4: 5. Available:
51             http://www.biomedcentral.eom/1471-2172/4/5 [18 February, 2003]
52       Wiester, M. J.; Watkinson, W. P.; Costa, D. L.; Crissman, K. M.; Richards, J. H.;  Winsett, D. W.; Highfill, J. W.
53             (1996) Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters. J.
54             Appl. Physiol. 81: 1691-1700.
         December 2005                                 5-83          DRAFT-DO NOT QUOTE OR CITE

-------
 1       Wilkins, C. K.; Clausen, P. A.; Wolkoff, P.; Larsen, S. T.; Hammer, M.; Larsen, K.; Hansen, V.; Nielsen, G. D.
 2             (2001) Formation of strong irritants in mixtures of isoprene/ozone and isoprene/ozone/nitrogen dioxide.
 3             Environ. Health Perspect. 109: 937-941.
 4       Witschi, H.; Espiritu, I.; Pinkerton, K. E.; Murphy, K.; Maronpot, R. R. (1999) Ozone carcinogenesis revisited.
 5             Toxicol. Sci. 52: 162-167.
 6       Wu, Z.-X.; Morton, R. F.; Lee, L.-Y. (1997) Role of tachykinins in ozone-induced airway hyperresponsiveness to
 7             cigarette smoke in guinea pigs. J. Appl. Physiol. 83: 958-965.
 8       Yamauchi, T.; Shima, M.; Kuwaki, T.; Ando, M.; Ohmichi, M.; Fukuda, Y.; Adachi, M. (2002) Acute effects of
 9             ozone exposure on lung function in mice sensitized to ovalbumin. Toxicology (Ireland) 172: 69-78.
10       Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
11             enhances ozone-induced lung injury in B6C3F1 mice. Toxicol.  Sci. 65: 99-106.
12       Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced airways
13             hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
14       Zhao, Q.; Simpson, L. G.; Driscoll, K. E.; Leikauf, G. D. (1998) Chemokine regulation of ozone-induced neutrophil
15             and monocyte inflammation. Am. J. Physiol. 274: L39-L46.
16
17
         December 2005                                 5-84          DRAFT-DO NOT QUOTE OR CITE

-------
 i           AX5.  ANNEX TO CHAPTER 5 OF OZONE AQCD
 2
 3
 4     AX5.1  INTRODUCTION
 5          This annex serves to provide supporting material for Chapter 5, Toxicological Effects of
 6     Ozone and Related Photochemical Oxidants in Laboratory Animals and In Vitro Test Systems.
 7     It includes tables that summarize new toxicological literature published since the last O3 criteria
 8     document (U.S. Environmental Protect!on Agency, 1996). In addition, it provides descriptions
 9     of those new finding, in many cases, with more detail than is provided in the chapter.
10
11
12     AX5.2  RESPIRATORY TRACT EFFECTS OF OZONE
13     AX5.2.1  Biochemical Effects
14          Biochemically detected effects of O3 are integrally involved in effects on both structure
15     and function (respiratory and nonrespiratory) of the respiratory tract.  However, even the few
16     relatively clear associations (e.g., increases in collagen metabolism/collagen and lung fibrosis)
17     are not fully understood, leading to difficulty in interpretation. For presentation, this section
18     summarizes studies designed to identify biochemical targets of O3, as well as biochemical
19     measurements of O3-induced changes in xenobiotic metabolism, antioxidant metabolism and
20     oxygen consumption, lipids and arachidonic acid metabolism, and collagen metabolism.
21     Only descriptions of new studies, published since the previous O3 CD are included. Detailed
22     discussions of older O3 literature are found in U.S. Environmental Protection Agency (1986,
23     1996).
24
25     AX5.2.1.1   Cellular Targets of Ozone Interaction
26          New studies characterizing the cellular targets of O3 interaction include the following.
27     Figure AX5-1 details the major secondary products of ozone interaction with lung cells.
28          Frampton et al. (1999) demonstrated the ozonation of PUFA to form nonanal and hexanal
29     in rat BAL after exposures to 0.22 ppm O3 for 4 h with exercise. Increases in nonanal were not
30     accompanied by significant changes in lung function, in epithelial permeability, or in airway

       December 2005                          AX5-1      DRAFT-DO NOT QUOTE OR CITE

-------
RHC = CH 4
PUFA
either In j
the 	 », RHC
absence N
ofH2O
Criegi
O O
\ I
• O3 	 > RHC — CH— -
ozone trioxolane
0~OX or in the
CH — presence 	 >
xo/ of H2O
se ozonide hyc
	 > RHC = O — O + RHC = O
carbonyl oxide aldehyde
RHC 	 > RHC = O + H2O2
OOH aldehyde hydrogen
iroxyhydroperoxy cpd. peroxide
         Figure AX5-1. Major secondary products of ozone interaction with lung cells.
 1      inflammation. Hexanal levels did not increase significantly and levels of both aldehydes
 2      returned to baseline by 18 h PE. Pryor et al. (1996) exposed rats to 0.5 to 10 ppm O3 both with
 3      and without 5% CO2 to measure the amount of aldehyde generated in BAL, and also the rate of
 4      disappearance of aldehydes from the ELF following the O3 exposure. Ozone exposure with CO2
 5      increased the tidal volume and the yield of aldehydes with a maximal aldehyde yield at 2.5 ppm
 6      for 1 h. Absolute yields were impossible to ascertain in this system because deposition of O3 is
 7      unknown and aldehyde recovery is not complete due to loss of aldehyde by volatization and by
 8      diffusion into underlying tissue. The data showed that at 0.5 ppm O3 with 5% CO2, levels of
 9      hexanal and nonanal increased at 30 minutes, decreased slightly from that level at 60 minutes,
10      was maximal at 90 minutes  and then dropped to 60 minutes levels at 120 minutes. Levels of
11      heptanal did not change appreciably during this time course.  Levels of these aldehydes were
12      dependent on a dynamic relationship between their production and the disappearance from the
13      ELF. The authors stated that O3 is the limiting reagent in this process because the amount of
14      PUFA far exceed the amount of O3 on a molar basis. Because of the limitations of measuring
15      aldehydes in this study paradigm, it is not useful for quantitative dosimetry; however, the authors
16      suggest the study does serve to demonstrate the use of aldehydes as biomarkers of O3 exposure
17      since nonanol is produced in an O3-specific pathway.
18          Postlethwait et al. (1998) utilized three biologically relevant models (isolated epithelial
19      lining fluid, intact lung, and liposome suspensions) to determine the O3-induced production of
20      heptanal, nonanal and hexanal in an attempt to estimate formation of lipid-derived bioactive
       December 2005
AX5-2
DRAFT-DO NOT QUOTE OR CITE

-------
 1      compounds.  Data suggest that PUFAs directly react with O3 and the amount of bioactive lipids
 2      produced is inversely related to ascorbic acid availability.  The authors caution that there are
 3      limitations to the use of measurements of these reactions products in determining O3 dose-
 4      response relationships due to the heterogenous nature of O3 reactions in the epithelial lining
 5      fluid.  Connor et al, (2004) have recently examined the reactive absorption of O3 within ELF
 6      using interfacial films composed of dipalmitoylglycero-3-phosphocholine (DPPC) and rat lung
 7      lavage fluid. The films reduced O3 reactive absorption by antioxidants. Further experiments
 8      using a human lung fibroblast cell line exposed to O3 demonstrated that ascorbic acid (AA)
 9      produced cell injury, that high levels of O3 and AA were needed to induce cell injury, and the
10      DPPC films reduced the amount of cell injury. From these data the authors suggest that O3
11      reactions with ELF substrates cause cell injury, that films of active, saturated phospholipids
12      reduce the local dose of O3-derived reaction products, and that these interfacial phospholipids
13      modulate the distribution of inhaled O3 and the extent of site-specific cell injury.
14          Recent studies have examined the formation of ozonation products such as
15      4-hydroxynonenal (FINE), a toxic aldehyde that reacts with cysteine, histamine, and lysine
16      amino acid residues and creates protein adducts.  Hamilton et al.  (1998) demonstrated (see
17      Chapter 6) using human AM exposed to 0.4 ppm O3 for 1 h that exposure caused apoptosis, an
18      increase in a 32-kDa protein adduct, and an increase in ferritin and a 72-kDa heat shock protein.
19      By exposing AM to FINE in vitro, all of these effects are replicated, which the authors interpret
20      to mean that creation of protein adducts and apoptotic cell death are cellular toxic effect of acute
21      O3 exposure and that it is mediated, at least in part by FINE.
22
23      AX5.2.1.2  Monooxygenases
24          Monooxygenases constitute a class of enzymes, including the cytochrome P-450 (C-P450)
25      or CYP system, that metabolize both endogenous and exogenous substances. Such metabolism
26      can result in detoxification (e.g., drugs like pentobarbital) or activation to more potent
27      metabolites (e.g., carcinogenic metabolites of benzo[a]pyrene, B[a]P). Although the liver  has
28      the greatest capacity for xenobiotic metabolism, the lung also has its complement.  The effects of
29      O3 on this metabolic system are summarized in Table AX5-1.
30          Lee et al. (1998) characterized the activities of various isoforms of CYPs in both rat and
31      rhesus monkey lung using microdissection techniques and found  regiospecific and species-

        December 2005                          AX5-3       DRAFT-DO NOT QUOTE OR CITE

-------
December 2005

X
-k


Table AX5-1.
Effects of Ozone on Lung Monooxygenases

Concentration
ppm
1.0
1.0
0.8
1
aCYP =
WT =
MT =
CCSP
ug/m3 Duration
1,960 8h
1,960 90 days
1,600 8 h/ day for
90 days
1,960 2h
= Cytochrome P-450
wild-type
metallothionein
= Clara Cell Secretary Protein
Species
Rat
male, SD
350-600 g
Rat,
male SD
275-300 g
Mice,
Clara cell secretory
protein deficient,
WT strain 129

Effects"
Increases CYP2E1 activity in lobar-bronchi and major daughter airway with 8 h
exposure. Decreased CYP2E1 activities in both major and minor daughter
airways with 90 day exposure. O3 does not result in consistent dramatic
alterations in CYP2E1 activities.
C YP2B activity increased. Linked to Clara cells in distal lung only — not in
trachea or intrapulmonary airway.
CCSP-1-mice had increases in IL-6 and MT mRNA that preceded decreases in
Clara cell C YP2F2 mRNA. WT mice had levels change but to a lesser degree.

Reference
Watt etal. (1998)
Paige et al. (2000a)
Mango etal. (1998)

 H

 6
 o


 o
 H

O

 O
 H
 W

 O


 O
 HH
 H
 W

-------
 1      specific differences. In rat parenchyma both CYP 1 Al and CYP 2B were highest, whereas, in
 2      rat airways, CYP 2E1 was highest. In rat airways and parenchyma P450 reductase activities
 3      were high, and conversely, low in trachea. Monkeys did not exhibit such site-selective
 4      differences in CYP 2B1, CYP 1A1, and P450 reductase; however, they had high CYP2E1
 5      activity in parenchyma and distal bronchioles.
 6          Watt et al. (1998) found that 1 ppm O3 in both short and long-term exposures in rat
 7      increased CYP 2E1 in a region-specific manner (a more detailed discussion is in the Morphology
 8      section of this chapter). This group (Paige et al., 2000a) further characterized CYP 2B
 9      expression and activity in a long-term O3 exposure (0.8 ppm 8 h/day for 90 days).  Activity of
10      CYP 2B increased 3-fold following this exposure, while CYP 2B like-immunoreactivity
11      increased 2-fold in microsomes prepared from distal lung.  Changes in immunodetectable CYP
12      2B protein and activity were limited to Clara cells in distal lung and not present in trachea or
13      intrapulmonary airways.
14          Studies have focused on P450 gene expression to examine possible genetic mechanisms
15      that  may explain differential O3-sensitivity (Mango et al., 1998). Mice (129 strain) deficient in
16      Clara cell secretory protein (CCSP-/-), which are oxidant-sensitive, were exposed to 1 ppm O3
17      for 2 h. The CCSP null mice demonstrated increases in IL-6 and metallothionein (MT) mRNA
18      that  preceded decreases in Clara cell CYP2F2 mRNA (normally expressed at high levels in
19      mouse lung) levels.  In 129 strain wild-type mice, RNA levels changed similarly, to a lesser
20      degree.
21
22      AX5.2.1.3  Antioxidants, Antioxidant Metabolism, and Mitochondrial
23                 Oxygen Consumption
24          Ozone is an oxidant that produces reactive oxygen species (ROS) involved in the
25      molecular mechanism(s) of toxicity.  Antioxidant chemical defenses are important in modulating
26      O3 toxicity.
27          Weller et al. (1997) studied the changes in distribution and abundance of copper-zinc
28      (Cu-Zn) and manganese (Mn) SOD in Fischer 344 rats exposed to 1.0 ppm O3 for up to
29      3 months. Using immunohistochemistry and immunogold labeling, they identified Cu-Zn SOD
30      positive and Mn SOD positive cells in exposed animals and compared them to controls.
31      In epithelial cells in airways and parenchyma they found reduced Cu-Zn SOD labeling in
       December 2005                           AX5-5       DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3-exposed rats.  In the CAR regions in both AMs and type II epithelial cells they found
 2      significantly increased levels of Mn SOD.  Mn SOD levels were not increased in type I epithelial
 3      cells, fibroblasts, or Clara cells. The authors suggest that the increased levels of Mn SOD in
 4      type II cells in the proximal alveolar duct confer tolerance and protection from further
 5      O3-induced injury.
 6           To look at the effect of antioxidants on modulating O3-induced oxidant stress, Freed et al.
 7      (1999) inhibited antioxidant transport using probenecid (an anion-transport inhibitor) in dogs
 8      exposed to 0.2 ppm in a 6h exposure. Blocking antioxidant transport caused heterogeneously
 9      distributed increases in peripheral airway resistance and reactivity, supporting the hypothesis
10      that in the lung periphery, endogenous antioxidants moderate the effects of O3 and that this
11      exposure is a subthreshold stimulus for producing effects on peripheral airway resistance and
12      reactivity in dogs. Treatment with probenecid also inhibited O3-induced neutrophilic
13      inflammation, which was present in untreated animals exposed to O3.  This finding provides
14      evidence of a dissociation between airway function and inflammation, suggests that O3-induced
15      inflammation and AHR are independent phenomena, and further, that O3-induced neutrophilic
16      influx is dependent on a probenecid-sensitive transport process. The authors postulated that
17      probenecid has either a direct or indirect effect on either cytokine or leukotriene transport.
18      Probenecid treatment also caused a 50-60% decrease in plasma urate, a decrease in ascorbate,
19      and a decrease in BALF protein.
20           Mudway and Kelly (1998) modeled the interactions of O3 with three ELF antioxidants,
21      AA, uric acid and GSH. They used a continually mixed, interfacial exposure set up in perpex
22      chambers with O3 concentrations of 0, 0.1, 0.25, 0.5, 1.0 or 1.5 ppm.  Exposures were carried out
23      with each oxidant individually, with the antioxidants as a mixture, and with and without human
24      albumin. In all three exposure conditions the ranking of reactivity with O3 was uric
25      acid > AA > GSH.  The reactions did not cause changes in sample pH and  no protein carbonyl
26      formations was observed with the antioxidants. They also observed consumption of the
27      antioxidants occurring in a linear fashion with time and a positive relationship to O3
28      concentration.  They concluded that GSH is not an important substrate for  O3, while  uric acid
29      appeared to be the most important substrate which confers protection from O3 by removing it
30      from inhaled air and limiting the amount that reaches the distal lung.  The authors acknowledge
        December 2005                           AX5-6       DRAFT-DO NOT QUOTE OR CITE

-------
 1      limitations in extrapolating these data to in vivo O3 exposures due to the absence of surfactant
 2      lipids and airway mucus in the model system.
 3
 4      AX5.2.1.4  Lipid Metabolism and Content of the Lung
 5          One of the major postulated molecular mechanisms of action of O3 is peroxidation of
 6      unsaturated fatty acids in the lung, prompting interest in measurement of lipids and lipid
 7      metabolism. Several new studies have examined the effects of O3  exposure on phospholipid in
 8      lung tissue.
 9          A new mechanism for the toxicity of O3 was proposed by Pryor et al. (1995). This
10      mechanism suggests that the biochemical changes due to O3 exposure are relayed to deeper
11      tissue in the lung by a cascade of ozonation products.  They consider lipid oxidation products as
12      the most likely molecules to do this because lipids are present in high concentrations in the ELF
13      and they react with  O3 to form stable molecules.  These lipid oxidation products cause activation
14      of specific Upases, which then trigger the activation of second messenger pathways (e.g.,
15      phospholipase A2 or phospholipase C). Experiments were completed by this group (Kafoury et
16      al., 1999) using exposures of cultured human bronchial epithelial (NHBE) cells to the lipid
17      ozonation product 1 -palmitoyl-2-(9-oxononanoyl)-sn-glycero-3-phosphocholine (PC-ALF) and
18      1-hydroxy-l-hydroperoxynonane (HHP-C9). Measurements of PAF, PGE2, IL-6 and IL-8 were
19      completed . PC-ALF  elicited release of platelet-activating factor (PAF) and prostaglandin E2,
20      but not IL-6. HHP-C9 caused release of PAF and IL-6 in these cells, but not prostaglandin E2.
21      These results suggest to the authors that O3-induced production of lipid ozonation products
22      causes release of proinflammatory mediators that then generate an  early inflammatory response.
23          Long et al. (2001) exposed hamsters to 0.12, 1.0 or 3.0 ppm O3 to evaluate lipid
24      peroxidation and antioxidant depletion. After 6 h exposures to the two higher levels resulted in
25      increased BALF neutrophil numbers and F2-isoprostanes. The highest exposure only caused
26      increased levels  of BALF urate and decreased plasma levels of ascorbate. Exposures to the
27      0.12 ppm had no effect on BALF neutrophils of F2-isoprostanes or on plasma antioxidants.
28      Exposures to 1.0 ppm O3 with 1  h of exercise caused increased levels of F2-isoprostanes.
29          Uhlson et al. (2002) evaluated the formation of oxidized phosphlipids by reacting O3 with
30      calf lung surfactant. Low levels of ozone (0.06, 0.125, and 0.25  ppm) for exposures of 2 to 48 h
31      created a dose- and  time-dependent increase in the formation of l-palmitoyl-2-(9'-oxo-

        December 2005                           AX5-7       DRAFT-DO NOT QUOTE OR CITE

-------
 1      nonanoyl)-glycerophosphocholine (16:Oa/9-al-GPCho), an oxidized phospholipid, which
 2      possessed biological activity in three assays.  The 16:Oa/9-al-GPCho: 1) decreased macrophage
 3      viability by necrosis at 6 |iM, 2) induced apoptosis in pulmonary epithelial-like A549 cells at
 4      100 to 200 |iM , and 3) elicited release of IL-8 from A549 cells at 50-100 |iM.
 5
 6      AX5.2.1.5 Protein Synthesis
 7           One new study of the effects of O3 on protein synthesis involved an examination of the
 8      time course of lung injury and changes in collagen content in rats exposed acutely or
 9      subchronically to 0.4 ppm O3 (van Bree et al., 2001). They observed centriacinar thickening of
10      septa after 7 days of exposure. This progressed at 28 and 56 days of exposure. After 28 days of
11      O3, the increase in collagen content in ductular was apparent and it increased progressively until
12      the 56 daytime point. While collagen content decreased with PE recovery, the structural fibrotic
13      changes in ductular septa did not return to control levels. Additionally, they observed the
14      presence of respiratory bronchioles after O3 exposure, which persisted after an 80-day recovery
15      period. These data suggest that subchronic O3 exposures in rats creates a progression of
16      structural lung injury that can evolve to a more chronic form, which included fibrosis.
17
18      AX5.2.1.6 Gene Expression
19           Gohil et al. (2003) have used genomic technology to examine differential gene expression
20      in C57BL/6 mice exposed to 1 ppm O3 for three consecutive nights for 8 h. Utilizing the
21      Affymetrix GeneChip, found O3-induced changes in the  expression of 260 genes, of which 80%
22      were repressed and 20% induced. A number of genes involved in progression of the cell cycle
23      were increased, including ribonucleotide reductase and S-adenosyl methionine decarboxylaseS.
24      Several NF-KB-activated genes were induced including inhibitor of apoptosis, platelet-derived
25      growth factor receptor a, monocyte chemoattractant protein 1, topoisomerase (DNA) II-a, and
26      serum amyloid 3. These genes are causatively linked with inflammation and in concert with the
27      induced cell cycle genes, may account for increased proliferation of Clara cells and Type II
28      pneumocytes.  Ozone caused suppression in the expression of several genes involved in
29      xenobiotic metabolism and in genes coding for major histocompatibility complex. These  data
30      suggest O3 exposure suppresses immune function and xenobiotic metabolism and enhances
31      cellular proliferation.

        December 2005                           AX5-8       DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX5.2.2   Lung Host Defenses
 2           A number of defense mechanisms operate in the respiratory tract to protect the host from
 3      infectious and neoplastic disease.  In humans and in animals, the conducting airways of the lungs
 4      are primarily protected by the mucociliary escalator.  The mucus layer acts to entrap many
 5      gaseous and particulate agents and they are cleared from the tract before they have the
 6      opportunity to reach underlying tissues. Defects in mucociliary transport can be caused by
 7      changing the chemical nature of the mucous secretions, by paralyzing the cilia, or by producing
 8      focal lesions in the ciliated epithelium, making it more susceptible to toxic inhalants. Within the
 9      gaseous exchange region of the lung, the first line of cellular defense against microbes and
10      nonviable particles is the alveolar macrophage (AM).  Impairment of AM function would alter
11      the lung's capabilities to maintain sterility, to clear the lungs of inhaled particles phagocytosized
12      by these cells, to mount an immune response, and to release immunologically-active soluble
13      mediators. Such effects would reduce the host's ability to resist infection and may be involved
14      in the pathogenesis of other chronic diseases. In addition to AM, other local humoral and
15      cell-mediated immune responses are active in protecting the host against such infectious insults,
16      as well as tumor cells.  Animal studies have shown that each one of the above defense systems
17      can be altered following exposure to O3. New studies are summarized in Table AX5-2 are
18      discussed below.
19
20      AX5.2.2.1  Clearance
21      Mucociliary Clearance
22           To ascertain the mechanism(s) by which O3 modulates uptake of particles, Churg et al.
23      (1996) prepared 2 mm SD rat tracheal explants and exposed them to either room air or 0.01,
24      0.05, 0.10, or 1.0 ppm O3 for 10 minutes.  After the O3 exposure, explants were then submerged
25      in either 5 mg/ml amosite asbestos or 4 mg/ml titanium dioxide for 1 h. Uptake of particles
26      assayed 7 days later indicated a dose-dependent uptake of TiO2 starting at 0.01 ppm and uptake
27      of asbestos at the two highest doses. To understand the potential role of oxidative stress in this
28      uptake, in some experiments at doses of 0.1 ppm O3, the reactive oxygen species scavengers
29      catalase or superoxide dismutase, or the iron chelator deferoxamine were added. Uptake of both
30      particles was inhibited by deferoxamine or catalase, but not by superoxide dismutase.  Based on
31      these results, the authors concluded that: (1) uptake of particles in the trachea is a direct effect of

        December 2005                           AX5-9        DRAFT-DO NOT QUOTE OR CITE

-------
o

O

to
O
O
                                          Table AX5-2. Effects of Ozone on Lung Host Defenses
            Concentration3       Duration
                                              Species
                                              Effects"
                                                                       Reference
Microbiologic Endpoints

0.1, 0.3ppm
         0.8 ppm
                               4 h/day,
                               5 days/week,
                               1 or 3 weeks
 Rat
(F344)
                      3h
 Rat
(F344)
No effect on cumulative mortality from subsequent lung infection with     Cohen et al.
4-8 x 106 Listeria monocytogenes, but concentration-related effects on     (2001, 2002)
morbidity onset and persistence.  One-week exposed rats: listeric
burdens trended higher than in controls; 0.3 ppm rats displayed
continual burden increases and no onset of resolution;  in situ IL-lct,
TNF-oc, and IFNy levels 48 and 96 h post-infection (4 x 106 level)
higher than controls. Three-week exposed rats: no O3-related change
in bacterial clearance; IL-la, TNFa, and IFNg levels higher than control
only at 48 h post-infection (4 * 106) and only with 0.3  ppm rats

Single exposure to S. zooepidemicus led to differential clearance           Dong et al.
patterns in exposed rats maintained on ad libitum or O3-mitigating         (1998)
calorie-restricted diets
X
         0.128 to 1.0
                      24 h/day for
                      1 week
 Rat
Rats infected with 3.8 x 108 Listeria monocytogenes (IT). Decreased clearance
of Listeria at highest exposure (1.0 ppm) and impaired cellular immune
response: decreased T/B cell ratios in lymph nodes; delayed-type
hypersensitivity response to Listeria-antigen; and depressed lympho-
proliferation response in spleen and lymph nodes.  Increased formation of
lung granulomas.
                                                                                                                                             Steerenberg
                                                                                                                                             etal. (1996)
H
6
o
o
H
O
         Clearance Endpoints (Non-Microbial)

         0.01-1.0 ppm          lOmin
         0.4 ppm
                      6h
                                               Rat
                                               (SD)
 Dog
Single 10 min exposure of trachea! explants, followed by 1 h incubation    Churg et al.
with particles, led to dose-related increases in uptake of amosite           (1996)
asbestos and titanium dioxide particles. Effect inhibited by added
catalase or desferoxamine, but not by superoxide dismutase.

Increased trachea! permeability to 99mTc-DTPA after direct sublobar       Foster and
exposure to O3. Clearance halftimes remained significantly lower for      Freed (1999)
1-7 d PE, but recovered by 14 days PE.
o
HH
H
W

-------
o

O

to
O
O
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
>
X
H
6
o

o
H
O
Concentration3
Alveolar Macrophage
0.8 ppm
0.8 ppm
1.0 ppm
2.0 ppm
Alveolar Macrophage
0.1, 0.3 ppm
0.1, 0.3 ppm
0.3 ppm
0.8 ppm
0.8 ppm
Duration
Endpoints (General)
3h
3h
Species

Rat (SD)
Rat (SD)
4 h Mouse cell line
3h
Endpoints (Functional)
4 h/day,
5 days/week,
1 or 3 weeks
4 h/day,
5 days/week,
1 or 3 weeks
5 h/day,
5 days/week,
4 weeks
3h
3h
Rat (SD)

Rat (F344)
Rat (F344)
Rat (F344)
Rat (SD)
Rat (F344)
Effects"

Increased ex vivo AM adherence to epithelial cultures mitigated by cell
pretreatment with anti-CD lib or anti-ICAM-1 antibodies.
Increased ex vivo AM adherence to epithelial cell cultures mitigated by
cell pretreatment with anti-TNF-cc/IL-la antibodies.
Increased intracellular calcium resting levels in WEHI-3 cells.
Decreased rates of calcium influx due to digitonin.
Decreased AM reduced glutathione content. Effect blocked by
pretreatment with bacterial endotoxin.

Superoxide anion: increased AM production (1 week; 0.1,0.3 ppm);
no intergroup differences noted after IFNy stimulation. H2O2: reduced
production (1 week; 0. 1, 0.3 ppm); further reduced production after
treatment with IFNy (0.1, 0.3 ppm, 1 and 3 weeks).
Increased AM superoxide anion production (1 week; 0.1, 0.3 ppm),
Lower H2O2 production (1 week; 0.1,0.3 ppm). Reduced production
after treatment with IFNy - superoxide (0.3 ppm, 1 week) and H2O2
(0.1 ppm, 1 week) - relative to cells without IFNy treatment. No effects
from 3 -week exposures.
No effect on AM endotoxin-stimulated IL-loc, IL-6, or
TNF-oc production. Decrease in stimulated, but not spontaneous,
superoxide formation; variable effects on H2O2 formation. No effect
on AM spontaneous, endotoxin-, or IFNy -stimulated, NO formation.
Increased AM motility in response to chemotaxin; effect mitigated by
cell pretreatment with anti-CD 1 Ib or anti-ICAM-1 antibodies.
Decrease in AM phagocytic activity
Reference

Bhalla (1996)
Pearson and
Bhalla (1997)
Cohen et al.
(1996)
Pendino et al.
(1996)

Cohen et al.
(2001)
Cohen et al.
(2002)
Cohen et al.
(1998)
Bhalla (1996)
Dong et al.
(1998)
O
HH
H
W

-------
O
O
1
1-1
to
O
O










:x!
rN
1
K>




o
c
H
6
o

0
H
O


Concentration3 Duration
Table AX5-2 (cont'd).

Species
Effects of Ozone on Lung Host Defenses

Effects"


Reference
Alveolar Macrophage Endpoints (Functional) (cont'd)

0.8 ppm 3 h




1.0 ppm 24 h/day,
3 days
Cytokines, Chemokines: Production,
0.1, 0.3 ppm 4 h/day,
5 days/week,
1 or 3 weeks
0.1, 0.3 ppm 4 h/day,
5 days/week,
1 or 3 weeks





0.3 ppm 5 h/day,
5 days/week,
4 weeks

0.3 ppm 24 or 96 h
1.0 ppm 1, 2, or4h, 2,
2.5 ppm 4, or 24 h

Mouse (B6J129SV)
(C57/BL6X 129 NOS"'")



Rat (Wistar)


Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production and PGE2 release. Initial decrease in ROI
production, with eventual rebound. Knockout (NOS"'") mice AM
incapable of similar response to O3 - no inducible NO or PGE2 above
control levels and consistent decreased ROI production
Lavage fluid from exposed rats subsequently inhibited IFNy -induced
AM NO production.

Fakhrzadeh
et al. (2002)



Koike et al.
(1998, 1999)
Binding, and Inducible Endpoints
Rat (F344)


Rat (F344)







Rat (F344)



Mouse (C57B1/6J)


Superoxide anion: no intergroup differences noted after IFNy
stimulation. H2O2: reduced production after treatment with IFNy.

Decreased expression of CDS among lung lymphocytes (0.1 ppm only;
3 weeks); effect exacerbated by stimulation with IFNcc (but not with
IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+ lymphocytes
(0.3 ppm only; 3 weeks); effect worsened by treatment with IL-lcc
(0.1, 0.3 ppm; 3 weeks). No effects on IL-2 -inducible lympho-
proliferation. Reduced AM production of ROIs after treatment with
IFNy; superoxide (0.3 ppm, 1 week) and H2O2 (0.1 ppm, 1 week) -
relative to untreated cells
No effect on AM endotoxin-stimulated IL-lcc, IL-6, or
TNF-cc production.


0.3 ppm: Increased lung: MIP-2, MCP-1, and eotaxin mRNA
expression.
1.0 ppm: After 4 h, increased lung: MIP-2, MCP-1, eotaxin, and
TL-6 mRNA exnression.
Cohen et al.
(2001)

Cohen et al.
(2002)






Cohen et al.
(1998)


Johnston et al.
(1999a)

                                                                     2.5 ppm: After 2 h, increased lung: MIP-2, MCP-1, eotaxin, and
                                                                     IL-6 mRNA expression.
                                                                     No exposure-related increases in lung IL-lcc, IL-lp, IL-lRcc, IL-10,
                                                                     IL-12, or IFNy mRNA expression.
O
HH
H
W

-------
                                   Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
V 1

(J '
CD
to
O
o


X
OJ
O
H
6
o
0
H
O
Concentration3 Duration Species
Cytokines, Chemokines: Production, Binding, and Inducible
1.0 ppm 4h Mouse cell line
1.0 ppm 6h Rat(SD)
1.0 ppm 24h/day, Rat(Wistar)
3 days
1.0 ppm 24 h Mouse (C57B1/6J)
1.0 ppm 24 h Mouse (C57B1/6J)
1.0 ppm 8 h/day, Mouse (C57B1/6)
3 days (C57Bl/6Ai' NOS^)
1.0, 2.5 ppm 4 or 24 h Mouse (C57B1/6J)


Effects"
Endpoints (cont'd)
Decreased binding of IFNy by WEHI-3 cells. Decreased superoxide
production by IFNy -treated cells; no similar effect on H2O2 production.
Decreased IFNy-stimulatable phagocytic activity. No effect on IFNy-
inducible la (MHC Class II) antigen expression.
Increased AM MIP-lcc, CINC, TNF-cc, and IL-lp mRNA expression.
Induced increase in MIP-lcc and CINC mRNA temporally inhibited by
cell treatment with anti-TNF-cc/IL-lp antibodies.
Lavage fluid from exposed rats subsequently inhibited: ConA-
stimulated lymphocyte IFNy production, but had no effect on IL-2
production; IL-2 -induced lymphopro-liferation; and, IFNy -induced
AM NO production.
Increased lung: MIP-2 (4 h PE) and MCP-1 (4 and 24 h PE) mRNA
expression.
Increased lung MIP-2 and MCP-1 mRNA expression (4 and 24 h PE);
no effects on mRNA levels of IL-lcc, IL-lp, IL-lRcc, IL-6, MIF,
MIP-lcc, MIP-lp, eotaxin, or RANTES at either timepoint in recovery
period. Enhanced expressions of some cytokines/chemokines were
maintained longer than normal by coexposure to endotoxin.
Knockout (NOS"'") mice have more lavageable MIP-2 after exposure
than wild-type; both greater than control.
Dose-related increases in cytokine/chemokine induction. Increased
lung MIP-lcc, MIP-2, eotaxin (4 and 24 h), IL-6 (4 h only), and iNOS
mRNA expression.


Reference
Cohen et al.
(1996)
Ishii et al.
(1997)
Koike et al.
(1998, 1999)
Johnston et al.
(2001)
Johnston et al.
(2002)
Kenyon et al.
(2002)
Johnston et al.
(2000)


O
HH
H
W

-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses

(J '
CD
to
o
o



X
O
H
6
o
0
H
O
o
s
0
o
H
W
Concentration3 Duration Species
Effects"
Cytokines, Chemokines: Production, Binding, and Inducible Endpoints (cont'd)
0.6, 2.0 ppm 3h Mouse (C57BL/6) Increased lung MIP-2 (4 hPE) and MCP-1 mRNA expression (24 h
and Rat PE); PMN and monocyte increased accumulation in lungs consistent
with sequential expression of the chemokines. NF-kB activation also
increased 20-24 h PE.
0.8 ppm 3 h Rat (SD)
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
2.0 ppm 3 h Rat (SD)
180-500 ug/m3 and BALB/C
1% OVA C57BL/6
Alveolar Macrophage/Lung NO- and iNOS-Related Endpoints
0.3 ppm 5 h/day, Rat (F344)
5d/week,
4 weeks
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")


Increased ex vivo AM adherence to epithelial cells mitigated by cell
treatment with anti-TNF-a or IL-la antibodies.
Increased AM IFNy+LPS-induced NOS expression and NO production,
as well as induced PGE2 release. Knockout (NOS"'") mice AM incapable
of similar response to O3 - no inducible NO or PGE2 above control
levels.
Increased AM IFNy+LPS-induced NOS expression and NO production.
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production. AM from exposed rats showed rapid
onset/prolonged activation of NF-KB.
O3 - dose-dependent increases in IgE, IL-4, IL-5; recruitment of
eosinophils and lymphocytes in BALB/c; O3 + OVA - increased IgG,
antibody liters, leukotrienes, airway responsiveness, immediate
culaneous hypersensitivity reactions in BALB/c. In C58BL/6 only
O3 + OVA caused culaneous hypersensitivity and altered IgG responses.
No effecl on AM spontaneous, endotoxin-, or IFNy -stimulated,
NO formation.
Increased AM IFNy+LPS-induced NOS expression and NO production
and PGE2 release. Knockoul (NOS"'") mice AM incapable of similar
response to O3 - no inducible NO or PGE2 above conlrol levels


Reference
Zhao el al.
(1998)
Pearson and
Bhalla (1997)
Fakhrzadeh
el al. (2002)
Laskin el al.
(2002)
Laskin el al.
(1998a)
Neuhaus-
Sleinmelz el al.
(2000)
Cohen el al.
(1998)
Fakhrzadeh
el al. (2002)



-------
Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses
V 1

(J '
CD
to
o
o



X

O
H
6
o
0
H
O
O
0
o
H
W
Concentration3 Duration Species
Alveolar Macrophage/Lung NO- and iNOS-Related Endpoints
0.8 ppm 3 h Mouse (B6J129SV)
(C57B1/6X 129 NOS"'")
1.0 ppm 8 h/day, Mouse (C57B1/6)
3 days (C57Bl/6Ai" NOS"'")
1.0 ppm 24 h/day, Rat(Wistar)
3 days
1.0, 2.5 ppm 4 or 24 h Mouse (C57B1/6J)
2.0 ppm 3 h Rat (SD)
2.0 ppm 3 h Rat (SD)
3.0 ppm 6h Rat (Brown Norway)
0.12, 0.5, or 2 ppm 3 h Mice BALB/c







Effects"
(cont'd)
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production. AM from exposed mice showed rapid and
prolonged activation of NF-KB, STAT-1 (expression, activity),
phosphoinositide 3-kinase, and protein kinase B.
Knockout (NOS"'") mice have more lavageable PMN, MIP-2, and
protein in lungs after exposure than wild-type.
Lavage fluid from exposed rats subsequently inhibited IFNy -induced
AM NO production.
Dose-related increase in lung iNOS mRNA expression
Increased AM spontaneous, IFNy, and LPS-induced NO production,
as well as spontaneous and LPS-induced NOS expression. Effect
somewhat ameliorated by pretreatment with bacterial endotoxin.
Increased AM spontaneous and IFNy+LPS-induced NOS expression
and NO production. AM from exposed rats showed rapid
onset/prolonged activation of NF-KB.
Increased lung iNOS mRNA expression. Effect blocked by
pretreatment with dexamethasone.
Dose-dependent increases in nitrate and Penh; increases in nNOS but
not iNOS or eNOS.







Reference
Laskin et al.
(2002)
Kenyon et al.
(2002)
Koike et al.
(1998, 1999)
Johnston et al.
(2000)
Pendino et al.
(1996)
Laskin et al.
(1998b)
Haddad et al.
(1995)
Jang et al.
(2002)








-------
Table AX5-2 (cont'd).  Effects of Ozone on Lung Host Defenses

(J '
CD
to
O
O



X
Oi

Concentration3
Duration
Surface Marker-Related Endpoints
0.8 ppm 3 h
1.0 ppm
1.0 ppm
1.0 ppm
0.1, 0.3 ppm
4h
2h
3 days
4 h/day,
5 days/week,
1 or 3 weeks
Species
Rat (SD)
Mouse cell line
Rat (SD)
Rat (Wistar)
Rat (F344)
Effects"
Increased expression of AM CD lib, but no effect on ICAM-1
No effect on IFNy-inducible la (MHC Class II) antigen expression on
WEHI-3 cells.
Decreased expression of integrins CD 18 on AM and CD1 Ib on PMN.
No effect on PMN CD62L selectin.
Increased expression of surface markers associated with antigen
presentation: la (MHC Class II) antigen, B7.1, B7.2, and CDllb/c on
BAL cells. Effect attributed to influx of monocytes.
Decreased expression of CDS among lung lymphocytes (0.1 ppm only;
3 weeks); effect exacerbated by stimulation of cells with IFNcc (but not
with IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+
lymphocytes (0.3 ppm only; 3 weeks); effect worsened by treatment of
cells with IL-lcc (0.1 and 0.3 ppm; 3 weeks).
Reference
Bhalla (1996)
Cohen et al.
(1996)
Hoffer et al.
(1999)
Koike et al.
(2001)
Cohen et al.
(2002)
NK- and Lymphocyte-Related Endpoints
0.1, 0.3 ppm
4 h/day,
5 days/week,
1 or 3 weeks
Rat (F344)
Decreased expression of CDS among lung lymphocytes (0.1 ppm only;
3 weeks); effect exacerbated by stimulation of cells with IFNcc (but not
with IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+
Cohen et al.
(2002)
                       lymphocytes (0.3 ppm only; 3 weeks); effect worsened by treatment of
                       cells with IL-lcc (0.1 and 0.3 ppm; 3 weeks). Lymphoproliferation: no
H
O
§ 0.4, 0.8, 1.6 ppm 12 h
H
O
C 1.0 ppm 24 h/day,
2 3 days
W
O
H
W
VAAWL \JAA OJJWAALCIAAVW LIO \JA AA_/-ZJ-AAAUUX-AU'AV_< AWAAAAO, \J . A JJJJAAA AAAVAVO.OVU.
response to ConA mitogen (1 week only); 0.3 ppm - decreased response
to ConA (1 week only).
Mouse (Balb/c) Decreased pulmonary delayed-type hypersensitivity reactions to low
MW agents, likely via activation of TH2 -dependent pathways.

Rat (Wistar) Lavage fluid from exposed rats subsequently inhibited ConA-stimulated
lymphocyte IFNy production, but had no effect on IL-2 production;
material also inhibited IL-2 -induced lymphoproliferation.





Garssen et al.
(1997)

Koike et al.
(1999)





-------
                                           Table AX5-2 (cont'd). Effects of Ozone on Lung Host Defenses

(J '
CD
to
O
O



Concentration3 Duration
Susceptibility Factors
0.3 ppm 24 to 72 h
1 ppm 4 h
10 ppm until death

Species
Mice
C57BL/6J
C3H/HeJ
C3H/HeOuJ
CHO-K1 cell line SP-A
Mice
A/J (O3 sensitive)
and C57BL/6J
(O3-resistant)
Effects"
Lavageable protein concentration lowered by inhibition of iNOS and by
targeted disruption of Nos2; reduced Nos2 and Tlr4 mRNA levels in the
O3-resistant C3H/HeJ mice.
Differences exist biochemically and functionally in SP-A variants.
O3 exposure affects the ability of variants to stimulate TNF-cc and IL-8.
No differences in histology or wet-to-dry lung weights; two loci- acute
lung injury -1 or -2 confer susceptibility.

Reference
Kleeberget et al.
(200 Ib)
Wang et al.
(2002)
Prows et al.
(1999) and
Prows et al.
(1997)
X
Conversion to ug/m3 ~ ppm value x 1960
bCommon abbreviations used:
 AM = Alveolar macrophage; PE = Postexposure (i.e., time after O3 exposure ceased); MIP = macrophage inflammatory protein;
 PMN = Polymorphonuclear leukocyte; MLN = Mediastinal lymph node; CINC = cytokine-induce neutrophil chemoattractant;
 BAL = Bronchoalveolar lavage; DTPA = diethylenetriaminepentaacetic acid; ROI = reactive oxygen intermediate/superoxide anion;
 IFN = Interferon; BALT = bronchus-associated lymphoid tissue; MCP = monocyte chemoattractant protein; CON A = Concanavalin A
H
6
O

O
H
O
O
HH
H
W

-------
 1      O3, as inflammatory cells are not present in the explants, (2) reactive oxygen species are
 2      important mediators of particle uptake, with hydrogen peroxide having a primary role in the
 3      process, and (3) due to the protective effect of deferoxamine, hydroxyl radical is probably
 4      involved in the uptake also.
 5           Pearson and Bhalla (1997) have utilized the radiolabled chelate 99mTc diethylenetriamine
 6      pentaacetic acid (Tc-DTPA) to assess the effect of O3 exposure on clearance across epithelial
 7      surfaces. 99mTc-DTPA clearance has been found to be significantly increased following a 3-h
 8      exposure to 0.8 ppm O3 in SD rats. Pretreatment with anti-IL-la and anti-TNF-a did not affect
 9      the permeability, suggesting that these soluble mediators are not involved in this process.  Foster
10      and Freed (1999) also used 99mTc-DTPA to examine regional clearance in dogs following a 6-h
11      isolated sublobar exposure to 0.4 ppm O3 or air.  Ozone decreased the clearance halftime of
12      99mTc-DTPA by 50% at 1 day following exposure. Seven days PE the halftime was still reduced
13      by 29% and by 14 days PE, clearance had recovered to normal levels. These data provide
14      evidence that a single local exposure to O3 increases transepithelial clearance, but without any
15      influence on contralateral segements, i.e., only for epithelia directly exposed to O3, and the
16      altered  permeability changes recover to normal levels in 2 weeks.
17
18      Alveolar Clearance
19           New evaluations of the effects of O3 on alveolar clearance have not been performed.
20
21      AX5.2.2.2  Alveolar Macrophages
22           Effects of O3 on other AM functions are summarized on Table AX5-2 and new studies are
23      discussed here.
24           Dong et al. (1998) reported that caloric restriction enhanced phagocytic function in
25      O3-exposed rats.  Whereas ad-libitum fed rats had a prolonged infection and pulmonary
26      inflammation from a Streptococcus challenge, calorie restricted rats had no infection and no
27      inflammation.
28           Ozone exposure has been implicated in altered chemotaxis and cell adhesion properties of
29      AM.  Bhalla (1996) reported that macrophages isolated from O3-exposed SD rats (0.8 ppm O3 or
30      air, nose-only, for 3 h and then examined AM from BALF at 12 h PE) showed greater mobility
31      and greater adhesion than AM isolated from air-exposed rats. This increased mobility and

        December 2005                          AX5-18      DRAFT-DO NOT QUOTE OR CITE

-------
 1      adhesion were attenuated when AMs were incubated with monoclonal antibodies to CD1 Ib
 2      (leucocyte adhesion molecules) or ICAM-1 (epithelial cell adhesion molecules).  The authors
 3      suggest that these observed changes in basic cell surface-associated macrophage properties are
 4      relevant to subsequent O3-induced lung inflammatory responses.  Using the same O3-exposure
 5      parameters, Pearson and Bhalla (1997) also observed that the increased AM adherence to
 6      epithelial cell cultures induced by O3 exposure was found to be mitigated by pretreatment with
 7      antibodies to TNF-a and IL-la, suggesting that the early inflammatory response to O3, in part,
 8      may be mediated by IL-la and/or TNF-a.
 9           Additional studies have been carried out to characterize the mechanisms by which O3
10      induces decreased lung resistance against  microbial  pathogens. Cohen et al. (1996) have
11      exposed the WEHI-3 cell line,  a BALB/c myelomonocytic AM-like cell, to 1 ppm O3 for 4 h to
12      determine the effects of O3 on AM activation by interferon-y (INF y). Ozone at this
13      concentration reduced binding  of INF y to AM and affected the AM functional parameters of
14      phagocytic activity, production of reactive oxygen intermediates, and elevation of intracellular
15      calcium.  Further,O3 increased  intracellular calcium  resting levels and decreased the rates of
16      calcium influx due to digitonin. The authors  concluded that this O3-induced modulation of AM
17      function could be responsible for the increased microbial pathogen survival following O3
18      exposure.  Pendino et al. (1996) studied the role of glutathione content in AM functions.  When
19      BALF recovered from female SD rats exposed to 2 ppm or air for 3 h was assayed for
20      intracellular glutathione (with the fluorescent indicator dye monochlorobimane [MCB]),
21      indicated that AM from O3-exposed rats had reduced levels of intracellular glutathione compared
22      to air-exposed rats.  This reduction in glutathione levels may be due to its interaction with
23      ozonation products from O3-induced lipid peroxidation.
24           Bactericidal activity of AM is mediated by hydrogen peroxide production.  To study the
25      effect of O3 on this response function of AM, Cohen et al. (2001,2002) exposed male  F-344 rats
26      to either 0.1 or 0.3 ppm O3 for  4 h/day, 5 days/week or either 1 or 3 weeks and assessed
27      superoxide anion and hydrogen peroxide production in AM recovered from BAL 24 h PE. They
28      found increased superoxide anion production at 1 week 0.1 and 0.3 ppm exposure and did not
29      observe any intergroup differences when stimulated by INFy.  Conversely, hydrogen  peroxide
30      production was reduced at both exposure concentrations and durations and was further reduced
        December 2005                          AX5-19      DRAFT-DO NOT QUOTE OR CITE

-------
 1      with INFy stimulation. The authors suggest that the compromised killing of bacteria by AM in
 2      O3-exposed rats may be due to the reduction in hydrogen peroxide production.
 3           Laskin et al. (1998a) have examined the activation of AM and type II epithelial cells in
 4      female SD rats exposed to 2 ppm O3 for 3  h. Ozone treatment caused a time-dependent increase
 5      in NO levels in both cell types that was correlated with increased expression of iNOS mRNA
 6      and protein. Laskin et al (1998b) hypothesized that the inflammatory mediators such as TNF-a
 7      and IL-lb may mediate the increase in NO release by activating the  expression of iNOS through
 8      NF-KB signaling. They demonstrated it by treating the cells with pyrrolidine dithiocarbamate,
 9      an inhibitor of NF-KB that caused a dose dependent inhibition of NO production and iNOS
10      expression. This group (Laskin et al., 2002) further investigated the mechanisms by which O3
11      activates AM using C57B16xl29 mice with a targeted disruption of the gene for iNOS. These
12      mice exposed to 0.8 ppm O3 for 3 h showed no toxicity as measured by BALF protein levels and
13      nitrotyrosine staining of the lung.  Additionally, mice overexpressing human Cu, Zn superoxide
14      dismutase (SOD) and mice with a targeted disruption of p50 NF-KB were also resistant to O3
15      toxicity.  Wild-type mice exposed to O3 showed an increase in expression of STAT-1, a protein
16      that binds to the regulatory region of iNOS.  Taken together, these results suggest to the authors
17      that a number  of proteins including NF-KB, phosphoinoside 3-kinase, and STAT-1 that bind to
18      and regulate expression of iNOS are modulated by O3 exposure.  Another study by this group
19      (Fakhrzadeh et al., 2002) used the same iNOS knockout mice strain  to further characterize O3
20      toxicity.  In wild-type mice O3 exposure causes an increase in AM superoxide anion and
21      prostaglandin (PG)E2, but in the knockouts, the reactive nitrogen intermediates were not
22      produced and (PG)E2 was at control levels. Further discussions of the role of nitric oxide
23      synthase and reactive nitrogen in  O3-induced inflammation are contained in  Section 5.2.3.5.
24      Additionally, cytokines and chemokines are very important components of the AM response to
25      O3 and are discussed in detail  in Section 5.2.3.4.
26
27      AX5.2.2.3 Immune System
28           Other than by natural protection (e.g., opsonizing antibody, nonspecific phagocytosis by
29      AM), the immune system defends the lung by mounting three major waves of response:  natural
30      killer (NK) cells (nonspecific  lymphocytes that kill viruses, bacteria, and tumor cells), followed
31      by cytotoxic T-lymphocytes (TCTL- lymphocytes that lyse specifically recognized microbial and

        December 2005                         AX5-20       DRAFT-DO NOT QUOTE OR CITE

-------
 1      tumor-cell targets), followed by antigen-specific antibodies.  These T-cell types are involved
 2      with other immunologically active cells (e.g., B-cells and AM), which in a complex manner,
 3      interact in immunological defense. To date, only a few of these mechanisms have been
 4      investigated in the context of their role in O3 susceptibility.  Effects on the systemic immune
 5      system can be different from those in the lung (see Section 5.3). New studies reporting
 6      O3-induced effects on the immune system in the lung are described here.
 7           Garssen et al. (1997) have studied the effects of O3 on non-IgE-mediated pulmonary hyper-
 8      immune reactions induced by picryl chloride (PCI). BALB/c mice sensitized with PCI, both
 9      actively  and passively (by adoptive transfer of lymphoid cells from pre-sensitized mice), were
10      then challenged with picryl sulfonic acid (PSA).  The mice were exposed to!2 h of 0.4, 0.8, or
11      1.6 mg/m3 O3 during one night, at 4 days or 7 days after skin sensitization (which was either just
12      before or just after PSA challenge, i.e., during the induction or effector phase).  Non-sensitized
13      mice showed no changes in tracheal reactivity to carbacol with O3 exposure.  Sensitized mice
14      were hyperreactive to carbachol 48 h after PSA challenge, whereas sensitized mice exposed to
15      all concentrations of O3 showed no significant tracheal hyperreactivity to carbachol.  The
16      sensitized mice also demonstrated a suppressed inflammatory reaction (PMN) with 1.6 mg O3
17      exposure.  O3 exposure following PSA challenge  also caused a suppression of tracheal
18      hyperresponsiveness. In a separate experiment wherein mice were exposed to O3 before
19      sensitization and then lymphoid cells from these mice were injected into non-exposed mice, the
20      recipients also demonstrated an inhibition of the induction of hyperreactivity.  These results are
21      opposite to the effect on type I (IgE-mediated) allergic reactions, which the authors suggest is
22      due to activation of Th-2 cell-dependent reactions that are possibly potentiated by O3 or to a
23      direct effect by O3 on Th-1 cells or other cells that are crucial for the tracheal hyperreactivity and
24      inflammation seen in this mouse model.
25           Recent evidence also point towards the potential interaction between the innate and
26      acquired immune system with O3  exposure. Kleeberger et al. (2000) performed a genome screen
27      on O3-susceptible (C57BL/6J) and O3-resistant (C3H/HeJ) mice and identified a candidate gene
28      on chromosome 4, Toll-like receptor 4 (Tlr4), a gene implicated in endotoxin susceptibility  and
29      innate immunity. When O3-resistant strain C3H/HeJ and C3H/HeOuJ (differing from the O3-
30      resistant strain by a polymorphism in the coding region of Tlr4) were exposed to 0.3 ppm for 24
31      to 72 h, greater protein concentrations were demonstrated in the OuJ strain.  The two strains

        December 2005                          AX5-21      DRAFT-DO NOT QUOTE OR CITE

-------
 1      exhibited differential expression of Tlr4 mRNA with O3 exposure.  These data point to a
 2      quantitative trait locus on chromosome 4 as being responsible for a significant portion of the
 3      genetic variance in O3 -induced lung hyperpermeability. Further investigation by this laboratory
 4      (Kleeberger et al., 2001a) using these mouse strains showed lavageable protein concentration
 5      was lowered by inhibition of inducible nitric oxide synthase (iNOS) and by targeted disruption
 6      ofNos2. Further comparisons on O3 exposure in these two strains (C3FI/HeJ and C3H/HeOuJ)
 7      demonstrated reduced Nos2 and Tlr4 mRNA levels in the O3-resistant C3FI/HeJ mice.  These
 8      data are consistent with the hypothesis that O3 -induced lung hyperpermeability is mediated by
 9      iNOS .  These studies also suggested a role for toll-like receptor 4 (TLR4) in the host response to
10      O3 similar to the role it has demonstrated in LPS sensitivity (Schwartz, 2002; Wells et  al., 2003).
11      TLR4 signaling is thought to be critical to linking the innate and acquired immune system
12      through antigen presenting cells and Thl/Th2 differentiation.
13          Neuhaus-Steinmetz et al.  (2000) compared the response to repeated  O3 (180-500 |ig/m3)
14      and OVA (1%) exposure in "IgE-high responder" (BALB/c) and "IgE-low responder "
15      (C57BL/6) mice. In BALB/c mice exposed to O3, a T-helper (Th)2-like response consisting of
16      dose-dependent increases in IgE, IL-4, IL-5, and recruitment of eosinophils and lymphocytes
17      into airways was generated. Concurrent O3/OVA exposures in BALB/c mice increased IgG,
18      antibody liters, leukotrienes, airway responsiveness, and immediate cutaneous hypersensitivity
19      reactions. In C57BL/6 mice only the combined O3/OVA exposure caused immediate cutaneous
20      hypersensitivity and altered IgG responses, thus demonstrating that O3 has the potential for
21      shifting the immune response toward a Th2-like pattern in two mouse strains with differing
22      potentials for developing allergic reactions. Becker et al. (1991) have demonstrated changes in
23      IgG production with O3 exposures of 1.0, 0.5, and 0.1 ppm for 2 h in vitro with human
24      lymphocytes.  Subsequent to O3 exposure, when lymphocytes were stimulated with pokeweed
25      mitogen (PWM, a T-cell-dependent stimulus) or Staphylococcus aureus Cowan 1 strain (SAC, a
26      T-cell-independent stimulus), both B and T cells were found to be affected by O3 preexposure.
27      T cells also  demonstrated an increase in IL-6 and a decrease in IL-2, suggesting that O3 may
28      have direct effects on IgG-producing cells and concurrently an effect that  is  mediated by altered
29      production of T cell immunoregulatory molecules.
30
       December 2005                          AX5-22       DRAFT-DO NOT QUOTE OR CITE

-------
 1           Surfactant protein A and D (SP-A and SP-D) are members of the collectin family, named
 2      for their composition of both collagens and lectins. The surfactant proteins are secreted by
 3      airway epithelial cells and are a part of the innate immune response with important
 4      immunomodulatory function.  Perturbations in lung immune defenses lead to a feedback loop
 5      between inflammation and SP-A and SP-D levels (reviewed in Hawgood and Poulain, 2001).
 6      Costing et al. (1991) investigated whether surfactant protein A (SP-A) is a target of O3 toxicity
 7      by exposing human and canine SP-A to 0.75 ppm for 4 h.  Functionally the in vitro exposure
 8      inhibited SP-A self-association and SP-A-mediated lipid vesicle aggregation.  Additionally SP-A
 9      decreased binding of SP-A to  mannose. Structurally, O3 oxidized tryptophan and methionine
10      residues on the protein. Additional work by Costing et al. (1992) further examined the effect of
11      O3 on the function of SP-A from dog and human.  In vitro O3 exposure at 0.4 or 0.75 ppm for 4 h
12      reduced the ability of SP-A to inhibit phospholipid secretion by alveolar type II cells and
13      reduced the capacity of SP-A to induce superoxide anion production and enhance phagocytosis
14      of herpes simplex virus (HSV).  In vivo exposures of rats (0.4 ppm for 12 h) generated SP-A less
15      capable of stimulating superoxide anion production by AM. These data suggested that inhibition
16      of interactions between SP-A  and alveolar cells may be one mode of toxicity of O3.  Bridges
17      et al. (2000) reported that both SP-A and SP-D directly protect surfactant phospholipids and
18      macrophages from oxidative damage. Both proteins were found to block accumulation of
19      TEARS and conjugated dienes generated during oxidation of surfactant lipids or low density
20      lipoprotein particles by a mechanism that does not involve metal chelation or oxidative
21      modification of the proteins. Wang et al. (2002) expressed human variants of PS-A in CHO cells
22      and then exposed the expressed protein to O3 (Ippm for 4 h).  All of the eight SP-A variants
23      studied showed decreased ability to stimulate cytokine (TNF- and IL-8) production in THP-1
24      cells, a macrophage-like cell line. Each variant exhibited a unique time- and dose-dependent
25      pattern of stimulation of cytokine production with O3 exposure, suggesting their potential role in
26      underlying susceptibility to O3 toxicity.  Targeted disruption of mouse SP- A and SP-D
27      (Hawgood et al, 2002) caused increases in bronchoalveolar lavage phospholipid, macrophage,
28      and protein through 24 weeks of age.  Further,  the deficient mice developed patchy lung
29      inflammation and air space  enlargement consistent with emphysema.
30
        December 2005                          AX5-23       DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX5.2.2.4  Interaction with Infectious Microorganisms
 2           Numerous investigations have sought to understand the effects of O3 on overall functioning
 3      of host defense systems by challenging animals with infectious agents before, during, or after
 4      exposure and observing the outcome, typically mortality. This work (summarized in
 5      Table AX5-2) shows that the results are dependent on microbial species, animal species, and O3
 6      exposure and its temporal relationship to infectious challenge.
 7           Recent studies of O3-induced modulation of cell-mediated immune responses showed
 8      effects on the onset and persistence of infection.  Cohen et al. (2001,2002) exposed male F-344
 9      rats to either 0.1 or 0.3 ppm O3 for 4 h/day, 5 days/week or either  1 or 3 weeks.  One day later
10      the rats were instilled with viable Listeria monocytogenes (4 x 106 for 8 x 106) and then tested at
11      1, 58, 72, or 96 h postinfection.  There was no observed effect on  cumulative mortality, but there
12      was a concentration-related effect on morbidity onset and persistence. In the one week-exposed
13      rats the listeric burdens trended higher than in controls and the high dose rats showed continual
14      burden increased and without resolution. Levels of IL-la, TNF-a, and IFNy were higher than
15      controls  at the 48 and 96 h time period.  In the 3-week-exposed rats there were no changes in
16      bacterial clearance. Levels of IL-la, TNF-a, and IFNy were higher than controls only at the 48
17      h time period only in the 0.3 ppm-exposed rats.  These observations suggest that exposure to O3
18      may be causing a possible imbalance between Th-1 and Th-2 cells, which can subsequently lead
19      to suppression of the resistance to intracellular pathogens. In this case, the host defense to
20      Listeria monocytogenes, which is predominantly a Th-1-type response, is adversely affected.
21
22      AX5.2.3   Inflammation and Lung Permeability Changes
23           Ozone has long been recognized to cause lung inflammation and increased permeability,
24      which are distinct events under control by independent mechanisms.  The normal lung has an
25      effective barrier function that controls bidirectional flow of fluids and cells between the air and
26      blood compartments.  Ozone disrupts this function, resulting in an increase in serum proteins,
27      bioactive mediators and polymorphonuclear leukocytes (PMNs) in the air spaces of the lung.
28      These inflammatory changes have been  detected microscopically  in tissue preparations and by
29      analyses of bronchoalveolar lavage fluid (BALF). Generally, the  initiation of inflammation is
30      part of a defense process; however, its persistence and/or repeated occurrence can result in health
        December 2005                          AX5-24       DRAFT-DO NOT QUOTE OR CITE

-------
 1      effects. For example, at the early stage, the increase in PMNs may enable more phagocytosis of
 2      microbes, and increased edema fluids or protein associated with the PMN influx could provide
 3      an enhanced medium for microbial growth, however, these are untested hypotheses. Mediators
 4      in the influxing fluid and secreted from the increased number of PMNs can recruit other cells
 5      (e.g., fibroblasts and AMs). If released, the large stores of proteolytic enzymes in these
 6      phagocytic cells could damage lung tissue over the longer term. Cheek et al. (1994) suggest that
 7      PMNs may also play a role in removal of O3-injured cells. The exact role of inflammation in
 8      causation of lung disease is not known, nor is the relationship between inflammation and
 9      changes in lung function.  However, it is associated with acute changes in pulmonary function
10      and chronic diseases such as asthma, chronic bronchitis and emphysema.  Table AX5-3
11      summarizes key new studies on the potential for O3 to increase lung permeability and to cause
12      inflammation.
13
14      AX5.2.3.1 Time Course  of Inflammation and Lung Permeability Changes
15          Studies on the time course of the inflammatory response (e.g., Cheng et al., 1995) indicated
16      that the maximal increases in BALF protein, albumin and number of PMNs occur 8 to 18 h
17      (depending on the study) post acute exposure. In rats, a single 3-h exposure to 0.5 ppm O3
18      produced a significant increase in both permeability and inflammation, but a comparable
19      exposure to 0.3 or 0.15 ppm did not produce an effect (Bhalla and Hoffman, 1997).  Dye et al.
20      (1999) reported that an acute exposure to 0.5 ppm O3 results in a significantly greater lung
21      injury, inflammation and BALF  levels of IL-6 in Wistar rats than in Sprague Dawley or F344
22      rats. BALF cell count,  PGE and IL-6 levels were consistently lower in F344 rats compared to
23      other strains. Sun et al. (1997) exposed OVA-sensitized male Dunkin-Hartley guinea pigs to 1
24      ppm O3 for 3 h and examined protein levels and PMN levels at 3 h PE. PMN levels were
25      significantly increased, without any change in BAL protein levels, suggesting a lack of
26      correlation between the two endpoints.  When guinea pigs exposed to 1 ppm O3 for 1 h were
27      evaluated 24 h PE, they exhibited the same trend, increase in PMN without a concordant
28      increase in BAL protein levels.  Increased AHR observed in the first group but not in the second
29      group suggest a dissociation between PMN levels and AHR.
30          Depending upon species tested and exposure regimens, continuous exposure for 3 to 7 days
31      resulted in an increase in BALF  protein and PMNs that typically peak after a few days and return

        December 2005                          AX5-25      DRAFT-DO NOT QUOTE OR CITE

-------
December 2
o
o









.
X
to
Oi


O
^
fe
H
6
0
n
s.**
H
O
0
H
W
O
/^
Table AX5-3. Effects of Ozone on Lung Permeability and Inflammation
Concentration
ppm

0.1
0.2
0.5
0.1
0.2
0.4
1.0

0.2
0.4
0.8

0.26





0.3





0.1
0.3
1.0

Hg/m3

196
392
980
196
322
784
1,960

392
784
1,568

510





588





196
588
1,960

Duration Species

0.5 h Primary rat alveolar
in vitro type II cells

1 h, in vitro Guinea pig (Hartley) and
human alveolar
macrophages


23 h/day for 1 week Guinea pigs, F (Hartley),
260-330 g


8 h/day, 5 days/week Mice, M (mast cell-
for 1-90 days deficient and
-sufficient),
6-8 weeks old


48 hand 72 h. Mice, M (C57BL/6J and
Exposures repeated C3H/HeJ) 6-8 weeks old
after 14 days



60 min Rat basophilic leukemia
cell line (RBL-2H3)


Effects3

Decreased resistance (Rt ) after 0.5 ppm from 2 to 24 h PE
and at 48 h in monolayers subjected to PMNs. Significantly
lower Rt after PMN treatment at 0.2 and 0.5 ppm.
Exposure of guinea pig alveolar macrophages to 0.4 ppm for
60 minutes produced a significant increase in IL-6 and TNF-
cc, and an exposure of human alveolar macrophages to
identical O3 concentration increased TNF-oc, IL-lb, IL-6 and
IL-8 protein and mRNA expression.
Increase in BALF protein and albumin immediately after 0.8
ppm exposure, with no effect of ascorbate deficiency in diet.
O3 -induced increase in BALF PMN number was only
slightly augmented by ascorbate deficiency.
Greater increases in lavageable macrophages, epithelial cells
and PMNs in mast cell -sufficient and mast cell-deficient
mice repleted of mast cells than in mast cell-deficient mice.
Cyinduced permeability increase was not different in
genotypic groups.

Greater BALF protein, inflammatory cell and LDH response
in C57BL/6J than in C3H/HeJ after initial exposure.
Repeated exposure caused a smaller increase in BALF
protein and number of macrophages, lymphocytes and
epithelial cells in both strains, but PMN number was greater
in both strains of mice compared to initial exposure.
O3 inhibited IgE- and A23 187 - indued degranulation.
Spontaneous release of serotonin and modest generation of
PGD2 occurred only under conditions that caused
cytotoxicity.
Reference

Cheek et al.
(1995)

Arsalane et al.
(1995)



Kodavanti
etal. (1995)


Kleeberger
etal. (200 Ib)




Paquette et al.
(1994)




Peden and
Dailey (1995)


H
W
0.3
2.0
588
3,920
72 h
3h
Mice (C57BL/6J and
C3H/HeJ)
Greater PMN response in C57BL/6J than in C3H/HeJ after
acute and subacute exposures. Responses of recombinant
mice were discordant and suggested two distinct genes
controlling acute and subacute responses.
Tankersley and
Kleeberger,
(1994)

-------
o

O

to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X

-------
December 1
o
o










^
X
to
oo

M
\^
%
H
1
O
0
!2
O
H
O
0
H
W
o

o
HH
—1
1 3
W

Table AX5-3 (cont'd). Effects of Ozone
Concentration
ppm Hg/m3 Duration Species

1.0-2.0 0-3,920 3h Mice (C57BL/6),
6-8 weeks old and
rats (Wistar),
14-16 weeks old


0.5 980 24 h following a 3 -day Mice, M (B6C3F1) 25 ±
(6 h/day) exposure to 2 g
cigarette smoke

0.5 980 8 h during nighttime Rat, M (Wistar, SD and
F344) 90 days old

0.8 1,568 2hand6h Rats, M (Fisher),
Juvenile (2 months;
180-250 g), Adult
(9 months; 370-420 g),
Old (18 months;
375-425 g), Senescent
(24 months; 400-450 g)


0.8 1,568 3h Rat, M (SD)
6-8 weeks old


0.8 1,568 3h Rat, M (SD)
6-8 weeks old


0.8 1,568 3h Rat, M (SD)
200-225 g

on Lung Permeability and Inflammation
Effects3

Steady state MCP-1 mRNA increase after 0.6 ppm, with
maximal increase after 2 ppm. After 2 ppm, MIP-2 mRNA
peaked at 4 h PE and MCP-1 mRNA peaked at 24 h PE.
BALF neutrophils and monocytes peaked at 24 and 72 h PE,
respectively. BALF MCP-1 activity induced by O3 was
inhibited by ananti-MCP-1 antibody.
BALF protein, neutrophils and lymphocytes were increased
in animals exposed to smoke and then to O3. Macrophages
from this group also responded with greater release of TNF-
a upon LPS stimulation
Exposure resulted in a significantly greater injury,
inflammation and BALF levels of IL-6 in Wistar than in SD
orF344 rats.
Comparable effect on the leakage of alveolar protein in rats
of different age groups, but a greater increase occurred in
interleukin-6 and N-acetyl-beta-D-glucosaminidase in
senescent animals than in juvenile and adult rats.





Increased adhesion of macrophages from exposed animals to
rat alveolar type II epithelial cells in culture. Treatment with
anti-TNF-a + anti-IL-la antibody decreased adhesion in
vitro, but not permeability in vivo
Increase in fibronectin protein in BALF and lung tissue, and
fibronectin mRNA in lung tissue. The increase produced by
O3 was amplified in animals pre-treated intra-tracheally with
rabbit serum to induce inflammation.
Treatment of animals with IL-10 prior to O3 exposure caused
a reduction in O3 induced BALF protein, albumin and
fibronectin and tissue fibronectin mRNA

Reference

Zhao et al.
(1998)




Yu et al.
(2002)


Dye et al.
(1999)

Vincent et al.
(1996).







Pearson and
Bhalla (1997)


Gupta et al.
(1998)


Reinhart et al.
(1999)


-------
o

O

to
O
O
                      Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
            Concentration
          ppm
                                     Duration
                                                   Species
                                                                                 Effects3
                                                                                                         Reference
X
to
VO
fe
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
0.8
0.8
0.8
           1.0
           1.0
          0.2
          0.5
          1.0
1,568
8h
          0.8        1,568      48 h

         1.0-2.0   1,960-3,920   3 h
                     1,568      8h
                     1,568      3 h
          1,960
          1,968
           392
           980
          1,960
          5 min exposure of
          airway segments
          following
          bronchoscopy

          8h
          In vitro at liquid/air
          interface
Monkey (Rhesus),
3 years 8 months-3 years
10 month old
                                           Rat, M (SD)
                                           6-8 weeks old
                                 Monkeys, M (Rhesus),
                                 3 years 8 months-3 years
                                 10 months old
                                 (5.1-7.6 kg)

                                 Mice, F
                                 (C57BL6X129NOSII
                                 knockout and wild-type
                                 B6J129SVF2)
                       Dogs, M (Mongrel)
                       Adult
                       Monkeys (Rhesus)
                       Primary TBE, BEAS-2b
                       SandHBEl
                                                                               Pretreatment of monkeys with a monoclonal anti-CD 18        Hyde et al.
                                                                               antibody resulted in a significant inhibition of O3-induced      (1999)
                                                                               neutrophil emigration and accumulation of necrotic airway
                                                                               epithelial cells.

                                                                               Cyclophosphamide treatment ameliorated O3-induced BALF   Bassett et al.
                                                                               neutrophils and albumin after short term and 1-day            (2001)
                                                                               exposure.  Anti-neutrophil serum reduced lavageable
                                                                               neutrophils but did not affect permeability.

                                                                               Trachea! epithelium of exposed animals expressed b6          Hyde et al.
                                                                               integrin.  The integrin expression was reduced or             (1999)
                                                                               undetectable in animals treated with CD-18 antibody.
                         Alveolar macrophages from O3 exposed wild-type mice        Fakhrzadeh
                         produced increased amounts of NO, peroxynitrite,             et al. (2002)
                         superoxide anion, and PGE2. Nitrogen intermediates were
                         not produced and PGE2 was at control level in exposed
                         NOSH knockout mice. These mice were also protected from
                         O3-induced inflammation and injury.

                         Mast cells from O3-exposed airways of ascaris sensitive       Spannhake
                         dogs released significantly less histamine and PGD2           (1996)
                         following in vitro challenge with ascaris  antigen or calcium
                         ionophore.

                         Increase in steady state IL-8 mRNA in airway epithelium.      Chang et al.
                         Increase in IL-8 protein staining declined at 24 h after         (1998)
                         exposure.

                         Dose related increase in IL-8 release in the conditioned
                         media.  Ozone produced greater toxicity  in cell lines than in
                         primary cultures.

-------
o

O

to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X

-------
o

O

to
O
O
Table AX5-3 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
X
fe
H

6
o


o
H

O


O
H
W

O


O
HH
H
W
Concentration
ppm Hg/m3 Duration Species
1.0 1,960 8h/nightforthree Mice, (C57B1/6 wild-
nights type and iNOS
knockout)

1.0 1,960 4h Mice, M (129 strain,
wild-type and clara cell
secretory protein-
deficient), 2-3 mo old
1.2 2,352 6h Rat, M (BN), 200-250 g




2.0 3,920 3h Mice, M (C57BL/6J)



2.0 3,920 3h Rat, F (SD)
6-8 weeks old


2.0 3,920 3h Rat, F (SD)
6-8 weeks old



2.0 3,920 4h Rat, M (Wistar),
200-225 g



Effects3
O3 exposure produced greater injury, as determined by
measurement of MIP-2, matrix metalloproteinases, total
protein, cell content and tyrosine nitration of whole lung
protein, in iNOS knockout mice than in wild-type mice.
Increases in abundance of mRNAs encoding eotaxin, MIP-
la and MIP-2 in CCSP-/-, but no change in wild-type mice.


Eotaxin mRNA expression in the lungs increased 1.6-fold
immediately after and 4-fold at 20 h. Number of lavageable
eosinophils increased 3- and 15-fold respectively at these
time points. Alveolar macrophages and bronchial epithelial
cells stained positively for eotaxin.
O3 -induced increase in protein and PMNs in BALF, and
pulmonary epithelial cell proliferation were significantly
reduced in animals pre-treated with UK-74505, a platelet
activating factor-receptor antagonist.
BALF cells from exposed animals released 2 to 3 times
greater IL-1 and TNF-cc, and greater fibronectin.
Immunocytochemistry showed greater staining of these
mediators in lung tissue from exposed rats.
Increase in BALF macrophage number and total protein.
Increase in iNOS expression, and increase in Fibronectin
and TNF-cc production by alveolar macrophages. O3 effects
were reduced by pretreatment with gadolinium chloride, a
macrophage inhibitor.
A transient increase in tissue neutrophils correlated with an
elevation and subsequent decline in airway
hyperresponsiveness. Pretreatment of rats with anti-
neutrophil serum protected the animals from O3-induced
airway hyperresponsiveness
Reference
Kenyon et al.
(2002)


(Johnston et al.
(1999b)


Ishii et al.
(1998)



Longphre et al.
(1999)


Pendino et al.
(1994)


Pendino et al.
(1995)



DeLorme et al.
(2002)




-------
o

O

to
O
O
                              Table AX5-3 (cont'd).  Effects of Ozone on Lung Permeability and Inflammation
X

-------
 1      towards control even with continuing exposure.  Van Bree et al. (2002) reported adaptation of
 2      rats to O3 following 5 days of exposure.  Animals exposed for 5 days had lower BALF proteins,
 3      fibronectin, IL-6, and inflammatory cells than animals exposed for 1 day. Postexposure
 4      challenge with single O3 exposures at different time points showed that a recovery of
 5      susceptibility to O3 (as measured by BALF levels of albumin, IL-6, and the number of
 6      macrophages and neutrophils) occurred at -15-20 days, but total protein and fibronectin levels
 7      remained attenuated even at 20 days post-5-day exposure. The recovery with regards to BrdU
 8      labeling occurred in  5-10 days after the 5 day exposure.  McKinney et al. (1998) investigated the
 9      role of IL-6 in the adaptive response induced by repeated O3 exposures and observed a
10      significant increases in IL-6 levels following a nighttime exposure of rats to O.Sppm O3 as
11      compared to a daytime exposure.  The kinetics of inflammation were similar following these
12      exposures, but a second exposure subsequent to the nighttime exposure resulted in lesser
13      inflammation than an exposure subsequent to a daytime exposure. Pretreatment of rats with an
14      anti-IL-6  receptor antibody prior to the nighttime exposure abolished O3-induced adaptation with
15      regards to IL-6.
16           The time course of the influx of PMNs into the lung and the BALF fluid levels of
17      macrophage inflammatory protein-2 were found to be roughly similar to that for proteins (Bhalla
18      and Gupta, 2000). Adherence of neutrophils to pulmonary vascular endothelium is maximal
19      within 2 h after exposure and returns to control levels by 12 h PE (Lavnikova et al., 1998).  In an
20      in vitro system utilizing rat alveolar type II cell monolayers, O3 produced a dose-dependent
21      increase in permeability (Cheek et al.,  1995). At higher O3 levels, neutrophils exacerbated the
22      injury, but their presence after the exposure expedited restoration of epithelial barrier.
23           Vesely et al. (1999a) have demonstrated that neutrophils contribute to the repair process in
24      O3 -injured airway epithelium. When rats were depleted of neutrophils by rabbit anti-rat
25      neutrophil serum, and exposed to  1 ppm O3 for 8 h, epithelial necrosis in the nasal cavity,
26      bronchi, and distal airways were observed, suggesting a role for neutrophils in repair processes.
27      Proliferation of terminal bronchiolar epithelial cells, as assessed by BrdU-incorporation, was
28      also decreased by O3 exposure, suggesting a role for neutrophils in this process.
29
        December 2005                          AX5-33       DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX5.2.3.2  Concentration and Time of Exposure
 2           The relative influence of concentration and duration of exposure (i.e., C x T) has been
 3     investigated extensively in rats, using BALF protein as an endpoint. Though the interaction
 4     between C and T is complex, concentration generally dominated the response. The impact of
 5     T was C-dependent (at higher Cs, the impact of T was greater); at the lowest C and T values, this
 6     dependence appeared to be lost.
 7           New studies evaluating C x T relationships have not been found.
 8
 9     AX5.3.3.3  Susceptibility Factors
10           Factors that have been studied for potential impact on the effects of O3 include age, gender,
11     nutritional status, genetic variability, exercise and exposure to co-pollutants.
12           The effects of age on lung inflammation are not well known.  Vincent et al. (1996) found
13     O3 did not differentially  affect the leakage of alveolar protein in rats of different age groups, but
14     an O3-induced increase in IL-6 and N-acetyl-beta-D-glucosaminidase (NAG) was observed in
15     senescent animals compared to juvenile and adult rats. Johnston et al. (2000b) compared gene
16     expression of chemokines and cytokine in newborn and 8-week-old C57B1/6J mice exposed to
17     1.0 or 2.5 ppm O3 for 4,  20, or 24 h.  The animals were killed immediately after exposure, total
18     RNA was isolated from lung tissue, ribonuclease protection assays were completed for a number
19     of cytokines/chemokines including IL-12, IL-10, IL-la, IL-lp, IL-IRa, MIF, IFN-y, MlP-la,
20     MIP-2,IL-6, and Mt.  The newborn mice displayed increased levels of Mt mRNA only, while the
21     8-week-old mice had increases in MlP-la, MIP-2, IL-6, and Mt mRNA. Comparisons were
22     made with mice  of the same age groups with exposures to endotoxin (10 min). Both age groups
23     displayed similar cytokine/chemokine profiles with endotoxin exposure. This suggested to the
24     authors that the responses to endotoxin, which does not cause epithelial injury, and the responses
25     to O3, which does, demonstrate that differences in inflammatory control between newborn and
26     adult mice is secondary to epithelial injury.
27           Ascorbate deficiency had been found to have only minimal effect on injury and
28     inflammation in  guinea pigs exposed to O3 (Kodavanti et al., 1995). Elsayed (2001)
29     demonstrated that a general dietary restriction to 20% of the freely-fed diet for 60 days caused an
30     extreme reduction in body weight in 1-month-old SD rats. These rats, exposed to 0.8 ppm
31     continuously for 3 days, had levels of antioxidants and detoxifying enzymes that were increased

       December 2005                          AX5-34     DRAFT-DO NOT QUOTE OR CITE

-------
 1      less than in freely fed animals.  Lung injury, as detected by cell proliferation in CAR and BALF
 2      levels of protein, number of neutrophils, was increased in sidestream cigarette smoke exposed
 3      mice that were subsequently exposed to O3 (Yu et al., 2002). Macrophages from smoke + O3
 4      exposed animals also responded by a greater release of TNF-a following LPS stimulation when
 5      compared to macrophages exposed to air, smoke or O3 alone.
 6           Recent lines of evidence illustrate the importance of genetic susceptibility in O3 health
 7      effects. The effects of acute and subacute exposures were studied by Tankersley and Kleeberger
 8      (1994) in inflammation-prone (susceptible) C57BL/6J(B6) and inflammation-resistant
 9      C3H/HeJ(C3) strains of mice. Based on the neutrophilic response to O3 in these two strains and
10      in recombinant mice, the authors concluded that the acute and subacute exposures are controlled
11      by two distinct genes, referred to as Inf-land Inf-2 respectively.  Exposures, when repeated
12      fourteen days after the initial exposures, caused a smaller increase in BALF protein and number
13      of macrophages, lymphocytes and epithelial cells in both strains, but PMN number was greater
14      in both strains compared to initial exposure (Paquette et al., 1994).
15           Further studies by Kleeberger et al. (1997) identified another potential  susceptibility gene,
16      tumor necrosis factor (TnJ), on a qualitative trait locus on mouse chromosome 17. Tnf codes for
17      the pro-inflammatory cytokine TNF-a. By neutralizing the function of TNF-a with  a specific
18      antibody, they were able to confer protection against O3 injury in susceptible mice. Cho et al.
19      (2001) demonstrated a role for tumor necrosis factor receptor  1 and 2 (TNFR1 and TNFR2,
20      respectively) signaling in subacute (0.3 ppm for 48  h) O3-induced pulmonary epithelial injury
21      and inflammation.  TNFR1 and TNFR2 knockouts were less sensitive to subacute O3 exposure
22      than wild-type C57BL/6J mice.  With acute exposures to O3 (2 pm for 3 h), airway
23      hyperreactivity was diminished in knockout mice compared to wild-type mice, but lung
24      inflammation and permeability were increased. Based on these studies, it has been
25      hyposthesized that in subacute O3 exposures, TNF-a is a susceptibility gene in mice, and further,
26      that independent mechanisms control lung inflammation and permeability. Further evidence for
27      the mechanistic separation of hyperresponsiveness and PMN infiltration was provided by Shore
28      et al. (2001) from studies using wild-type and TNFR knockout mice exposed to 2 ppm O3 for
29      3 h. Numbers of PMN  in BAL collected 21 h PE were not changed in air exposures, but were
30      increased to the same extent in both wild-type and TNFR knockout mice, whereas
31      hyperresponsiveness was increased only in wild-type mice, but not in the knockouts.

        December 2005                          AX5-35       DRAFT-DO NOT QUOTE OR CITE

-------
 1          Prows et al (1999) characterized the differences between A/J (O3-sensitive) and C57BL/6J
 2      (O3-resistant) mice exposed to 10 ppm O3, which continued until the animals died. Sensitive
 3      animals survived < 13 h and resistant mice survived > 13 h. Histological examination of the
 4      lungs and wet-to-dry lung weight ratios did not differ between the two strains. Though the dose
 5      given was not environmentally relevant, the genome wide scans confirmed earlier findings
 6      (Prows et al. 1997) that two loci (acute lung injury-1 and -3, Alii and AH3, respectively) on
 7      chromosome 11 control susceptibility to death after O3 exposure.
 8          A recent comprehensive characterization of lung injury in nine inbred mouse strains (Savov
 9      et al., 2004) has been summarized in Table AX5-4. The exposure consisted of 3 h of 2.0 ppm O3,
10      followed by room air, for 6 h or 24 h PE and analyzed for plethysmography, MCh challenge,
11      BALF, histology, and single nucleotide polymorphisms (SNPs). This group identified two strains
12      (C3H/HeJ and A/J) as consistently O3-resistant, and two strains (C57BL/6J and 129/SvIm) as
13      consistently O3-vulnerable. Five strains were characterized as having inconsistent phenotypes
14      with intermediate responses to O3 (CAST/Ei, BTBR, DBA/2J, FVB/NJ, and BALB/cJ).  Their
15      in silico genome scan identified on chromosome 1, a 170- to  189-Mb region associated with the
16      6-h airway hyperreactivity response and the 24-h inflammatory response; on chromosome 7,  a 30-
17      to 40-Mb region associated with the 6-h inflammatory response and the 6-h protein increase; and
18      on chromosome 17, a 30- to 40-Mb region associated with the 24-h airway hyperreactivity
19      response. They found no consistent correlation between the concentration of total protein in
20      BALF and influx of inflammatory cells (PMN), which they attribute to be regulated by different
21      genes. They did find a correlation between O3-induced increases in IL-6 and PMN concentration.
22          Ozone-induced changes in CCSP (called CC16 by this group) expression were evaluated in
23      five inbred mouse strains: C57BL/6J and CBA both considered sensitive to acute O3-induced
24      inflammation, C3H/HeJ and AKR/J both considered resistant, and  SJL/J considered intermediate
25      (Broeckaert et al., 2003).  Two exposures paradigms were used, 1.8 ppm O3 for 3 h or 0.11 ppm
26      O3, 24/h day for up to 3 days, and BALF and serum were assayed immediately after exposure or
27      at 6 h PE. Both exposure levels caused a transient increase in CC16 in serum that correlated with
28      BALF changes in protein, LDH, and inflammatory cells. There was an inverse relationship
29      between preexposure levels of CC16 in BALF and epithelial  damage based on serum CC16 levels
30      and BALF markers of inflammation. There was also an inverse relationship between
       December 2005                          AX5-36      DRAFT-DO NOT QUOTE OR CITE

-------
Table AX5-4.  Inbred Mouse Strain Susceptibility
Mouse Strain
C57BL/6J
129/SvIm
BTBR
BALB/cJ
DBA/2J
A/J
FVB/NJ
CAST/Ei
C3H/HeJ

Baseline
hyporeactive
hyperreactive
hyperreactive
intermediate
intermediate
very hyperreactive
intermediate
intermediate
intermediate
Penh
O3 only
susceptible
susceptible
susceptible
susceptible
resistant
resistant
resistant
resistant
resistant
PMN Protein
OjthenMCh 6h 24h 6h 24h
much more t 1 11 ns ns
responsive
more responsive 1 1 1 1 t ns
more responsive 1 1 1 1 1 1 1 1
more responsive t ns t ns ns t
less responsive 11 1 ns 1
much less 1 1 1 1 ns ns
responsive
less responsive 11 It t 1
less responsive 1 1 1 1 1 1 t 1 1
less responsive t I ns I
IL-6 % PCNA
6h 24h 24h
II 1 >4
It t >2
H t t < 1
t ns <1
1 1 >1
It ns > 4
1 ns > 1
H t t < 1
II ns < 1
Overall
Response
toO3
highly
sensitive
highly
sensitive
intermediate
intermediate
intermediate
highly
resistant
intermediate
intermediate
highly
resistant

-------
 1      preexposure levels of albumin in BALF and lung epithelium damage. Based on these results, the
 2      authors conclude that a major determinant of susceptibility to O3 is basal lung epithelial
 3      permeability. As all of the mouse strains had similar levels of preexposure CC16 mRNA, they
 4      explored the possible role of CC16 isozymes in differences among strains.  The CC16 monomer a
 5      7kD protein exist in two isoforms with differing pi values, CC16a (4.9) and CC16b (5.2).
 6      To evaluate the role of CC16 isoform profiles in permeability differences between C57BL/6J and
 7      C3H/HeJ, this group  evaluated the CC16 protein profiles in BALF of the strains before and after
 8      O3 exposure following two-dimensional protein electrophoresis analysis. C57BL/6J mice had
 9      lower levels of CC16a (the more acidic form) than C3FI/HeJ. But both the strains had similar
10      levels of CC16b. Based on these observations Broeckaert et al (2003) conclude that greater
11      epithelial permeability observed in C57BL/6J may be due to difference in the expression of
12      CC16a and possibly other antioxidant/inflammatory proteins.
13           Wattiez et al. (2003) examined BALF protein from C57BL/6J (O3-sensitive) and C3H/HeJ
14      (O3-resistant) mice exposed to filtered air using a two-dimensional polyacrylamide gel approach
15      to analyze the protein profiles.  C3H/HeJ mice expressed 1.3 times more Clara cell protein 16
16      (CC16) than C57BL/6J mice, and further, expressed more of the acidic isoform of CC16.  Strain-
17      specific differential expression of isoforms of the antioxidant protein 2 (AOP2), the isoelectric
18      point 5.7 isoform in C3Ft/HeJ and isoelectric point 6.0 isoform in C57BL/6J were observed.
19      These studies suggested a potential role for the strain-specific differential expression in their
20      protein toward differential  susceptibility to oxidative stress.
21
22      AX5.2.3.4  Mediators of Inflammatory Response and Injury
23           While neutrophils in the lung characterize an inflammatory response to O3, the release of
24      chemotactic mediators by inflammatory cells indicates their state of activation and their role in
25      continued inflammation and injury. Studies in recent years have placed a greater focus on these
26      mediators to understand the mechanisms implicated in O3-induced inflammation and injury.
27      Cytokines and chemokines have been shown to be released as a result of stimulation or injury of
28      macrophages, epithelial cells and PMNs.  Many  of these mediators have been implicated in PMN
29      recruitment in the lung following O3 exposure.  The expression of macrophage inflammatory
30      protein 2 (MIP-2) mRNA or BALF levels of MIP-2 increased in mice and rats exposed to O3
31      concentrations equal  to or greater than  1 ppm (Driscoll et al. 1993; Haddad et al., 1995; Bhalla

        December 2005                          AX5-38       DRAFT-DO NOT QUOTE OR CITE

-------
 1      and Gupta, 2000).  The increased mRNA expression was associated with an increased
 2      neutrophilia in the lung. Ozone exposure also caused an increase in monocyte chemotactic
 3      protein-1 (MCP-1) mRNA in mice and rats (Zhao et al., 1998). These studies implicate MCP-1 in
 4      O3-induced monocyte accumulation in the lung and suggest a role of NFkB in MCP-1 gene
 5      expression. Fibronectin, an extracellular matrix glycoprotein, has been studied for its role in lung
 6      inflammation and inflammatory disorders.  Gupta et al. (1998) observed an increase in both
 7      fibronectin protein and mRNA expression in the lung of rats exposed to 0.8 ppm O3.  A
 8      mechanistic role of fibronectin in O3-induced inflammation and injury was suggested on the basis
 9      of comparability of temporal changes in B ALF protein, fibronectin and alkaline phosphatase
10      activity (Bhalla et al., 1999). Numerous studies have reported O3-induced differential expression
11      of various cytokines and inflammatory mediators both in vivo and ex vivo: increased expression
12      of cytokine-induced neutrophil chemoattractant (CINC) and NF-KB expression in vivo (Haddad
13      et al., 1996; Koto et al., 1997); IL-8 both in vivo and in vitro (Chang et al., 1998); tumor necrosis
14      factor (TNF-a), fibronectin, interleukin-1 (IL-1), and CINC release by macrophages  ex vivo
15      (Pendino et al., 1994; Ishii et al., 1997), and NF-aB and TNF-a (Nichols et al., 2001). An
16      increase in lung CINC mRNA occurred within 2 h after the end of a 3 h exposure of rats to 1 ppm
17      O3. The CINC mRNA expression was associated  with neutrophilia at 24 h post-O3 exposure.
18      Exposure of guinea pig alveolar macrophages recovered in BALF and exposed in vitro to 0.4 ppm
19      O3 for 60 minutes produced a significant increase  in IL-6 and TNF-a (Arsalane et al., 1995). An
20      exposure of human AMs to an identical O3 concentration increased TNF-a, IL-lb, IL-6 and IL-8.
21      This exposure also caused an increase in mRNA expression for TNF-a, IL-lb, IL-6 and IL-8 in
22      human cells. Ozone exposure caused an increase in IL-6, MlP-la, MIP-2, eotaxin and
23      metallothionein (MT) expression in mice (Johnston et al.,  1999a). The IL-6 and MT increase was
24      enhanced in mice deficient in Clara cell secretory  protein (CCSP), suggesting a protective role of
25      Clara cells and their secretions (Mango et al., 1998).  CCSP deficiency also increased sensitivity
26      of mice to O3, as determined by an increase in abundance of MlP-la and MIP-2 following a 4 h
27      exposure (Johnston et al., 1999b).
28           A role for mast cells in airway responses is proposed on the basis of chronic exposure
29      studies demonstrating greater increases in lavageable macrophages,  epithelial cells and PMNs in
30      mast cell-sufficient mice compared to mast cell-deficient mice exposed to 0.26 ppm O3
31      (Kleeberger et al., 2001b). Similar results were earlier reported by the same group using a higher

        December 2005                           AX5-39      DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3 concentration (Longphre et al., 1996). Increases in inflammatory cells were also observed in
 2      mast cell-deficient mice repleted with mast cells, but O3-induced permeability increase was not
 3      different in genotypic groups exposed to 0.26 ppm. When a mast cell line was exposed to varying
 4      O3 concentrations, spontaneous release of serotonin and modest generation of PGD2 occurred
 5      only under conditions that caused cytotoxicity (Peden and Dailey, 1995). Additionally, O3
 6      inhibited IgE- and A23187-induced degranulation. Mast cells recovered from O3-exposed
 7      peripheral airways of ascaris sensitive dogs released significantly less histamine and PGD2
 8      following in vitro challenge with ascaris antigen or calcium ionophore (Spannhake, 1996).  Ozone
 9      exposure also promoted eosinophil recruitment in the nose and airways in response to instillation
10      of ovalbumin or ovalbumin-pulsed dendritic cells and aggravated allergy like symptoms in guinea
11      pigs (lijima et al., 2001).
12           Treatment of rats with cyclophosphamide prior to O3 exposure resulted in a decreased
13      recovery of PMNs in the BALF  and attenuated permeability induced by O3 (Bassett et al., 2001).
14      Additionally, they found that pretreatment of animals with antiserum against rat neutrophils
15      abrogated PMN accumulation in the lung, but did not alter permeability changes produced by O3.
16      DeLorme et al. (2002) showed a relationship between neutrophilic inflammation and airway
17      hyperresponsiveness.  Treatment of rats with anti-neutrophil serum protected the animals from
18      O3-induced airway hyperresponsiveness. Studies utilizing antibodies to selected pro- or anti-
19      inflammatory cytokines suggest a role of TNF-a, interleukin-10 (IL-10) and IL-lb in O3-induced
20      changes in permeability, inflammation and cytokine release (Ishii et al., 1997; Reinhart et al.,
21      1999; Bhalla et al., 2002). An attenuation of O3-induced increase in permeability and
22      inflammation was also observed in mice treated, either before or after exposure, with UK-74505,
23      a platelet-activating factor receptor antagonist (Longphre et al., 1999), suggesting that O3-induced
24      epithelial and inflammatory changes are mediated in part by activation of PAF receptors.
25           Ozone exposure stimulates macrophage motility towards a chemotactic gradient, and
26      macrophages from rats exposed  to 0.8 ppm O3 adhered to epithelial  cells (ARL-14) in culture to a
27      greater extent than macrophages from air-exposed controls (Bhalla,  1996). Both macrophage
28      motility and chemotaxis were attenuated by antibodies to cell adhesion molecules CD-I Ib and
29      ICAM-1.  An exposure of female rats to O3 had an attenuating effect on CD-18 expression on
30      alveolar macrophages and vascular PMNs, but the expression of CD62L, a member of selectin
31      family, on vascular PMNs was not affected (Hoffer et al., 1999). In monkeys, the O3-induced

        December 2005                          AX5-40      DRAFT-DO NOT QUOTE OR CITE

-------
 1      inflammation was blocked by treatment with a monocolonal antibody to CD 18, suggesting
 2      dependence of PMN recruitment on this adhesion molecule (Hyde et al., 1999). Treatment of
 3      monkeys with CD 18 antibody also reduced tracheal expression of the beta6 integrin (Miller
 4      et al., 2001). A single 3 h exposure of rats to O3 resulted in an increase in neutrophil adhesion to
 5      epithelial cells in culture (Bhalla and Young, 1992) and caused an elevation in concentration of
 6      ICAM-1, but not CD-18, in the BALF (Bhalla and Gupta, 2000). Takahashi et al. (1995a) found
 7      an increase in tissue expression of ICAM-1 in mice exposed to 2 ppm O3. They noted a temporal
 8      correlation of inflammatory activity and ICAM-1 expression which varied in different regions of
 9      the lung. A comparable pattern of time-related changes in total protein, fibronectin and alkaline
10      phosphatase activity in the BALF of rats exposed to 0.8 ppm O3 was also observed by Bhalla
11      etal. (1999).
12
13      AX5.2.3.5  Role of Nitric Oxide Synthase and Reactive Nitrogen in Inflammation
14          An acute exposure of rats to 2 ppm O3 caused an increase in the expression of iNOS activity
15      with an increase in BALF macrophage number and total protein and increase in fibronectin and
16      TNF-a production by AMs (Pendino et al.,  1995). All of these effects of O3 were reduced by
17      pretreatment with gadolinium chloride, a macrophage inhibitor. Macrophages isolated from O3-
18      exposed mice produced increased amounts of nitric oxide, superoxide anion and PGE2, but
19      production of these mediators by macrophages from NOS knockout mice was not elevated
                                                                                     s~<,
20      (Fakhrzadeh et al., 2002).  Additionally, mice deficient in NOS or mice treated with N -
21      monomethyl-L-arginine, an inhibitor of total NOS, were protected from O3-induced permeability,
22      inflammation and injury, suggesting a role of nitric oxide in the production of O3 effects
23      (Kleeberger et al., 2001b; Fakhrzadeh et al., 2002). Another study demonstrated greater injury
24      (as determined by measurement of MIP-2, matrix metalloproteinases, total protein, cell content
25      and tyrosine nitration of whole lung protein) in iNOS knockout mice than in wild-type mice on O3
26      exposure (Kenyon et al., 2002).  They proposed that protein nitration differences are related to
27      inflammation and may not be dependent on iNOS-derived NO.
28          Ishii et al. (2000a) performed studies using pretreated rats with ebselen (a potent anti-
29      inflammatory, immunomodulator and NO/peroxynitrite scavenger) and then exposed to 2 ppm O3
30      for 4 h. The pretreated rats had decreased numbers of neutrophils, lowered albumin levels, and
31      inhibited nitration of tyrosine residues in BALF 18 h PE, without changes in macrophage iNOS

        December 2005                           AX5-41       DRAFT-DO NOT QUOTE OR CITE

-------
 1      expression. These results suggest that an iNOS-independent mechanism may be involved in O3-
 2      induced inflammation. Inoue et al. (2000) demonstrated in human transformed bronchial
 3      epithelial cells that NO-generating compounds (TNF-a, IL-lp, and INF-y) induce IL-8
 4      production and that NOS inhibitors inhibit IL-8 production.  In vivo experiments in the same
 5      study using male Hartley-strain guinea pigs exposed to 3 ppm O3 for 2 h showed that NOS
 6      inhibitor pretreatment attenuated-O3 induced neutrophil recruitment and airway
 7      hyperresponsiveness at 5 h after exposure.  The NOS inhibitors also blunted the increase in
 8      nitrate/nitrite levels and in IL-8 mRNA, at the 5 h PE.  The authors hypothesize that NO, or its
 9      derivatives, facilitate airway hyperresponsiveness and inflammation after O3 exposure, possibly
10      mediated by IL-8.  Jang et al. (2002) have attempted to characterize the mechanism by which
11      short-term O3 exposures (0.12, 0.5, 1, or 2 ppm for 3 h) cause airway inflammation and
12      responsiveness in BALB/c mice. Using a modified Griess reaction, measurement of nitrate and
13      nitrite in BAL fluid after O3 exposure showed dose-dependent increases in nitrate, which is
14      indicative of in vivo NO generation. Functional studies of enhanced pause (Penh) demonstrated
15      increases with O3 which were also dose-dependent.  Western blot analysis of lung tissue showed
16      increases in NOS-1, but not in NOS -3 or iNOS isoforms.  The authors conclude that in mice
17      NOS-1 may induce airway responsiveness by a neutrophilic airway inflammation.
18
19      AX5.2.4  Morphological Effects
20      AX5.2.4.1  Introduction
21           All  laboratory animal species studied to date show generally similar morphological
22      responses to < 1 ppm O3. The precise  characteristics of the structural changes due to O3 are
23      dependent on  the exposure regimen, time of examination, distribution of sensitive cells, and the
24      type of centriacinar region (i.e., junction between the end of the terminal bronchioles and the first
25      few generations of either respiratory bronchioles or alveolar ducts, depending on the species).
26           The presentation of this morphology section will begin with effects of short-term exposure
27      (<  1 week). The subsequent discussion of long-term exposure effects (>1  week) is an artificial
28      separation, but consistent with the division of studies in other effects sections. The long-term
29      studies that included evaluation of various durations of exposures are presented totally in this
30      subsection to illustrate the effects of exposure duration and to provide a better understanding of
31      possible chronic effects of multi-seasonal, ambient O3 exposures in the population (Chapter 7).

        December 2005                          AX5-42      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX5.2.4.2  Short-Term Exposure Effects on Morphology
 2           Morphological effects of key new exposure studies generally lasting less than 1 week are
 3      summarized in Table AX5-5. The following discussion is by region of the respiratory tract first,
 4      followed by exposure and susceptibility factors.
 5           Hotchkiss et al. (1998) explored the efficacy of a topical anti-inflammatory corticosteriod,
 6      fluticasone propionate (FP), in preventing the inflammation and mucous cell metaplasia in rats
 7      after cumulative O3 exposure.  Male F-344 rats were exposed to filtered air or 0.5 ppm O3,
 8      8 h/day, for 3 or 5 days. Immediately before and after exposure, the rats were given FP (25 jig)
 9      by intranasal instillation (50 jiL/nasal passage) or an equivalent amount of vehicle only. Nasal
10      tissues were processed for light microscopy 2 h and 3 days after the 3- and 5-day exposures,
11      respectively. Rats treated with FP had 30 to 60 % less nasal inflammation after 3 and 5 days of
12      O3 exposure and 85% less mucous cell metaplasia after the 5-day exposure compared with
13      vehicle-instilled, O3-exposed controls.
14           Fanucchi et al. (1998) reported that exposure to bacterial endotoxin, a common ambient air
15      toxin, can potentiate mucous cell metaplasia in the nasal transitional epithelium of rats caused by
16      a previous O3 exposure. Male F344/N Hsd rats were intranasally instilled with saline or
17      100 |ig/ml endotoxin after exposure to filtered air or 0.5 ppm O3, 8 h/day for 3 days, and
18      evaluated 6 h and 3 days PE.  Mucous cell metaplasia was not found in the air/endotoxin group,
19      but was found in the O3/saline group and was most severe in the O3/endotoxin group.  A similar
20      synergistic effect of O3 and endotoxin on the nasal epithelium was reported in Fischer rats by
21      Wagner et al. (2001a,b). When exposed to O3 alone (0.5 ppm, 8 h/day for 3 days), rats developed
22      epithelial lesions in the nasal transitional epithelium. Exposure to endotoxin alone (20 jig) caused
23      lesions in the respiratory epithelium of the nose and conducting  airways. Endotoxin enhancement
24      of the O3-induced mucous cell metaplasia was related to neutrophilic inflammation.
25           Cho et al. (1999a, 2000) reported that O3-induced mucous cell metaplasia in the transitional
26      epithelium of rats may be dependent on pre-metaplastic responses, such as mucin mRNA
27      upregulation, neutrophilic inflammation, and epithelial proliferation. Male F344/N rats were
28      exposed to 0.5 ppm O3, 8 h/day for 1, 2, or 3 days and 2 h, or 1, 2, or 4 days PE were assayed for
29      the parameters listed above. A rapid increase in an airway-specific mucin gene (rMuc-5AC
30      mRNA) occurred rapidly after exposure to O3, both before and during the onset of mucous cell
31      metaplasia.  Neutrophilic inflammation coincided with epithelial DNA synthesis and

        December 2005                           AX5-43     DRAFT-DO NOT QUOTE OR CITE

-------
                                  Table AX5-5. Effects of Ozone on Lung Structure:  Short-Term Exposures

to
O
o
X
Concentration
ppm
0.1
0.5
1.0



0.2
0.4






0.2
0.4
0.8


0.4



0.4
1.0
Hg/m3
196
980
1,960



392
784






392
784
1,568


784



784
1,960
Duration
8 h/ day x 1 day
8 h/day x 1 day
8 h/day x 1, 10,
75, and 90 days


3, 7, 28, and
56 days; 3-, 7-,
and 28-day
recovery from
28 days of
exposure


23 h/day for
7 days



12 h/day;
1- or 7-day
exposure

2h

Species
Rat;
male;
Sprague-Dawley



Mouse;
male; Nffl;
Rat;
male; Wistar
RIV:Tox
Guinea pig;
male; Hartley
Crl:(HA)BR
Guinea pig;
female; Hartley;
±AH2 diet


Rat;
(Wistar RiV:TOX;
M&F; 1,3,9, &
18 months of age
Monkey;
adult male rhesus
Effects3
No dose-related response on CYP2E1, one of six P450 enzymes identified
in respiratory tissue. CYP2E1 activity was elevated (250% and 280%) in the
lobar bronchi / major daughters airways immediately after 1.0 ppm O3 exposure
for Iday and 10 days, respectively, but not in the trachea or distal bronchioles;
CYPE1 activity was unchanged and decreased after 1.0 ppm O3 exposure for 75
and 90 days, respectively.
Concentration-related centriacinar inflammation, with a maximum after 3 days
of exposure; number of alveolar macrophages and pulmonary cell density
increased progressively until 56 days of exposure, with the guinea pig the most
sensitive species. Concentration and exposure-time dependent hypertrophy of
bronchiolar epithelium in mouse only. Exposure to 0.2 ppm for 3 and 7 days
caused significant histological and morphometric changes in all 3 species;
exposure for 56 days caused alveolar duct fibrosis in rat and guinea pigs. Total
recovery in rats after 28-day exposure, but not in guinea pigs or mice .
Treatment-related lesions were observed after exposure to 0.4 and 0.8 ppm O3;
lesions were primarily seen in the terminal bronchioles and consisted
of mononuclear cell and neutrophilic infiltrate and thickening of the
peribronchiolar interstitium. Effects were only marginally exacerbated by the
AH2 (ascorbic acid) deficient diet and lesions were resolved after 1 week in FA.
Centriacinar inflammation (increased alveolar macrophages and PMNs;
increased proximal and ductular septal density) was greatest in young rats
(1 month and 3 months for 1-and 7-day exposures, respectively) and decreased
with age. No major gender differences were noted.
Reduced glutathione (GSH) increased in the proximal intrapulmonary bronchus
after 0.4 ppm O3 and in the respiratory bronchiole after 1.0 ppm O3. Local O3
Reference
Watt et al.
(1998)




Dormans
etal. (1999)






Kodavanti
etal. (1995)



Dormans
etal. (1996)


Plopper et al.
(1998)
                                                            dose (measured as excess 18O) varied by as much as a factor of three in different
                                                            airways of monkeys exposed to 1.0 ppm, with respiratory bronchioles having
                                                            the highest concentration and the parenchyma the lowest concentration. After
                                                            exposure to 0.4 ppm, the O3 dose was 60% to 70% less and epithelial injury was
                                                            minimal, except in the respiratory bronchiole, where cell loss and necrosis
                                                            occurred, but was 50% less than found at 1.0 ppm.

-------
o

O

to
O
O
Table AX5-5 (cont'd). Effects of Ozone on Lung Structure:  Short-Term Exposures
X
Concentration
ppm fig/m3
0.5 980




0.5 980





0.5 980




0.5 980











Duration Species
8 h Rat;
+ BrdU to label male; F344
epithelial cells


8 h/day, Rat;
3 or 5 days; male; F344
+ fluticasone
propionate (FP)
intranasally

8 h/day, Rat;
3 days male; F344/N Hsd
+ endotoxin
(100 ug/mL )
intranasally
8 h/day, Rat;
1, 2, or 3 days male; F344/N
+ BrdU to label
epithelial cells
+ antirat
neutrophil
antiserum





Effects3
O3 exposure induced a transient influx of neutrophils and a significant (17%)
loss of NTE cells 2-4 h after exposure. Increased epithelial DNA synthesis was
first detected 12 h PE. LI and ULLI indices of epithelial cell DNA synthesis
were greatest 20-24 h and still elevated 36 h PE; numeric density of NTE cells
returned to control levels 20-24 h PE.
No significant difference of FP on morphometry of the maxilloturbinates;
O3 exposure caused neutrophilic rhinitis with 3.3- and 1.6-fold more
intraepithelial neutrophils (3-day and 5-day exposure, respectively) and marked
mucous cell metaplasia (5 -day exposure only) with numerous mucous cells and
approximately 60 times more IM in the nasal transitional epithelium; FP-treated
rats exposed to O3 had minimal nasal inflammation and mucous cell metaplasia.
Endotoxin-induced neutrophilia in nasal mucosa with NTE; mucous cell
metaplasia was not detected in air/endotoxin-exposed rats, was observed
in O3/saline-exposed rats, and was most severe in O3/endotoxin-exposed rats.


Acute O3 exposure induced a rapid increase in rMuc-5AC mRNA levels prior to
the onset of mucous cell metaplasia; neutrophilic inflammation coincided with
epithelial DNA synthesis and upregulation, but was resolved when mucous cell
metaplasis first appeared in the NTE.
Maxilloturbinates lined with NTE determined the epithelial labeling
index, numeric densities of neutrophils, total epithelial and mucous secretory
cells, amount of stored intraepithelial mucosubstances, and steady-state
ratMUC-5AC (mucin) mRNA levels. Four days after a 3-d exposure,
antiserum-treated, O3-exposed rats had 66% less stored intraepithelial
mucosubstances and 58% fewer mucous cells in their NTE than did controls.
Antiserum treatment had no effects on O3-induced epithelial cell proliferation or
mucin mRNA upregulation.
Reference
Hotchkiss
etal. (1997)



Hotchkiss
etal. (1998)




Fanucchi
etal. (1998)



Cho et al.
(1999a,
2000)










-------
December
to
o
o

X
v\
|V
4^
ON

Table AX5-5 (cont'd). Effects of Ozone on Lung Structure: Short-Term Exposures
Concentration
ppm fig/m3 Duration

0.5 980 8 h/day,
3 days
+ endotoxin




0.5 980 8 h/day,
1 and 3 days
+ OVA
(1%,
50 uL/nasal
passage)
1 1,960 8 h




Species

Rat;
F-344





Rat;
Brown Norway




Rat;
Sprague-Dawley
Ferret; young male
Monkey; young male
rhesus
Effects3

Enhanced epithelial lesions in the NTE and respiratory epithelium of the nose
and conducting airways by endotoxin and O3 exposures, respectively;
synergistic effects of coexposure mediated by neutrophils. Endotoxin increased
rMuc-5 AC mRNA levels in the NTE of O3 -exposed rats; neutrophil depletion,
however, had no effect on endotoxin-induced upregulation of mucin gene
mRNA levels. Endotoxin enhanced the O3-induced increase in stored
mucosubstances (4-fold increase), but only in neutrophil-sufficient rats
O3 enhanced the appearance of eosinophils in the maxilloturbinates of
OVA-challenged rats but did not increase inflammation in other nasal tissues;
O3/OVA coexposures for 3 days increased the number of epithelial cells as well
as the appearance of mucus-containing cells in the NTE lining the
maxilloturbinates.

Severe, acute infiltration of neutrophils along with necrotic bronchiolar
epithelium in all lung regions, especially in the centriacinar region; necrosis
and inflammation was more severe in ferrets and monkeys than in rats.


Reference

Wagner et al.
(2001a,b)





Wagner et al.
(2002)




Sterner-Kock
et al. (2000)



aAM = Alveolar macrophage.
PE = Postexposure (i.e., time after O3 exposure ceased).
LM = Light microscopy.
EM = Electron microscopy.
RB = Respiratory bronchiole.
TB = Terminal bronchiole.
IAS = Interalveolar septum.
PMN = Polymorphonuclear leukocyte.

-------
 1      upregulation of rMuc-5AC, but was resolved before the development of epithelial metaplasia. In
 2      the follow-up study, the investigators found that only the mucous cell metaplasia was neutrophil-
 3      dependent, whereas O3-induced epithelial cell proliferation and mucin gene upregulation were
 4      neutrophil-independent.
 5           In the centriacinar region., different species have similar responses to low levels of O3
 6      (> 0.2 ppm for 1 week; Dormans et al., 1999). Dormans et al. (1999) compared the extent and
 7      time course of fibrotic changes in mice, rats, and guinea pigs exposed to 0.2 and 0.4 ppm O3 for 3,
 8      7, 28, and 56 days.  They found a concentration-related centriacinar inflammation in all three
 9      species, with a maximum occurring after 3 days of exposure and total recovery within 3 days.
10           The effects of exposure duration are complex and are likely responsible for the similar
11      patterns of biochemical responses (see Section 5.2.1).  Repair of the damage by removal of
12      injured epithelial cells is enhanced by the influx of neutrophils (Hyde et al., 1999; Veseley et al.,
13      1999a; Miller et al., 2001; see Section 5.2.3).
14           Hotchkiss et al.  (1997) reported that labeling indices for rat nasal transitional epithelial cell
15      DNA were greatest 20 to 24 h after an 8-h exposure to 0.5  ppm O3, but still greater than control
16      by36hPE.
17           Exploring the role of susceptibility factors on morphological changes, Dormans et al.
18      (1999) compared morphological, histological, and biochemical effects in the rat, mouse, and
19      guinea pig after O3 exposure and after recovery in clean air. Wistar RIV:Tox male rats, NIH male
20      mice, and Hartley Crl:(HA)BR male guinea pigs were continuously exposed to filtered air, 0.2, or
21      0.4 ppm for 3, 7, 28, and 56 days. Recovery from 28 days of exposure was studied at intervals of
22      3, 7, and 28 days PE. Morphometric analysis was performed only on lung parenchyma with
23      proximal alveoli and smaller alveolar ducts and no distinct species-specific differences were
24      noted.  The mouse was the most sensitive as shown by a concentration and exposure-time
25      dependent persistence of bronchiolar epithelial hypertrophy, elevated lung enzymes, and slow
26      recovery from exposure.  In both rats and guinea pigs, 56 days of exposure to 0.4 ppm O3 caused
27      increased amounts of collagen in ductal septa and large lamellar bodies in Type II cells; however,
28      the inflammatory response was greater in the guinea pig. Overall, the authors rated mice as most
29      susceptible, followed  by guinea pigs and rats.
30           In another comparative study of airway effects, Sterner-Kock et al. (2000) exposed ferrets,
31      monkeys and rats to 1.0 ppm O3 for 8 h.  The ferrets developed epithelial necrosis and

        December 2005                          AX5-47      DRAFT-DO NOT QUOTE OR  CITE

-------
 1      inflammation that was similar to the monkey, and more severe than that found in rats.  Because
 2      ferrets have a similar pulmonary structure as humans (e.g., well-developed respiratory
 3      bronchioles and submucosal glands), the authors concluded that the ferret would be a better
 4      model than rodents for O3-induced airway effects.
 5           Younger rats were found to have larger centriacinar lesions than older rats after a 1- or 7-
 6      day exposure to 0.4 ppm O3 (Dormans et al., 1996).  Thus, age susceptibility is dependent on the
 7      endpoint examined.
 8           Rats with endotoxin-induced rhinitis were more susceptible to mucous cell metaplasia in the
 9      nasal transitional epithelium caused by a 3-day exposure to 0.5 ppm O3 (Cho et al.,1999b).
10      Wagner et al. (2002) reported a similar O3-induced enhancement of inflammatory and epithelial
11      responses associated with allergic rhinitis. Brown Norway rats were exposed to 0.5 ppm O3,
12      8 h/day for 1 day or 3 consecutive days and then immediately challenged intranasally with either
13      saline or ovalbumin.
14           More recent research has focused on the concept of O3 susceptible and non-susceptible sites
15      within the respiratory tract, including in situ antioxidant status and metabolic activity. Plopper
16      et al. (1998)  examined whether the variability of acute epithelial injury to short-term O3 exposure
17      within the tracheobronchial tree is related to local tissue doses of O3 or to local concentrations of
18      reduced glutathione (GSH). Adult male rhesus monkeys were exposed for 2 h to filtered air or O3
19      (0.4 or 1.0 ppm). The O3 was generated by  18 O2 for determination of local O3 dose in the trachea,
20      proximal bronchi, distal bronchi, and proximal respiratory bronchioles. Analyses of GSH and
21      extracellular components (BAL) also were performed.  Significant cellular injury was found at all
22      sites, but the most damage, along with increased inflammatory cells,  occurred in the proximal
23      respiratory bronchiole.  A significant reduction in GSH was found in the  proximal bronchus at 0.4
24      ppm O3, and in the respiratory bronchiole at 1.0 ppm O3. A significant decrease in the percent of
25      macrophages, along with significant increases in the percent of neutrophils and eosinophils, and a
26      doubling of total lavage protein, were found after exposure to 1.0 ppm O3 only.  The authors
27      concluded that the variability of local O3 dose in the respiratory tract was related to inhaled O3
28      concentration and was closely associated with local GSH depletion and with the degree of
29      epithelial injury.
30           Plopper et al.  (e.g., Watt et al.,1998; Paige et al., 2000)  explored the site-specific
31      relationship between epithelial effects of O3 exposure and the metabolism of bioactivated

        December 2005                           AX5-48      DRAFT-DO NOT QUOTE OR CITE

-------
 1      compounds within the respiratory tract of rats. The distribution of cytochrome P450 (CYP2E1)-
 2      dependent activity, measured with a selective substrate (p-nitrocatechol), was found to be highest
 3      in the distal bronchioles and minor daughter airways, and lower in the lobar bronchi and major
 4      daughter airways.  Short-term O3 exposure (1 ppm for 8 h) increased CYP2E1 activity in the
 5      lobar bronchi/major daughter airways only; however, long-term O3 exposure (1 ppm for 90 days)
 6      decreased CYP2E1 activity in the major and minor airways, further complicating the
 7      interpretation of O3 effects based on concentration and duration of exposure and recovery. Rats
 8      treated i.p. with 1-nitronaphthalene, a pulmonary toxicant requiring metabolic activation, and
 9      exposed to 0.8 ppm O3, 8 h/day for 90 days showed greater histopathologic and morphometric
10      effects in the centriacinar region of the lung (Paige et al., 2000b).  Despite reported tolerance to
11      oxidant stress after long-term O3 exposure, there was increased severity of ciliated cell toxicity.
12
13      AX5.2.4.3 Long-Term Exposure Effects on Morphology
14          Key new exposure studies describing the morphological effects of O3 exposures lasting
15      longer than 1 week are summarized in Table AX5-6.
16          Marked mucous cell metaplasia was found in F344 rats exposed to 0.5 and 1.0 ppm O3, but
17      not 0.12 ppm for 20 months (Harkema et al., 1997a).  In a follow-up study, hyperplasia was found
18      in the nasal epithelium of rats exposed to 0.25 and 0.5 ppm,  8 h/day, 7 days/week, for  13 weeks
19      (Harkema et al.,  1999).  The mucous cell metaplasia,  and associated intraepithelial
20      mucosubstances, induced by 0.5 ppm O3 persisted for 13 weeks after exposure. An acute (8-h)
21      exposure to 0.5 ppm O3  13 weeks after the chronic exposure induced an additional increase of
22      mucosubstances in the nasal epithelium of rats, but not in rats chronically exposed to 0 or
23      0.25 ppm O3.
24          Rats continuously  exposed for 6 months to the ambient air of Sao Paulo, Brazil (11 ppb O3;
25      1.25 ppm CO; 35 |ig/m3 PM; 29 |ig/m3 SO2) also developed  secretory hyperplasia in the upper
26      airways (Lemos et al., 1994).  No significant changes in nasal tissue, however, were seen in rats
27      continuously exposed for 49 days to the ambient air of Mexico City, Mexico (Moss et al., 2001).
28          Apoptosis regulators like Bcl-2 may play a role in the development and resolution of
29      mucous cell metaplasia in the nasal airway (Tesfaigzi et al.,  1998). In rats exposed to  0.5 ppm  O3
30      for 1 month, Bcl-2 was found in protein extracts of nasal epithelium. After 3 and 6 months of
31      exposure, the number of metaplastic mucous cells in the transitional epithelium was indirectly

        December 2005                           AX5-49      DRAFT-DO NOT QUOTE OR CITE

-------
                           Table AX5-6. Effects of Ozone on Lung Structure: Long-Term Exposures
1, J
tt>
<^ '
CD
>-l
to
O
o












^
X
Y1
o



O

i-rj
H
6
o
0
H
O

o
d
Concentration
ppm

Mexico
City
Ambient:
0.018
(> 0.12
for 18 1-h
intervals)
0.12
0.5
1.0









0.12
0.50
1.0




0.12
1.0


jig/m3 Duration

23 h/day
for 7 weeks

35.3



235 6 h/day,
980 5 days/week
1,960 for
20 months








235 6 h/day,
980 5 days/week
1.960 for 24 and
30 months



235 6 h/day,
1,960 5 days/week,
for 2 or 3
months
Species

Rat;
male;
F344;
8 weeks old



Rat;
male;
F344;
6-8 weeks old








Mouse; male
and female;
B6C3F1;
6-7 weeks old



Rat;
F344/N


Effects3

No inflammatory or epithelial lesions in nasal airways or respiratory tract.






LM morphometry of CAR remodeling. Thickened tips of alveolar septa lining
ADs (alveolar entrance rings) 0.2 mm from TB in rats exposed to 0. 12 ppm and
to 0.6 mm in rats exposed to 1.0 ppm. At 0.5 and 1.0 ppm, atrophy of nasal
turbinates, mucous cell metaplasia in NTE, increased volume of interstitium and
epithelium along ADs due to epithelial metaplasia, and bronchiolar epithelial
hyperplasia. At 1.0 ppm, increased AMs and mild fibrotic response (increase in
interstitial matrix and cellular interstitium; the latter due to increase in volume in
interstitial fibroblasts). More effects in PAR than in terminal bronchioles.
Effects not influenced by gender or by aging. Effects similar to, or model of,
early fibrotic human disease (e.g., idiopathic pulmonary fibrosis).


Effects in the nose and centriacinar region of the lung at 0.5 and 1.0 ppm. Nasal
lesions were mild: hyaline degeneration, hyperplasia, squamous metaplasia,
fibrosis, suppurative inflammation of transitional and respiratory epithelium; and
atrophy of olfactory epithelium. Lung lesions: alveolar/bronchiolar epithelial
metaplasia and histiocytosis in terminal bronchioles, alveolar ducts, and
proximal alveoli. Severity was greatest in mice exposed to 1.0 ppm O3, but there
was minimal interstitial fibrosis.
Morphometric changes (epithelial thickening, bronchiolarization) occurred after
2 or 3 months exposure to 1.0 ppm O3; effects were similar to those found with
20 months exposure (see Pinkerton et al., 1995)

Reference

Moss et al.
(2001)





Catalano et al.
(1995a,b);
Chang et al.
(1995);
Harkema et al.
(1994, 1997a,b)
Pinkerton et al.
(1995);
Plopper et al.
(1994a);
Stockstill et al.
(1995)
Herbert et al.
(1996)





Pinkerton et al.
(1998)


O
HH
H
W

-------
o

O

to
O
O
Table AX5-6 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures
X
        Concentration
        ppm
                          Duration
              Species
Effects3
Reference
0.25 490 8 h/day,
0.5 980 7 days/week
for
13 weeks

0.4 784 23.5 h/day
for 1, 3, 7, 28,
or 56 days







0.5 980 8 h/day for 1,
3, and 6
months
Rat;
male;
F344/NHSD;
10-14 weeks
old
Rat;
Wistar
7 weeks old







Rat;
male;
F344/N
Mucous cell hyperplasia in nasal epithelium after exposure to 0.25 and 0.5 ppm
O3; still evident after 13 weeks recovery from 0.5 ppm O3 exposure. Mucous cell
metaplasia found only after 0.5 ppm O3, but still detectable 13 weeks PE.


Acute inflammatory response (increased PMNs and plasma protein in B ALF)
reached a maximum at day 1 and resolved within 6 days during exposure;
AMs in B ALF increased progressively up to day 56, and slowly returned to near
control levels with PE recovery. Histological examination and morphometry of
the lungs revealed CAR inflammatory responses throughout O3 exposure;
thickening of septa was observed at day 7. Ductular septa thickened
progressively at days 7, 28, and 56 of exposure; showed increased collagen at
day 28, which was further enhanced at day 56. Increased RBs with continuous
exposure. Collagen and bronchiolization remained present after a recovery
period.
Increased Bcl-2, a regulator of apotosis, after 1 month, decreasing somewhat
thereafter, returning to baseline by 13 weeks PE; increased number of
metaplastic mucous cells in NTE after 3 and 6 months.
Harkema
(1999)



Van Bree
(2002)








Tesfaigzi
(1998)

etal.




etal.









etal.


0.5 980 8 h/day for 1,
3, and 6
months
0.5 980 8 h/day for
5 days,
every 5 days
for a total of
1 1 episodes





Rat;
male;
F344/N
Monkey;
bonnet;
30-day-old
infants






Increased Bcl-2, a regulator of apotosis, after 1 month, decreasing somewhat
thereafter, returning to baseline by 13 weeks PE; increased number of
metaplastic mucous cells in NTE after 3 and 6 months.
Increased density and distribution of goblet cells in RB whole mounts stained
with AB/PAS; extensive remodeling of distal airway with O3 and O3 + HDMA
challenge; increased airways resistance and reactivity, and respiratory motor
adaptation also occurred. Authors conclude that periodic cycles of acute injury
and repair associated with the episodic nature of environmental patterns of O3
exposure alters postnatal morphogenesis and epithelial differentiation in the
distal lung of infant primates.



Tesfaigzi et al.
(1998)

Evans et al.
(2003);
Schelegle et al.
(2003a);
Chen et al.
(2003);
Fanucchi et al.
(2000);
Plopper and
Fanucchi (2000)

-------
o

O

to
O
O
Table AX5-6 (cont'd). Effects of Ozone on Lung Structure: Long-Term Exposures
>
X
(Si
 I
to
Concentration
ppm
0.8



0.5





0.5





jig/m3 Duration
1,568 8 h/day for
90 days
+ 1-NN
(100 mg/kg)
980 11 episodes of
5 days each,
8 h/day
followed by
9 days of
recovery
980 1 1 episodes of
5 days each,
8 h/day
followed by
9 days of
recovery
Species
Rat;
male;
Sprague-
Dawley
Monkey;
Macaca
mulatto;
30 days old


Monkey;
Macaca
mulatto;
30 days old


Effects3
Increased O3-induced centriacinar toxicity (histopathology, TEM, morphometry)
of 1-Nitronaphthalene (1-NN), a pulmonary cytotoxicant requiring metabolic
activation, especially to ciliated cells.

In small conducting airways O3 caused decrements in density of airway epithelial
nerves. Reduction greater with HDMA + O3. O3 or HDMA+O3 caused increase
in number of PGP 9.5 (pan-neuronal marker) in airway. CGRP-IR nerves were
in close contact with the PGP9.5 positive cells. Appearance of clusters of
PGP9.5+/CGRP cells. Suggests episodic O3 alters developmental pattern of
neural innervation of epithelial compartment.
Abnormalities in the BMZ included: (1) irregular and thin collagen throughout
the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
immunoreactivity was reduced; (4) FGF-2 immunoreactivity was absent in
perlecan-deficient BMZ, but was present in the lateral intercelluar space (LIS),
in basal cells, and in attenuated fibroblasts; (5) syndecan-4 immunoreactivity
was increased in basal cells.
Reference
Paige et al.
(2000b)


Larson et al.
(2004)




Evans et al.
(2003)




H
6
o

O
H
/O
        aTB = Terminal bronchiole.

        PE = Postexposure (i.e., time after O3 exposure ceased).

        AM = Alveolar macrophage.

        LM = Light microscopy.
                               EM = Electron microscopy.
                               RB = Respiratory bronchiole.

                               IAS = Interalveolar septum.

                               C x T = Product of concentration and time.
O
HH
H
W

-------
 1      related to the percentage of cells that were Bcl-2 positive. Cells from rats exposed to filtered air
 2      did not express any Bcl-2.
 3           A similar spectrum of lesions also was reported (Herbert et al., 1996) in the nasal cavity and
 4      centriacinar lung of male and female mice exposed to 0.5 or 1.0 ppm of O3 for 2 years, which
 5      persisted with continued exposure for 30 months.  Few changes, however, were found in other
 6      endpoints (e.g., lung function or lung biochemistry) examined in these rats.  The investigators'
 7      interpretation of the entire study is that rodents exposed to the two higher O3 concentrations had
 8      some structural hallmarks of chronic airway disease in humans.
 9           A fifth long-term study was reported in infant monkeys by Plopper et al. (Evans et al., 2003;
10      Schelegle et al., 2003a, 2003b; Chen et al., 2003; Plopper and Fanucchi, 2000; Fanucchi et al.,
11      2000) using a shorter simulated, seasonal O3-exposure pattern, but at a higher O3 concentration
12      (0.5 ppm) than the protocol used by Tyler et al. (1988, 1991a).  Infant rhesus monkeys (30 days
13      old) were exposed to filtered air, house dust mite allergen aerosol (HDMA), or O3 + HDMA.  The
14      O3 exposures were 8 h/day for 5 days, every 14 days for a total of 11 O3 episodes. Half of the
15      monkeys were sensitized to house dust mite allergen (Dermatophagoides farinae) at 14 and 28
16      days of age.  The sensitized monkeys were exposed to HDMA for 2h/day on days 3-5 of the FA
17      or O3 exposures. The lungs were removed during the last filtered air exposure and the right and
18      left cranial and right middle lobes were separately inflation fixed. Microdisection and
19      morphometric analyses were performed on the conducting airways to the level of the most
20      proximal respiratory bronchiole.  Repeated exposures to O3 or O3 + HDMA over a 6-month
21      period resulted in an atypical development of the basement membrane zone of airways in
22      nonsensitized developing monkeys. A profound remodeling in the distal conducting airways was
23      found in the sensitized monkeys as a result of the damage and repair processes occurring with
24      repeated exposure  (Evans et al., 2003; Schelegle et al., 2003a; Fanucchi et al., 2000).
25           Schelegle et al. (2003a) reported the lung histopathology results from the O3 exposures to
26      infant monkeys. At necropsy, cross sections of the left caudal lobe were prepared from each
27      animal. The accumulation of eosinophils and mucous cells within the combined epithelium and
28      interstitium compartments was determined in the conducting airways and in the
29      terminal/respiratory bronchioles.  House dust mite sensitization and HDMA challenge alone, or
30      combined with O3  exposure, resulted in significantly greater eosinophil accumulation in the
31      conducting airways when compared to FA and O3 only exposures. A significant accumulation of

        December 2005                          AX5-53      DRAFT-DO NOT QUOTE OR CITE

-------
 1      eosinophils was found in the terminal/respiratory bronchioles of the sensitized monkeys
 2      challenged with HDMA when compared to monkeys exposed to FA, O3, and HDMA + O3. The
 3      mean mass of mucous cells increased in the fifth generation conducting airways of sensitized
 4      animals challenged with HDMA alone and when combined with O3 exposure, and in the terminal
 5      bronchioles of sensitized animals exposed to HDMA + O3. The tracheal basement membrane of
 6      house dust mite-sensitized monkeys exposed to HDMA or to HDMA + O3 was significantly
 7      increased over controls; however, there were no significant changes in the airway diameter of
 8      proximal and mid-level airways. The authors interpreted these findings to indicate that the
 9      combination of cyclic O3 exposure and HDMA challenge in house dust mite-sensitized infant
10      monkeys act synergistically to produce an allergic-reactive airway phenotype characterized by
11      significant eosinophilia of midlevel conducting airways, transmigration of eosinophils into the
12      lumen, and an altered structural development of conducting airways. Exposures of sensitized
13      young monkeys to HDMA alone, or to O3 alone, resulted in eosinophilia of the mid-level
14      conducting airways and the terminal/respiratory bronchioles, but without alterations  in airway
15      structure or function.  Evans et al. (2003) examined development of the tracheal basement
16      membrane zone (BMZ) in these monkeys and found that with exposures to either O3 or HDMA +
17      O3 BMZ development was affected. Abnormalities in the  BMZ included: (1) irregular and thin
18      collagen throughout the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
19      immunoreactivity was reduced;  (4) FGF-2 immunoreactivity was absent in perlecan-deficient
20      BMZ, but was present in the lateral intercelluar space (LIS), in basal cells, and in attenuated
21      fibroblasts; (5) syndecan-4 immunoreactivity was increased in basal cells. The authors interpret
22      these data to suggest that O3 targets cells associated with synthesis of epithelial BMZ perlecan.
23      The absence of FGF-2, normally stored in the BMZ, could affect downstream signaling in airway
24      epithelium and could be responsible for the abnormal development of the airway seen in this
25      study, and thus be  an important  mechanism modulating O3-induced injury.
26           Mid-level bronchi and bronchioles  from these monkeys were examined for alterations in
27      airway innervation (Larson et al., 2004).  They found decrements in the density of epithelial
28      nerves in the axial  path between the sixth and seventh airway generations in exposures to O3.
29      Combined O3+HDMA exposures exacerbated this reduction. They attribute this loss of nerve
30      plexuses to neural  regression or stunted nerve development, the latter corroborated by the Evans
31      et al. (2003) finding of decreased growth factors following O3 exposure. Additionally, they found

        December 2005                           AX5-54      DRAFT-DO NOT QUOTE OR CITE

-------
 1      streaks or clusters of cells immunoreactive for protein gene product 9.5 (PGP 9.5, a pan-neuronal
 2      marker) and negative for calcitonin gene-related peptide. The functional significance of this is
 3      unknown, and presumed to be implicated in injury-repair process induced by O3.
 4           Recently, bronchiolization was reported in rats exposed to 0.4 ppm O3 for only 56 days (van
 5      Bree et al., 2001).  Collagen formation progressively increased with increasing O3 exposure, and
 6      remained increased into PE recovery. In addition to centriacinar remodeling, Pinkerton et al.
 7      (1998) reported thickening of tracheal, bronchial, and bronchiolar epithelium after 3 or 20 months
 8      exposure to 1 ppm. No such responses were observed at either time point for exposures to 0.12
 9      ppm O3.
10
11      AX5.2.5   Effects on Pulmonary  Function
12      AX5.2.5.1  Introduction
13           Numerous pulmonary function studies of the effects of short-term O3 exposure (defined here
14      as < 1 week of exposure) in several animal species have been conducted and generally show
15      responses similar to those of humans (e.g., increased breathing frequency, decreased tidal volume,
16      increased resistance, decreased forced vital capacity [FVC] and changes in the expiratory flow-
17      volume curve). The breathing pattern returns to normal after O3 exposure.
18           This section will provide a brief overview of acute and short-term exposure effects and then
19      focus on functional changes observed after long-term exposure to  O3 (defined here as >1 week of
20      exposure) and on the new studies on O3-induced airway hyperresponsiveness (AHR).
21
22      AX5.2.5.2  Acute and Short-Term Exposure Effects on Pulmonary Function
23           Wiester et al. (1996) exposed male Fischer 344 rats to 0.5 ppm O3 for either 6 or 23 h/day
24      over 5 days. Ozone-induced changes in lung volume were attenuated during the 5 exposure days
25      and returned to control levels after 7 days recovery. The responses to repeated O3 exposure in
26      rats were exacerbated by reduced ambient temperature, presumably as a result of increased
27      metabolic activity.
28           Recent work has utilized inbred mouse strains with varying ventilatory responses to O3 to
29      attempt to model susceptible populations. As differences were seen in inflammatory responses to
30      acute O3 exposures in C57BL/6J and C3H/HeJ mice, comparisons were made of their ventilatory
31      responses also (Tankersley et al., 1993). Following an exposure of 2 ppm O3 for 3 h, breathing

        December 2005                          AX5-55      DRAFT-DO NOT QUOTE OR CITE

-------
 1      frequency (f), tidal volume (VT), and minute ventilation were measured 1 and 24 h in both
 2      normocapnia (or air at -0% CO2) and hypercapnia (5 or 8% CO2).  They demonstrated that acute
 3      O3 exposures caused altered hypercapnic ventilatory control, which varied between strains. The
 4      observations from this study indicate that control of ventilation is at least in part regulated by
 5      other genetic factors.
 6           The Paquette et al. (1994) study discussed in 5.3.3.3 also measured ventilatory responses in
 7      C57BL/6J and C3H/HeJ mice on repeated subacute exposures to O3.  C57BL/6J and C3H/HeJ
 8      had differing responses to both normocapnia and hypercapnia. Normocapnic VE was greater
 9      following subacute O3 exposure in C57BL/6J mice than in C3H/HeJ mice, due to increased and
10      reduced VT, respectively. The authors speculated that this increased VT in C57BL/6J mice may
11      contribute to the increased susceptibility to lung injury due to a greater dose of O3 reaching the
12      lower lung.  Hypercapnic ventilatory responses following subacute O3 exposures demonstrated
13      reduced VE (due to decreased VT) in C57BL/6J only.  Evaluations of O3 dosimetry were
14      performed in these two strains using 18O-labeled O3 (Slade et al., 1997).  Immediately after
15      exposures to 2 ppm 18O3 for 2-3 h, C3H/HeJ mice had 46% less 18O in lungs and 61% less in
16      trachea, than C57BL/6J. Additionally, C3H/HeJ mice had a greater body temperature decrease
17      following O3 exposure than C57BL/6J mice.  The authors suggested that the differences in
18      susceptibility to O3 are due to differences the ability to decrease body temperature and,
19      consequently decrease the dose of O3 to the lung.
20           Takahashi et al. (1995b) measured tracheal transepithelial potential (VT) in eight mouse
21      strains 6 h after exposure to 2 ppm O3 for 3 h.  AKR/J, C3H/HeJ, and CBA/J were identified as
22      resistant strains and 129/J, NJ, C57BL/6J, C3HeB/FeJ and SJL/J were identified as susceptible
23      strains. The pattern of inheritance for this trait suggested an autosomal recessive pattern.  The
24      authors noted that strains' responses to this parameter did not show concordance with
25      inflammatory responses, suggesting to the authors that the two phenotypes are not controlled by
26      the same genetic factors.
27           The Savov et al. (2004) study discussed in 5.2.3.3 (2.0 ppm O3 for  3 h) characterized
28      ventilatory responses using whole body plethysmography and enhanced pause index (Penh) in the
29      nine mouse strains evaluated.  Table AX5-4 lists the baseline Penh, the Penh following O3, and the
30      Penh response to methacholine (MCh) following O3. C57BL/6J was hyporeactive to MCh prior to
        December 2005                           AX5-56      DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3, but was very responsive to MCh following O3.  Conversely, C3H/HeJ had an intermediate
 2      baseline Penh and a small response to MCh following O3 exposure.
 3
 4      AX5.2.5.3 Long-Term Exposure Effects on Pulmonary Function
 5           New long-term O3 exposure studies evaluating pulmonary function are not available.
 6
 7      AX5.3.5.4 Acute and Chronic Exposure Effects on Airway Responsiveness
 8           New studies in laboratory animals allow possible ways of predicting, with increased
 9      specificity, the effects of O3 exposure on the exacerbation of asthma symptoms and the risk of
10      developing asthma in humans. A variety of methods have been used to assess airway
11      responsiveness in humans, including airway challenge with nonspecific bronchoconstrictors (e.g.,
12      inhaled methacholine or histamine) and with indirect (e.g., inhalation of adenosine
13      monophosphate, hypertonic saline, mannitol) stimuli to bronchoconstriction (Anderson, 1996).
14      Although inhaled agonist challenges are preferred in humans, laboratory animals studies have
15      employed intravenous (i.v.) agonist challenges as well as inhalation challenges.  The
16      comparability of these two routes for bronchoconstrictor administration has not been well studied,
17      however differences have been reported (e.g., Sommer et al., 2001). Most challenge tests require
18      an outcome measure that reflects airway function, such  as pulmonary resistance, dynamic lung
19      compliance, or deceased forced expiratory flow and volume. In a generalized sense, resistance is
20      a measure of large airway function, and dynamic compliance is a measure of small airway
21      function.  In regards to pulmonary mechanics, human infants are much like laboratory rodents
22      because both have very compliant chest walls. Therefore, this discussion will make comparisons
23      from the published literature on airway responsiveness in human infants as recently reviewed, for
24      example, by Stick, 2002.
25           As with infants, a limitation of the traditional types of studies in laboratory rodents is the
26      requirement for sedation. The need for artificial ventilation in laboratory animal studies may
27      cause breathing patterns that affect O3 deposition (see Section 5.2). load et al. (2000) reported
28      that when 1 ppm O3 for 90 min is administered to isolated rat lung at either 2.4 ml/40 bpm or
29      1.2 m/80 bpm, the more rapid breathing pattern elicits less epithelial cell injury than the slower
30      breathing pattern.  They further showed greater reduction in injury in the proximal axial airway
31      compared to its adjacent airway branch and terminal bronchiole.  For rats, normal respiration is

        December 2005                           AX5-57      DRAFT-DO NOT QUOTE OR CITE

-------
 1      approximately 100 bpm, so this paradigm does not really model rapid shallow breathing elicited
 2      in the intact animal.  Schelegle et al. (2001) showed that the large conducting airways of rats are
 3      protected by rapid, shallow breathing, but there is a more even distribution of epithelial cell injury
 4      to the terminal bronchioles.  Recent observations (Postlethwait et al., 2000) demonstrate that the
 5      conducting airways are the primary site of acute cytotoxicity from O3 exposure.  By utilizing a
 6      new analytic approach of three-dimensional mapping of the airway tree in SD  rat isolated lung
 7      exposed to 0, 0.25, 0.5, or 1.0 ppm O3 for 20 to 90 minutes, they showed a concentration-
 8      dependent increase in injured cells. Injury was evident in proximal and distal  conduction
 9      airways, lowest in terminal bronchioles, and highest in the small side branches downstream of
10      bifurcations. These exposure levels did not concurrently elicit changes in LDH activity or total
11      protein in BALF.
12           More recent methods of studying laboratory animals utilize unanesthetized, unrestrained
13      rodents in a whole-body plethysmograph (e.g., Shore et al., 2001, 2002; Goldsmith et al., 2002;
14      Jang et al., 2002), but pulmonary resistance is measured indirectly using several indices of
15      inspiratory/expiratory pressure differences, including enhanced pause (Penh), that may be less
16      sensitive than direct measurements of lung airflow resistance (Murphy, 2002). Also, in another
17      study by Sommer et al. (1998), unrestrained guinea pigs were shown to have a daily variability in
18      pulmonary resistance that is similar to that occurring in humans. Therefore, circadian rhythms of
19      airway caliber must be considered when performing airway challenge tests in any species.
20      Animals with  acute viral illness have morphological  evidence of inflammatory cell infiltration,
21      bronchiolar wall edema, epithelial hyperplasia, and increased airway mucous plugs that can cause
22      airway narrowing, air trapping, and serious functional changes in the lung (Folkerts et al., 1998).
23           Exercise-induced bronchoconstriction in humans appears to be mediated by changes in the
24      tonicity of the airway lining fluid (Anderson and Daviskas, 2000) and, therefore, a test in
25      laboratory animals based on the inhalation of mannitol aerosol (hyperosmolar) might be feasible
26      and provide information similar to that from exercise challenges in cooperative children and
27      adults (Brannan et al., 1998). In active humans with asthma, adenosine monophosphate
28      challenges appear to better reflect ongoing airway inflammation than histamine or methacholine
29      challenges (Polosa and Holgate,  1997; Avital et al, 1995a,b), and might be useful in identifying
30      mechanisms of asthma in laboratory animals and their responsiveness to environmental
31      pollutants.

        December 2005                           AX5-58      DRAFT-DO NOT QUOTE OR CITE

-------
 1     Airway responsiveness in asthma
 2           The increased responsiveness to bronchoconstrictor challenge in asthma is thought to result
 3     from a combination of structural and physiological factors that include increased inner-wall
 4     thickness, increased smooth-muscle responsiveness, and mucus secretion.  This baseline
 5     responsiveness is thought to be modulated in asthma by chronic inflammation and airway
 6     remodeling (Stick, 2002). Longitudinal studies in adults have shown that the development of
 7     airway responsiveness is associated with persistence of symptoms (O'Conner et al., 1995).
 8
 9     Airway responsiveness in infants
10           The age at which nonspecific airway hyperresponsiveness first appears in humans is
11     unknown, although both genetic and environmental factors are most likely to play a role.
12     Although underlying physiological or structural factors may determine this relative increase in
13     responsiveness in infants compared with older children, the most likely explanation is that infants
14     receive a relatively larger dose of inhaled challenge agent than older children.  Thus, when a
15     correction is made for this dose effect, infants and older children appear to have a similar
16     response to inhaled histamine (Stick et al., 1990; Stick, 2002). Airway responsiveness at one
17     month is a predictor of lung function at six years (Palmer et al., 2001). Data from this study also
18     show that the genetic determinants of atopy and airway responsiveness are independent (Palmer
19     et al., 2000).  In another study of infants with wheeze, persistence of airway hyperresponsiveness
20     was associated with persistence of symptoms, although airway responsiveness  at one month of
21     age was neither a sensitive nor a specific predictor of outcome (Delacourt et al., 2001).
22           The human studies imply that airway responsiveness is a key factor in asthma, but it is not
23     clear if the factors that are important for airway responsiveness in early life are related to
24     inflammation, structure or physiology of the airways, or the combination of all three.
25     Furthermore, it is not clear how viruses, allergens and irritants in the environment modify  innate
26     airway responses (Holt et al., 1999), but they are known to be important.
27
28     Airway responsiveness in laboratory animals
29           Laboratory animals, including rodents (mice, rats, guinea pigs), rabbits, cats, dogs, and
30     nonhuman primates have been used to study the effects of O3 exposure on airway
31     bronchoconstriction. New studies examining airway responsiveness in laboratory animals are

       December 2005                          AX5-59      DRAFT-DO NOT  QUOTE  OR CITE

-------
 1      listed in Table AX5-7.  Ozone-induced AHR in guinea pigs has been used as a model of
 2      bronchospasm (e.g., Kudo et al., 1996; van Hoof et al., 1996; 1997a,b; Matsubara et al., 1997a,b;
 3      Sun and Chung, 1997; Aizawa et al., 1999a,b; Tsai et al., 1998; Nakano et al., 2000). In this
 4      model, the guinea pigs are acutely exposed for 1 or 2 h to high O3 concentrations (2 to 3 ppm).
 5      The model is useful for understanding mechanisms of bronchospasm, but are not directly relevant
 6      for extrapolation to potential airway responses in humans exposed to ambient levels of O3.  Dye
 7      et al. (1999) showed hyperresponsiveness to methacholine in rats 2 h after exposure to 2 ppm
 8      O3 for 2 h.
 9           Shore et al. (2000) have shown that O3-induced AHR is reduced in immature SD rats. The
10      animals were exposed to 2 ppm O3 for 3 h in nose-only-exposure plethysmographs and baseline
11      VE was normalized for body weight.  VE was reduced, primarily as a result of decreased VT, with
12      O3 exposure.  Adults rats had 40-50% decreases in VE, 6-wk-old rats had smaller decreases, and
13      2- and 4 week-old rats had no significant changes in VE.. This suggested to the authors that the
14      higher baseline VE in the young rats, combined with the smaller decreases in VE with O3
15      exposure, created a much larger dose in the immature rats. Shore et al. (2002) completed
16      complementary studies in A/J mice at ages 2, 4,  8, or 12 weeks using exposures of 0.3 to 3 ppm
17      for 3 h. Ozone caused a similar concentration-related decreases in VE except in the 2- and 4-week
18      old mice.  This suggested that the young mice are less sensitive than adult mice to O3 in terms of
19      AHR.  Lean and obese mice were also compared for differences in AHR response to O3 exposure
20      (2.0 ppm O3 for 3 h). Shore et al. (2003) exposed lean, WT C57BL/6J mice and mice with a
21      genetic defect in the gene that codes for leptin, the satiety hormone. These ob/ob mice had
22      enhanced AHR and inflammation compared to the WT mice.
23          Airway hyperresponsiveness can be induced by specific antigens as well as O3. The most
24      commonly used laboratory animal model is the ovalbumin (OVA) sensitized guinea pig.
25      Animals sensitized with OVA have been shown to have similar responses to nonspecific
26      bronchoconstrictors (e.g., carbachol) as control animals; however, OVA-sensitized guinea pigs
27      exposed to O3 showed increased AHR to histamine (Vargas et al., 1994). Guinea pigs were
28      sensitized by inhalation exposure to ovalbumin and subsequently challenged with histamine; the
29      main endpoint was specific airway resistance. When exposed to O3 before sensitization, repeated
30      exposures to very high levels (5.0 ppm) decreased the OVA sensitization threshold; however, in
31      already sensitized animals, a 2-h exposure to > 1.0 ppm enhanced airway responsiveness to

        December 2005                         AX5-60      DRAFT-DO NOT  QUOTE OR CITE

-------
                                                   Table AX5-7.  Effects of Ozone on Airway Responsiveness
1, J
i
cr
^
to
o
o















^
X
ON
^




O
^
"Tl
H
6
o
2|
0
H
O
Cj
O
H
Ozone
Concentration"

ppm ug/m

0.1 196
0.3 588






0.15 294
0.30 588
0.60 1,176
1.2 2,352
0.3 588




0.5 980



1 1,960


1 1,960



2 3,920



3 5,880




Exposure
Duration

4 h/day,
4 days/week
for 24 weeks





4h



4 h/day for 1,
3, 6, 12, 24,
or 38 days


8 h/day for 5 days,
repeated every
14 days for
6 months
Ih


Ih



2h



Ih



Challenge"

Agent Route

ACh inh
OVA inh






ACh iv
Hist iv
SP iv

SP iv




Hist inh



Ach inh
OVA inh

Mch inh
OVA inh


Mch inh



Hist iv
OVA iv



Species, Sex, Strain,
Drugs and Agec

none Guinea pig, M & F,
Hartley






Na Guinea pig, M
Pentobarbital Hartley,
500-600g

Na Guinea pig, M
Pentobarbital Hartley,
500-600 g


Ketamine + Rhesus monkey,
Diprivan M, 30 days old


Urethane Guinea pig, M,
Dunkin-Hartley,
250-300 g
Urethane Mouse, M,
C57BL/6,
6 weeks old

Ketamine + Rat, M,
Zylazine F344
14 months old

Guinea pig, M,
Hartley,
500-700 g



Observed Effect(s)

O3 exposure did not produce airway hyperresponsiveness to
ACh in nonsensitized animals; in OVA-sensitized animals,
there was increased responsiveness to both nonspecific (ACh)
and specific (OVA) airway challenge that persisted for 4 weeks
after exposure 0. 1 and 0.3 ppm O3. Effects were not gender
specific and were not associated with BALF inflammatory
indicators, but were associated with antigen-specific antibodies
in blood.
Increased airway responsiveness to Hist, but not Ach, 16-18 h
after 1.2 ppm O3 exposure only. Increased responsiveness to
SP occurred after exposure to > 0.3 ppm O3

Increased airway responsiveness to SP occurred 16-18 h after
exposure to 0.3 ppm O3 for 1, 3, 6, 12, and 24 days; but not
after 48 days. Highly significant correlation between airway
responsiveness and BALF total cells, Ams, neutrophils, and
eosinophils, suggesting that airway inflammation is involved.
Increased airway responsiveness to Hist after 10 episodes of
exposure to O3 + HDMA in sensitized infant monkeys.


Increased bronchial responsiveness at 3 h, but not 24 h after O3;
OVA had no effect on baseline, but enhanced airway
responsiveness 24 h after O3
Ozone caused increased Cdyn and VE, and decreased PaO2 in
OVA- sensitized mice


Increased airway responsiveness to MCh 2 h PE



Increased airway responsiveness to histamine after O3 exposure
in OVA sensitized guinea pigs, with enhanced responsiveness
after OVA challenge



Reference

Schlesinger
et al. (2002a,b)






Segura et al.
(1997)


Vargas et al.
(1998



Schelegle et al.
(2003b)


Sun et al.
(1997)

Yamauchi et al.
(2002)


Dye et al.
(1999)


Vargas et al.
(1994)

O
HH
H
W
"Table ordered according to ozone concentration.
bMCh = methylcholine, ACh = acetylcholine, Hist = histamine, 5-HT = 5-hydroxytryptamine, SP = substance P, FS = field stimulation, CCh = carbachol, TX = thromboxane,
KC1 = potassium chloride, Pt = platinum; Route:  iv = intravenous, inh = inhalation., sc = subcutaneous, ip = intraperitoneal
°Age or body weight at start of exposure.

-------
 1      ovalbumin.  Thus, O3 exposure does not modify the development of antigen-induced AHR and, in
 2      fact, may enhance AHR at high levels of exposure.
 3           The enhancement of antigen-induced bronchoconstriction by acute, high-level O3 was
 4      further explored in OVA-sensitized guinea pigs (Sun et al., 1997) and mice (Yamauchi et al.,
 5      2002). Male Dunkin-Hartley guinea pigs were sensitized by i.p. injection of OVA (1  mL 20 jig)
 6      and exposed to filtered air or 1  ppm O3 for 1 h (Sun et al.,  1997). Airway responsiveness to
 7      inhaled acetylcholine was measured 3 h and 24 h after air and O3 exposures and BAL was
 8      performed.  Four other groups of OVA-sensitized animals were exposed to OVA aerosol or to O3
 9      alone, or in combination,  and airway responsiveness to acetylcholine was measured 3 h and
10      24 h PE.  In the combined exposure groups, OVA aerosol exposure was initiated either
11      immediately after, or 21 h after O3 exposure. In this study, O3 exposure increased bronchial
12      responsiveness to acetylcholine at 3 h, but not 24 h, while  OVA alone had no effect. Combined
13      exposure to  O3 and OVA  (1 ppm for 1 h, then 3 min OVA) increased bronchial responsiveness to
14      acetylcholine 3 h after O3 exposure. At 24 h following O3 exposure, AHR increased when OVA
15      challenge was performed  at 21  h, suggesting to authors that O3 pre-exposure can potentiate OVA-
16      induced AHR. Neutrophil counts in the BALF increased at 3 and 24 h after O3 exposure alone,
17      but were not further increased when O3 exposure was combined with OVA airway challenge;
18      however protein content of the  BALF did increase at 3  and 24 h in the O3 and OVA groups.
19      Thus, this study also indicates that high-ambient O3 exposure can augment antigen(OVA)-
20      induced AHR in guinea pigs.
21           Male C57BL/6 mice were sensitized by i.p. injection of OVA (50 jig) and exposed to
22      filtered air or 1 ppm O3 for 1 h  (Yamauchi et al., 2002). Airway responsiveness to methacholine
23      was measured 24 h after an inhalation challenge to OVA (10 mg/mL) in OVA-sensitized and
24      control groups, and 3 h after the OVA inhalation  challenge in animals exposed to O3 to maximize
25      the susceptible time period for AHR (Yamaguchi et al., 1994).  Pulmonary function was
26      measured by plethysmography  before and at 10-min intervals during the 1-h air and O3 exposures.
27      Blood gases and BAL fluid were monitored immediately after exposure. Mice sensitized to OVA
28      had AHR to methacholine.  Ozone exposure caused significant decreases in dynamic lung
29      compliance, minute ventilation, and PaO2 in OVA-sensitized mice, but not in controls. A marker
30      of inflammation (soluble intercellular adhesion molecule-1 [sICAM-1]) was elevated in the BAL
31      fluid of OVA-sensitized mice, but sICAM-1 levels were not significantly changed by O3

        December 2005                          AX5-62       DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposure, indicating that the O3-induced AHR to methacholine was not caused by O3-induced
 2      inflammation.
 3           Ozone-induced AHR may be temporally associated with neutrophils (DeLorme et al., 2002)
 4      and other inflammatory cells stimulated by leukotrienes (Stevens et al., 1995a), cytokines (Koto
 5      et al., 1997), mast cells (Igarashi et al., 1998; Noviski et al., 1999), or by oxygen radicals
 6      (Takahashi et al., 1993; Stevens et al., 1995b; Tsukagoshi et al., 1995; Kudo et al., 1996). Two
 7      new studies have shown that inflammation is not a prerequisite of AHR (Stevens et al., 1994;
 8      Koto et al., 1997), and some investigators have suggested that O3-induced AHR may be
 9      epithelium dependent (Takata et al., 1995; Matsubara et al., 1995; McGraw et al., 2000). For
10      example, neonatal rats pretreated with capsaicin, which will permanently destroy C-fibers and
11      prevent O3-induced release of neuropeptides (Vesely et al., 1999b), and then exposed to O3 when
12      adults, showed a marked increase in airway responsiveness to inhaled aerosolized methacholine
13      (Jimba et al., 1995). Some investigators (Matsumoto et al., 1999; DeLorme et al., 2002) have
14      shown that respective intravenous pretreatment with neutrophil elastase inhibitor or PMN
15      antiserum can block O3-induced AHR; other investigators (Koto et al., 1995; Aizawa et al., 1997;
16      Takebayashi et al.,  1998) have shown that depletion of tachykinins by capsaicin treatment, or by a
17      specific tachykinin  receptor  antagonist, can block the induction of AHR by O3.  The seemingly
18      disparate responses in laboratory animals  may be due to species- or strain-specific differences in
19      inherent reactivity to bronchoconstrictors, or to inherent differences in susceptibility to O3-
20      induced inflammation (Zhang et al., 1995; Depuydt et al., 1999; Dye et al., 1999).
21           The studies referenced above are useful for gaining an understanding of how acute
22      O3 exposure modulates specific airway responsiveness to allergen challenge, but the
23      O3 concentrations used in these studies are not typical of ambient exposures in the population.
24      More recently published studies that may  be potentially relevant to ambient levels of O3 were
25      conducted in vivo, in an isolated perfused lung model, and in ex vivo lung segments using
26      multihour and repeated multihour exposures with ambient levels of O3. A study on the
27      relationship between O3-induced AHR and tracheal epithelial function was conducted in New
28      Zealand white rabbits by Freed et al. (1996).  Rabbits were exposed to filtered air or to 0.2 ppm
29      O3 for 7 h. Tracheal transepithelial potential  difference (PD) was measured 3 h after exposure
30      and lung resistance and reactivity were partitioned into central and peripheral components using
31      forced oscillation. Exposure to O3 significantly decreased PD, but did not change lung resistance.

        December 2005                          AX5-63      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Changes in the compartmentalized lung resistance, measured in response to bronchoconstrictor
 2      aerosol challenge (acetylcholine) before and after bilateral vagotomy, were not significantly
 3      different in air-exposed rabbits; however, bilateral vagotomy did enhance peripheral lung
 4      reactivity in O3-exposed rabbits. The acetylcholine-induced increase in lung resistance with O3
 5      exposure (140%) was two times higher than with air exposure, indicating that ambient-level O3
 6      exposure affects tracheal epithelial function in rabbits and increases central airway reactivity,
 7      possibly through vagally-mediated mechanisms.
 8           Delaunois et al. (1998) studied pulmonary mechanics and hemodynamics in the isolated
 9      perfused lung model that allowed partitioning of the total pressure gradient into arterial, pre- and
10      post-capillary, and venous components. New Zealand white rabbits were exposed to filtered air
11      or to 0.4 ppm O3 for 4 h and evaluated for airway responsiveness to acetylcholine, substance P, or
12      histamine immediately or 48 h later.  Ozone exposure did not significantly change baseline values
13      of pulmonary resistance and dynamic compliance, but inhibited pulmonary mechanical reactivity
14      to all three bronchoconstrictors that persisted for 48 h. Ozone also modified vasoreactivity of the
15      vascular bed, but only at 48 h PE.  Arterial segmental pressure, normally insensitive to
16      acetylcholine and substance P, was significantly elevated by O3; precapillary segmental pressure
17      decreased in response to acetylcholine. The authors concluded that O3  can induce direct vascular
18      constriction, but the vascular responses are variable and depend on the  agonist used and on the
19      species studied.
20           Guinea pigs were  exposed to filtered air, 0.15,  0.3,  0.6, or 1.2 ppm O3 for 4 h and evaluated
21      for airway responsiveness to acetylcholine,  substance P,  or histamine 16 to 18 h later (Segura
22      et al., 1997). Ozone did not cause airway hyperresponsiveness to acetylcholine or histamine,
23      except at the highest concentration (1.2 ppm O3) for histamine.  However, O3 did cause
24      hyperresponsiveness to  substance P at > 0.3 ppm.  The authors speculated that O3 destroys neutral
25      endopeptidases, responsible for substance P inactivation, that are located in airway epithelial
26      cells. In a follow-up study at the same laboratory, Vargas et al. (1998)  reported that guinea pigs
27      chronically exposed to 0.3 ppm O3 for 4 h/day became adapted to substance P-induced AHR.
28      Ozone caused increased sensitivity to substance P after 1, 3, 6, 12, and  24 days of exposure that
29      was associated with airway inflammation; however,  after 48 days of exposure, the increased
30      sensitivity to substance  P was lost.
        December 2005                           AX5-64      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The effects of repeated short-term exposure or long-term exposure to O3 on airway
 2      responsiveness have been investigated in several laboratory animal studies. Both Vargas et al.
 3      (1998), discussed above, and Szarek et al. (1995) reported that the AHR associated with acute O3
 4      exposures does not persist during long-term exposure to ambient-levels of O3 (< 1 ppm). In the
 5      Szarek et al. (1995) study, Fischer 344 rats were exposed to 0.0, 0.12, 0.5, or 1.0 ppm O3, 6h/day,
 6      5 days/week for 20 months. Eighth generation airway segments were isolated from the exposed
 7      rats and circumferential tension development was measured in response to bethanechol,
 8      acetylcholine, and electrical field stimulation and normalized to smooth muscle area. Maximum
 9      responses of the small bronchi of male rats were significantly reduced after exposure to 0.12 and
10      0.5 ppm O3, suggesting some adaption had taken place during long-term exposure, possibly
11      increased inner wall thickness.
12           Changes in breathing pattern and lung function caused by O3 are attenuated with repeated
13      daily exposures for at least 3 to 5 days. load et al. (1998), however, reported that repeated daily
14      O3 exposure enhances,  rather than diminishes, the responsiveness of rapidly adapting airway
15      receptors. Guinea pigs were exposed to 0.5 ppm O3,  8 h/day for 7 days and then studied for
16      measurement of impulse activity of the rapidly adapting receptors, dynamic lung compliance, and
17      lung resistance at baseline and in response to substance P, methacholine, hyperinflation, and
18      removal of end-expired airway pressure. Repeated exposure increased receptor activity to
19      substance P, methacholine, and hyperinflation; there  were no significant effects on baseline or
20      substance P- and methacholine-induced changes in lung compliance and resistance.  Because
21      agonist-induced changes in receptor activity precede lung function changes, the authors
22      concluded that the responsiveness of rapidly adapting receptors was enhanced.
23           Schlesinger et al. (2002a,b) evaluated airway responsiveness following acetylcholine or
24      OVA inhalation challenges in male and female Hartley guinea pigs exposed to 0.1 and 0.3 ppm
25      O3, 4 h/day, 4 days/week for 24 weeks. The bronchoprovocation tests were performed at 4 week
26      intervals during exposure and at 4- to 8-week intervals during the PE period in nonsensitized
27      animals and in animals sensitized to allergen (OVA)  prior to, or concurrent with, O3 exposure.
28      Ozone exposure did not cause AHR in nonsensitized animals, but did exacerbate AHR to both
29      acetylcholine and OVA in sensitized animals that persisted for 4 weeks after exposure.  The
30      effects of O3 on airway responsiveness were gender independent and they  were concentration-
31      related for the acetylcholine challenges. The study did not show any evidence of adaptation.

        December 2005                          AX5-65     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Schelegle et al. (2003a) evaluated airway responsiveness in infant rhesus monkeys exposed
 2      to a 5 day O3 episode repeated every 14 days over a 6-month period.  Half of the monkeys were
 3      sensitized to house dust mite allergen (Dermatophagoides farinae) at 14 and 28 days of age
 4      before exposure to a total of 11 episodes of O3 (0.5 ppm, 8 h/day for 5 days followed by 9 days of
 5      FA), house dust mite allergen aerosol (HDMA), or O3 + HDMA. Monkeys were sedated for
 6      measurement of airway responsiveness and then anesthetized for measurement of pulmonary
 7      mechanics (e.g., Raw, R,.s) using a head-out body plethysmograph.  A necropsy was performed
 8      after all pulmonary function measurement were taken (see the previous Section 5.3.4 for results).
 9      The HDMA and histamine aerosol challenges were administered until Raw doubled. Data were
10      expressed as the concentration increasing Raw by 150% (EC 150 Raw). Other measurements
11      included VT, fB, and SaO2 (estimated by pulse oximeter).  Baseline Raw was significantly elevated
12      after 10 exposure episodes in the HDMA + O3 group compared to the FA, HDMA, and O3
13      exposure groups.  Aerosol challenge with HDMA at the end of the 10th episode did not
14      significantly affect Raw, VT, fB, or SaO2. Aerosol challenge with histamine was not significantly
15      different after 6 episodes; however, the EC 150 Raw for the HDMA + O3 group was significantly
16      reduced after 10 episodes when compared to the FA, HDMA, and O3 exposure groups, indicating
17      the development of airway hyperresponsiveness in this group sometime between episodes 6
18      and 10.
19          Using 18O exposures at 1 ppm for 2 h and breathing frequencies of 80, 120, 160, or
20      200 breaths/minute, Alfaro et al., (2004) examined the site-specific deposition of 18O. At all
21      frequencies, parenchymal areas had a lower content of 18O than trachea and bronchi.
22      As breathing frequency increased from 80 to 160 bpm, the deposition showed a reduction in
23      midlevel trachea and an increase in both mainstream bronchi. At this frequency there was also an
24      increase in deposition in parenchyma  supplied by short (cranial) airway paths. At 200 bpm 18O
25      deposition in trachea increased, concurrent with increases in right cranial and caudal bronchi
26      regions.  Right cranial parenchymal content decreased at 200 bpm, whereas right caudal
27      parenchymal levels did not change at  any breathing frequency.  The authors list some limitations
28      of this study, such as the possible effect on regional distribution of ventilation by the use of the
29      negative-pressure ventilator, the effect of paralysis on airway geometry, and possible
30      translocation of 18O during the 2 h exposure period. But the evidence provided by these studies
31      strongly suggests that the effect of rapid, shallow breathing is to create a more evenly distributed

        December 2005                           AX5-66     DRAFT-DO NOT QUOTE OR CITE

-------
 1      injury pattern, with possibly greater protection from focal injury to the large conducting airways
 2      including the trachea and the left mainstem bronchus.
 3           Schelegle et al. (2003b) examined adaptive phenomena in SD rats using an exposure
 4      paradigm consisting of 5 days of daily 8 h 1 ppm O3 exposures followed by 9 days of recovery in
 5      filtered air.  This O3/FA pattern was repeated for 4 cycles and animals were analyzed on day 1
 6      and day 5 of each exposure and at the end of the filtered air period.  The O3 -induced rapid
 7      shallow breathing pattern followed by adaptation occurred with each cycle, however, the release
 8      of substance P from the trachea , the neutrophil content, and cell proliferation, as visualized by
 9      BrdU labeling, became attenuated after the first cycle, thus displaying a disconnect from the rapid
10      shallow breathing response.  The repeated cycles of O3 also created hypercellularity of the CAR
11      epithelium and thickening of the CAR interstitium, not linked to changes in cell proliferation.
12      The authors hypothesize a mechanism of injury from repeated O3 exposures that consists of: (1)
13      diminished neutrophilic inflammation/and or release of mitogenic neuropeptides, (2) depressed
14      cell proliferative response, and (3) cumulative distal airway lesion.
15           Using a subset of monkeys from the same study (Schelegle et al.,  2003b) reported above,
16      Chen et al. (2003) reported that attenuation of O3-induced rapid shallow breathing and lung
17      function changes typically seen with repeated O3 exposure may be caused by the adaptation of the
18      respiratory motor responses. The monkeys were killed 3 to 5 days after exposure to
19      11 "episodes" of O3 and brain stem coronal slices were prepared.  Whole cell recordings were
20      performed on neurons from the nucleus tractus solitarius (NTS), the brain stem region that
21      processes lung sensory signals.  Episodic O3 exposure resulted in neuroplasticity of the NTS
22      including increased nonspecific excitability of the NTS neurons, an increased input resistance,
23      and an increased spiking response to intracellular injections of depolarizing current.
24
25      AX5.2.6  Genotoxicity Potential of Ozone
26           Many experimental studies have been conducted to explore the mutagenic and carcinogenic
27      potential of O3.  Recently published in vivo exposure studies found increased DNA strand breaks
28      in respiratory cells from guinea pigs (Ferng et al., 1997) and mice (Bornholdt et al., 2002) but
29      only after high O3 exposures (1  ppm for 72 h and 1  or 2 ppm for 90 min, respectively).
30           Witschi et al.  (1999) exposed female strain A/J mice to 0.12, 0.50, and 1.0 ppm O3 for 6
31      h/day, 5 days/week for up to 9 months. After 5 months, one-third of the O3-exposed mice were

        December 2005                           AX5-67     DRAFT-DO NOT QUOTE OR CITE

-------
 1      compared to one-half of the controls exposed to filtered air.  There was no statistically significant
 2      difference in lung tumor multiplicity or incidence. The remaining O3-exposed mice were split
 3      into two groups. Ozone exposure continued in one group for an additional 4 months and the mice
 4      in the second group were allowed to recover in filtered air.  Again, there were no statistically
 5      significant differences in  lung tumor multiplicity between control mice and mice exposed to any
 6      concentration of O3 for 9  months. The highest, and only statistically significant lung tumor
 7      incidence, was found in the mice exposed to 0.5 ppm O3. In the O3-exposed mice allowed to
 8      recover in filtered air, only the mice exposed to 0.12 ppm O3 had statistically significant increases
 9      in lung tumor incidence and multiplicity. These results were considered by the authors to be
10      spurious and of no significance for data interpretation.
11          Kim et al. (2001) evaluated the effects of O3 inhalation exposure in B6C3FJ mice.
12      No increased incidence of lung tumors was found after exposure to 0.5 ppm O3 for 6 h/day,
13      5 days/week for 12 weeks.  There were statistically significant differences in mean body weight
14      between O3-exposed mice and air-exposed controls, as well as significant differences in the mean
15      absolute and relative weights of several organs (e.g., liver, spleen, kidney, testes, and ovary).
16      Histopathologic examination of major organs revealed oviductal carcinomas in 3/10 O3-exposed
17      female mice.
18
19
20      AX5.3   SYSTEMIC EFFECTS OF OZONE EXPOSURE
21          Mathematical models of O3 dosimetry predict that essentially no O3  penetrates to the blood
22      of the alveolar capillaries, and thus is unlikely to enter the bloodstream (see Section 5.2).
23      However, numerous studies have indicated that inhalation of O3 can produce effects in
24      lymphocytes, erythrocytes, and serum, as well as several organ systems.  The mechanism by
25      which O3 causes such systemic changes is unknown but it seems most likely that some reaction
26      product of O3, which then penetrates to the  blood and is transported to some target site, is a
27      probable mechanism. Extra-pulmonary effects could also be due to the exposure-related
28      production of mediators, metabolic products and cell trafficking. Although systemic effects are
29      of interest and indicate a very broad array of O3 effects, they are of limited influence and difficult
30      to interpret.  By protecting from respiratory tract effects, these systemic effects will likely be
        December 2005                           AX5-68      DRAFT-DO NOT QUOTE OR CITE

-------
 1      protected against also.  New studies of systemic effects are discussed here and summarized in
 2      Table AX5-8.
 3
 4      AX5.3.1  Neurobehaviorial Effects
 5           Acute exposures to increasing O3 concentrations affect animal behavior (see Table AX5-8
 6      for details of studies). Recently reported studies in adult laboratory animals confirm that relevant
 7      O3 concentrations from 0.2 to 1.0 ppm can decrease motor activity and affect short- and long-term
 8      memory.  This has been shown in passive avoidance conditioning studies in rats (Rivas-Arancibia
 9      et al., 1998; Avila-Costa et al., 1999; Dorado-Martinez et al., 2001), and in water-maze learning
10      tasks in mice (Sorace et al., 2001). The effects have been attributed to reactive oxygen/nitrogen
11      species and/or ozonation products. The memory deficits could be blocked by administration of
12      vitamin E (Guerrero et al, 1999) or taurine (Rivas-Arancibia et al., 2000).  Rivas-Arancibia et al.
13      (2003) demonstrated in rats that 1 ppm of O3 for 4 h caused increased freezing and decreased
14      exploratory behaviors that were accompanied by decreased serotonin levels and increased levels
15      of NO, glutamate, dopamine and striatal lipoperoxidation.  Morphological changes were also
16      observed in O3-exposed animals, including neuronal cytoplasm and dendrite vacuolation and
17      dilation of rough endoplasmic reticulum cisterns,  which the authors interpret as a
18      neurodegenerative process resulting from the oxidative stress of acute O3 exposure. A recent
19      study by Nino-Cabrera et al. (2002) reports that 0.7 ppm O3 exposure for 4 h can induce
20      ultrastructural alterations in the hippocampus and prefrontal cortex in aged rats, areas of the  brain
21      where degenerative age-related changes in learning and memory functions have been reported
22      (Bimonte et al., 2003).
23           In a series of studies reviewed by Paz (1997), animals acutely exposed to O3 concentrations
24      from 0.35 to 1.0 ppm demonstrated significant alterations of electroencephalographic (EEG)
25      patterns during sleep. For example, rats and cats both showed loss of paradoxical sleep time after
26      2 to 8 h of O3 exposure (Paz and Bazan-Perkins, 1992; Paz and Huitron-Resendiz,  1996).
27      A permanent 50% loss of paradoxical sleep time,  as well as increased total wakefulness and
28      alterations in circadian rhythm, were shown in rat pups born to dams exposed to 1.0 ppm O3
29      during the entire period of gestation (Haro and Paz, 1993).  The sleep pattern effects were
30      associated with alterations in brain neurotransmitter levels (Huitron-Resendiz et al., 1994;
31      Gonzalez-Pina and Paz, 1997) and most likely caused by O3 reaction products or prostaglandins

        December 2005                          AX5-69      DRAFT-DO NOT QUOTE OR CITE

-------
o

O

to
O
O
Table AX5-8.  Systemic Effects of Ozone
X
Ozone Concentration
ppm
jig/m3 Duration
Species
Effects3
Reference
NEUROBEHAVIORAL EFFECTS
0.1
0.2
0.5
1.0

0.1
0.4
0.7
1.1
1.5
0.3
0.6

196 4h
392
980
1,960

196 4h
784
1,372
2,156
2,940
588 30 days
1,176

Rat
Wistar
male


Rat
Wistar
male


Mouse
CD-I
M, F
Rats exposed for 4 h to 0.2, 0.5, and 1 ppm O3 showed long-term memory
deterioration and decreased motor activity, which was reversed 24 h later.
Brain and pulmonary Cu/Zn SOD levels were increased in animals
exposed to 0.1, 0.2, and 0.5 ppm O3, but decreased in animals exposed
to 1 ppm O3.
O3 caused memory impairment at > 0.7 ppm (one trial passive avoidance
test), decreased motor activity at > 1.1 ppm, and increased lipid
peroxidation at > 0.4 ppm. Lipid perioxidation levels from the frontal
cortex, hippocampus, striatum and cerebellum increased with increasing
O3 concentration.
O3 exposure slightly but selectively affected neurobehavioral performance
in male mice assessed with a 5-min open-field test on exposure days 4 and
19 and on day 3 after the end of the exposure. O3 exposure, however, did
Rivas-Arancibia
etal. (1998)



Dorado-Martinez
etal. (2001)



Sorace et al.
(2001)

                                                                 not grossly affect neurobehavioral development.  Reversal learning in the
                                                                 Morris water maze test was consistently impaired in both prenatally and
                                                                 adult exposed mice. In addition, longer latency to step-through in the first
                                                                 trial of the passive avoidance test and a decrease in wall rearing in the hot-
                                                                 plate test were recorded in O3 prenatally exposed mice.  Except for the
                                                                 first open-field test, altered responses were observed only in animals
                                                                 exposed to 0.3 ppm O3.
0.35
0.75
1.5

0.7



686 12 h
1,470
2,940

1,372 4 h



Rat
Wistar
male

Rat



O3 exposure decreased paradoxical sleep after 2 h of exposure, and
increased slow wave sleep after 12 h of exposure at all O3 concentrations;
5-HT concentrations in the pons increased with increasing O3
concentration.
Vitamin E administered before or after O3 exposure blocked memory
deterioration (passive avoidance)and increases in lipid peroxidation levels
in the striatum, hippocampus and frontal cortex that were associated with
oxidative stress.
Paz and Huitron-
Resendiz (1996)


Guerrero et al.
(1999)



-------
o

O

to
O
O
Table AX5-8 (cont'd).  Systemic Effects of Ozone
X
Ozone Concentration
ppm Hg/m3 Duration Species
NEUROBEHAVIORAL EFFECTS (cont'd)
0.7 1,372 4h Rat
Wistar
male
27 months old
0.7 1,372 4h Rat
0.8 1,568
1 1,960 12h/day Rat
during dark
period
1 1,960 4h Rat
Wistar
male
1 1,960 3h Rat
1.5 2,940 24 h Rat
Wistar
male
Effects3

O3 exposure increased ultrastructural alterations in the hippocampus and
prefrontal cortex in aged rats compared with controls. These areas are
related to learning and memory functions, which are the first degenerative
aging changes observed.
Taurine (43 mg/kg) given before or after O3 exposure improved memory
deterioration in an age-specific manner. Old rats showed peroxidation in
all control groups and an improvement in memory with taurine. When
taurine was applied before O3, peroxidation levels were high in the frontal
cortex of old rats and the hippocampus of young rats; in the striatum,
peroxidation caused by O3 was blocked when taurine was applied either
before or after exposure.
O3 exposure during pregnancy affects the neural regulation of paradoxical
sleep and circadian rhythm of rat pups 30, 60, and 90 days after birth.
O3 caused alterations in long-term memory and a significant reduction of
dendritic spines. Results provide evidence that deterioration in memory is
probably due to the reduction in spine density in the pyramidal neurons of
the hippocampus.
O3 or its reaction products affect the metabolism of major neurotransmitter
systems as rapidly as after 1 h of exposure. There were significant
increases in dopamine (DA), and its metabolites noradrenaline (NA) and
3,4 dihydroxyphenylacetic acid (DOPAC), and 5-hydroxyindolacetic acid
(5-HIAA) in the midbrain and the striatum.
Adult rats exposed to O3 spend decreased time in wakefulness and
paradoxical sleep and a significant increase in time in slow-wave sleep.
Neurochemical changes include increased metabolism of serotonin in
the medulla oblongata, pons, and midbrain.
Reference

Nino-Cabrera
et al. (2002)
Rivas-Arancibia
et al. ( 2000)
Haro and Paz
(1993)
Avila-Costa et al.
(1999)
Gonzalez-Pina
and Paz (1997)
Huitron-Resendiz
etal. (1994)

-------
o

O

to
O
O
Table AX5-8 (cont'd).  Systemic Effects of Ozone
X
to
Ozone
ppm
Concentration
jig/m3 Duration
Species
Effects3
Reference
NEUROENDOCRINE EFFECTS
0.5
0.5 to 3
1.0
980 20 h/day
for 5 days
.0 980 to 3 h
5,880
1,960 24 h
Rat
Rat
Sprague-
Dawley
male
Rat
Sprague-
Dawley
male
O3 produced marked neural disturbances in structures involved in the
integration of chemosensory inputs, arousal, and motor control. O3
inhibited tyrosine hydroxylase activity in noradrenergic brainstem cell
groups, including the locus ceruleus (-62%) and the caudal A2 subset
(- 57%). Catecholamine turnover was decreased by O3 in the cortex
(- 49%) and striatum (- 18%) but not in the hypothalamus.
Hyperthyroid, T4-treated rats (0.1 - 1.0 mg/kg/day for 7 days) had
increased pulmonary injury (BALF LDH, albumin, PMNs) at 18 h PE
compared to control rats.
Hyperthyroid, T3 -treated rats had increased metabolic activity and
Oj-induced pulmonary injury, but lipid peroxidation, as assessed by
alkane generation, was not affected.
Cottet-Emard
etal. (1997)
Huffman et al.
(2001)
Sen etal. (1993)
CARDIOVASCULAR EFFECTS
0.1
0.3
0.5
196 5h
588
980
Rat
Wistar young
(4-6 month)
and old
(22-24 month)
Transient rapid shallow breathing with slightly increased HR appeared
1-2 min after the start of O3 exposure, possibly due to olfactory sensation;
persistent rapid shallow breathing with a progressive decrease in HR
occurred with a latent period of 1-2 h. The last 90-min averaged values
for relative minute ventilation tended to decrease with the increase in the
Arito etal. (1997)
                                                               level of exposure to O3 and these values for young rats were significantly
                                                               lower than those for old rats. An exposure of young rats to 0.1 ppm O3 for
                                                               shorter than 5 h significantly decreased the tidal volume and HR and
                                                               increased breathing frequency, but no significant changes were observed
                                                               in old rats.  There were no differences between young and old rats in
                                                               non-observable-adverse-effect-levels (NOAELs) for the O3-induced
                                                               persistent ventilatory and HR responses, when the NOAELs were
                                                               determined by exposure to 0.3 and 0.5 ppm O3.

-------
o

O

to
O
O
Table AX5-8 (cont'd).  Systemic Effects of Ozone
X
Ozone Concentration
ppm Hg/m3 Duration
CARDIOVASCULAR EFFECTS (cont'd)
0.1 196 8h/day
0.3 588 for 4 days
0.5 980
0.25 to 2.0 490 to 2 h to 5 days
3920
0.5 588 6 h/day
23 h/day
for 5 days
0.5 588 8h
Species

Rat
Wistar
male
Rat
Mouse
Guinea pig
Rat
F-344
male
Rat
F-344
male
Effects3

Orcadian rhythms of HR and core body temperature were significantly
decreased on the first and second O3 exposure days in a concentration
dependent manner, and returned to control levels on the third and fourth
days.
Robust and consistent decreases in HR and core body temperature; smaller
decreases in metabolism, minute ventilation, blood pressure, and cardiac
output that vary inversely with ambient temperature and body mass.
Minimal extrapulmonary effects were observed at a core body temperature
of 34 °C; O3 exposures at 22 and 10 °C produced significant decreases in
heart rate (160 and 210 beats/min, respectively), core body temperature
(2.0 and 3.5 °C, respectively), and body weight (15 and 40 g,
respectively). Decreases in these functional parameters reached their
maxima over the first 2 exposure days and returned to control levels after
the 3rd day of exposure.
O3 exposure increased atrial natriuretic peptides in the heart, lung, and
circulation, suggesting they mediate the decreased BP and pulmonary
edema observed with similar O3 exposures.
Reference

Iwasaki et al.
(1998)
Watkinson et al.
(2001)
Watkinson et al.
(1995);
Highfill and
Watkinson (1996)
Vesely et al.
(1994a,b,c)

-------
o

O

to
O
O
Table AX5-8 (cont'd).  Systemic Effects of Ozone
X
Ozone Concentration
ppm
Hg/m3
REPRODUCTIVE AND
0.2
0.4
0.6

0.3
0.6
0.9


0.3
0.6



0.4
0.8
1.2

392
784
1,176

588
1,176
1,764


588
1,176



784
1,568
2,352

Duration
Species
Effects3
Reference

DEVELOPMENTAL EFFECTS
Continuous up
to day 17 of
pregnancy

Continuous up
to postnatal
day 26


Continuous
until
gestational
day 17

Continuous
during
gestation
days 7-17
Mouse
CD-I


Mouse
CD-I



Mouse
CD-I



Mouse
CD-I


No significant effects on either reproductive performance, postnatal
somatic and neurobehavioral development (as assessed by a Fox test
battery) or adult motor activity (including within-session habituation);
some subtle or borderline behavioral deficits were noted, however.
O3 caused subtle CNS effects but did not affect the animals' capability to
learn a reflexive response (limb withdrawal); females exposed to 0.6 ppm
O3 showed a reduced preference for the right paw than both their same-sex
controls and 0.6 ppm males. The effect was more robust in the case of an
organised avoidance response (wall-rearing).
Exposure to O3 did not grossly affect neurobehavioral development,
as assessed by somatic and sensorimotor development (postnatal day
(PND) 2-20), homing performance (PND 12), motor activity (PND 21),
passive avoidance (PND 22-23), water maze performances (PND 70-74),
and response to a nociceptive stimulus (PND 100).
No effect of O3 on reproductive performance; no significant somatic
developmental effects in Cyexposed pups except for a delay in eye
opening that was not concentration dependent.

Petruzzi et
(1995)


Petruzzi et
(1999)



al.



al.




Sorace et al.
(2001)



Bignami et
(1994)






al.




-------
o

O

to
O
O
                                           Table AX5-8 (cont'd).  Systemic Effects of Ozone
        Ozone Concentration
          ppm
              jig/m3     Duration
                Species
                                                                Effects3
                                                                                    Reference
X
REPRODUCTIVE AND DEVELOPMENTAL EFFECTS (cont'd)
           0.6
               1,176
Continuous
from birth to
weaning
Mouse
CD-I
           1.0
           2.0
               1,960
               3,920
3h
Rat
Sprague-
Dawley
female
                                                                    DeH'Omo et al.
                                                                    (1995a,b)
Exposure to O3 did not produce any significant impairment of the
acquisition phase during swimming navigation, a sensitive indicator
for hippocampal damage; however, O3 slightly increased the swimming
paths during the last day of the reversal phase. Mice exposed to O3
showed a slightly but significantly higher swimming speed during all
the days, which was unrelated to differences in body weight and to
navigational performances. Moreover, mice exposed to O3 (with the
exception of one animal) had a strong tendency to make turns to the left
while the controls, independent of sex, preferred clockwise turns.

High O3 exposure stimulates hepatocytes to produce increased amounts of   Laskin et al.
nitric oxide as well as protein, possibly mediated by cytokines such         (1994, 1996,
as TNF-oc produced by alveolar macrophages. When macrophage function  1998); Laskin and
is blocked, hepatic injury induced by O3 is prevented.                     Laskin (2001)
           2.0
              3,920
2h
Rat
F-344
Utilizing electron paramagnetic resonance (EPR) spectroscopy of
chloroform extracts of liver homogenates, a significant flux of hydrogen
peroxide produced from the reaction of O3 with lipids of the extracellular
lining could be a source of biologically relevant amounts of hydroxyl
radical.  EPR signals for carbon-centred alkoxyl and alkyl adducts were
detected with C-phenyl N-tert-butyl nitrone (PEN) in the liver of animals
exposed to O3.
                                                                    Vincent et al.
                                                                    (1996)

-------
o

O

to
O
O
Table AX5-8 (cont'd).  Systemic Effects of Ozone
X
Ozone Concentration
ppm
EFFECTS
0.5
0.8
0.8
1.0
10.0
1.0
5.0
10.0
1.0
5.0
10.0
jig/m3 Duration
ON CUTANEOUS TISSUE
980 2h
1,568 6 h
1,568 2 h
1,960
19,600
1,960 2 h
9,800
19,600
1,960 2 h
9,800
19,600
Species

Mouse
hairless
female
Mouse
SKH-1
hairless
Mouse
SKH-1
hairless
Mouse
SKH-1 hairless
Mouse
Effects3

a tocopherol levels in the stratum corneum (SC) were not affected by O3
exposure (0.5 ppm) alone, but were significantly depleted by combined
exposure to UV and O3.
Increased lipid peroxidation in the skin epidermis and dermis activated
stress proteins HSP27 and HO-1, and activated a proteolytic enzyme
system (MMP-9) related to matrix injury and repair processes.
High O3 depletes hydrophilic antioxidants in the SC: Vit. C decreased to
80%, GSH decreased to 41%, and uric acid decreased to 44% of control
levels after exposure to > 1.0 ppm O3
Vit. E levels decreased and malondialdehyde levels increased in the SC
with increasing O3 concentration.
High O3 exerts an oxidizing effect on the outermost layer of the skin (SC);
depletes low-molecular-weight antioxidants (a tocopherol, vit. C,
glutathione, uric acid) in a concentration dependent manner; increases
malondialdehyde levels associated with lipid peroxidation
Reference

Valacchi et al.
(2000)
Valacchi et al.
(2003)
Weber et al.
(2000)
Thiele et al.
(1997a)
Weber et al.
(2000)
       aRER = Rough endoplasmic reticulum.
       PE = Postexposure (i.e., time after O3 exposure ceased).
       TSH = Thyroid stimulating hormone.
       T3 = Triiodothyronine.
       T4 = Thyroxine.
       cyt. = Cytochrome.
       NADPH = Reduced nucotinamide adenine dinucleotide phosphate.
       NADH =
       B[a]P = Benzo[a]pyrene.
                            NK = Natural killer.
                            PHA = Phytohemagglutin.
                            ConA = Concanavalin A.
                            LPS = Lipopolysaccharide.
                            SRBC = Sheep red blood cell.
                            TEA = Thiobarbituric acid.
                            ONP =
                            IgE = Immunoglobulin E.

-------
 1      (Koyama and Hayaishi, 1994). The permanent effects in pups caused by high O3 exposure during
 2      gestation were attributed to the diminished antioxidant capability of fetal tissue (Giinther et al.,
 3      1993).
 4
 5      AX5.3.2  Neuroendocrine Effects
 6           Several new studies have examined the interaction of O3 with the pituitary-thyroid-adrenal
 7      axis. Sen et al. (1993) found that T3 supplementation increased metabolic rate and pulmonary
 8      injury in the lungs of O3-treated animals. Increased toxicity to O3 was later reported in
 9      hyperthyroid rats by Huffman et al. (2001).
10           Mechanisms involved in the interaction of O3 and the neuroendocrine system are still are
11      not well understood. Cottet-Emard et al. (1997) studied the effects of exposure to 0.5 ppm O3 for
12      5 days on catecholamine activity in rat sympathetic efferents and brain areas of prime importance
13      to adaptation to environmental stressors. Catecholamine activity was assessed by estimating the
14      turnover rate of catecholamines and in vivo tyrosine hydroxylase activity in peripheral and central
15      structures (i.e., heart, lungs, superior cervical ganglia, cerebral cortex, hypothalamus and
16      striatum), and in A2 cell groups within the nucleus tractus solitarius (NTS) and locus ceruleus
17      (A6). Ozone inhibited norepinephrine turnover in heart (-48% of the control level) but not in
18      lungs and failed to modify the tyrosine hydroxylase activity in superior cervical ganglia, and the
19      catecholamine content in the adrenal glands.  In the central nervous system, O3 inhibited tyrosine
20      hydroxylase activity in noradrenergic brainstem cell groups, including the locus ceruleus (-62%)
21      and the caudal A2 subset (- 57%). Catecholamine turnover was decreased by O3 in the cortex
22      (-49%) and striatum (-18%) but not in the hypothalamus. These data show that high ambient
23      levels of O3 can produce marked neural disturbances in structures involved in the integration of
24      chemosensory inputs, arousal, and motor control, effects that may be responsible for some of the
25      behavioral effects described in Section 5.2.1.
26           High, non-ambient levels of O3 (e.g., > 1.0 ppm) have been shown to affect visual and
27      olfactory neural pathways in the  rat. Custodio-Ramierez and Paz (1997) reported a significant
28      delay in visual evoked potentials recorded in the visual cortex and the lateral geniculate nucleus
29      of male Wistar rats acutely exposed for 4 h to high levels of O3 (1.5, and 3.0 ppm). Also using
30      Wistar rats, Colin-Barenque et al. (1999) reported cytological and ultrastructural changes in the
31      granule layer of the olfactory bulb after a 4-h exposure to 1 to 1.5 ppm O3.

        December 2005                           AX5-77     DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX5.3.3  Cardiovascular Effects
 2           New studies evaluating the cardiovascular effects of O3 have monitored continuous
 3      cardiovascular and ventilatory measurements in unanesthetized rats exposed to O3. Using the
 4      head-out plethysmograph for continuous measurements, Arito et al. (1997) exposed rats to
 5      filtered air for 6 h, followed 2 days later by a 5-h exposure to 0.1 ppm O3, 5 days later by a 5-h
 6      exposure to 0.3 ppm O3, and 10 days later by a 5-h exposure to 0.5 ppm O3. Each of the O3
 7      exposures was preceded by a 1-h exposure to filtered air. Transient rapid shallow breathing with
 8      slightly increased HR appeared 1-2 min after the start of O3 exposures and was attributed to an
 9      olfactory response. Persistent rapid shallow breathing with a progressive decrease in HR
10      occurred with a latent period of 1-2 h.  During the last 90-min of exposure, averaged values for
11      relative minute ventilation tended to decrease with the increase in O3 concentration for young
12      (4-6 months) but not old (20-22 months) rats.
13           In a series of studies utilizing radiotelemetry transmitters for monitoring ECG, HR, core
14      body temperature (Tco), and motor activity in unanesthetized and unrestrained rats, Watkinson
15      et al. (1995; 2001) and Highfill and Watkinson (1996) demonstrated that when HR was reduced
16      during O3 exposure, the Tco decreased in association with reduced activity. The decreases in body
17      temperature and associated decreases in blood pressure reported by Watkinson et al., and also by
18      Arito et al. (1997), suggested that the pattern and magnitude of the ventilation and HR responses
19      were mediated through some physiological and behavioral defense mechanisms acting to
20      minimize the irritant effects of O3 inhalation.
21           The adaptation of cardiovascular and thermoregulatory responses to O3 also was reported by
22      Iwasaki et al. (1998) in ECG electrode- and thermistor sensor implanted rats after repeated
23      exposure to 0.1, 0.3, and 0.5 ppm O3 8 h/day for 4 consecutive days.  Circadian rhythms of HR
24      and Tco were disrupted on the first and second O3 exposure days in a concentration dependent
25      manner. The 8-h and 12-h averaged values of HR and Tco decreased significantly on the first and
26      second exposure days and recovered to control values after small but significant rebound
27      increases  on the third and fourth days of O3 exposure.
28           More recent reports by Watkinson et al. (2003) further examined the thermoregulatory
29      response to O3 exposures. Male Fischer-344 rats were exposed to one of three possible O3 levels (
30      0.0 ppm x 24 h/day [air], 0.5 ppm x 6h/day [intermittent] or 0.5 ppm x 23 h/day [continuous]) at
31      one of three temperatures (10° C [cold], 22° C [room], or 34° C [warm]) for a total of 9 treatment

        December 2005                          AX5-78       DRAFT-DO NOT QUOTE OR CITE

-------
 1      groups. Another protocol examined the effects of O3 exposure (0.5 ppm) and exercise described
 2      as rest, moderate, heavy or CO2-stimulated ventilation. Both intermittent and continuous O3
 3      exposure caused decreases in HR and Tco and increases in BALF inflammatory markers. Exercise
 4      in filtered air caused increases in HR and Tco while exercise in O3 caused decreases in those
 5      parameters. Carbon dioxide and O3 induced the greatest deficits in HR and Tco. The authors
 6      discuss several factors which may modulate the hypothermic response, including: 1) dose, 2)
 7      animal mass (i.e., smaller animals show a greater response), and 3) environmental stress (e.g.,
 8      restraint, exercise).  The authors further discuss possible problems with extrapolation to humans
 9      that may be caused by this response (discussed in Chapter 4).
10           The tissue edema reported in the heart and lungs of laboratory animals exposed to relatively
11      high ambient O3 concentrations (> 0.5 ppm) may be  caused by increased circulating levels of
12      atrial natriuretic factor (ANF) which is known to be  a possible mediator of increased capillary
13      permeability, vasodilation, and decreased blood pressure (Daly et al., 2002; Vesely et al.
14      (1994a,b,c) reported increased levels of ANF in the heart, lungs, and circulation of rats exposed to
15      0.5 ppm O3 for 8 h.
16
17      AX5.3.4   Reproductive and Developmental Effects
18           New studies on the developmental effects of O3 demonstrate that prenatal exposures to  O3
19      concentrations < 1.0 ppm do not cause major or widespread somatic or neurobehavioral effects in
20      the offspring of laboratory animals. Animal studies  evaluating O3-induced reproductive effects
21      have not been completed.
22
23      Developmental Effects
24           A study of somatic and neurobehavioral development was reported by Bignami et al. (1994)
25      in female CD-I mice exposed during pregnancy (days 7 to 17) to O3 concentrations of 0, 0.4, 0.8,
26      or 1.2 ppm. They did not find any O3 effects on reproductive or behavioral performance, or on
27      neonatal mortality, but found a significant decrease in body weight gain and delayed eye opening
28      in pups in the  1.2-ppm exposure group.  A follow-up study by Petruzzi et al. (1995) did not find
29      any significant deficits in reproductive performance, postnatal somatic and neurobehavioral
30      development, or adult motor activity in CD-I mice exposed in utero from conception through day
31      17 of pregnancy to 0, 0.2, 0.4, and 0.6 ppm O3. In a  subsequent study by Petruzzi et al. (1999),

        December 2005                          AX5-79      DRAFT-DO NOT QUOTE OR CITE

-------
 1      subtle changes in handedness and morphine reactivity were found when the O3 exposures (0.3,
 2      0.6, or 0.9 ppm) continued postnatally until weaning [post natal day (PND) 26].  Female mice
 3      exposed to 0.6 ppm O3 showed a reduced preference for the right paw at PND 70, and 0.9 ppm O3
 4      altered hot plate avoidance after IP treatment with morphine (10 mg/kg) at PND 100.
 5           Dell'Omo et al. (1995a,b) exposed CD-I mice to 0.6 ppm O3 from birth through weaning
 6      (PND 22 or 26). Swimming navigation was tested (DeH'Omo et al., 1995a) at 12 to 13 weeks of
 7      age using acquisition and reversal trials. Exposure to O3 did not produce any significant
 8      impairment of navigational performance during acquisition and only subtle changes during
 9      reversal. As in previous studies, there were no significant effects of O3 on reproductive
10      performance but O3 offspring showed a significant reduction in body weight. Ozone effects on
11      neurobehavioral development were not large and very selective, with some attenuation of activity
12      responses and impairment of passive avoidance acquisition (DeH'Omo et al.  (1995b). Similarly,
13      only small  and selective effects on somatic and sensorimotor development were found in the
14      offspring of CD-I mice continuously exposed from 30 days prior to the formation of breeding
15      pairs  until PND 17 to 0.0, 0.3, or 0.6 ppm O3 (Sorace et al., 2001).
16
17      AX5.3.5  Effects on the Liver, Spleen, and Thymus
18      Liver
19           New  studies on the effects of O3 on liver showed that, in rats, high (1 to 2 ppm) acute O3
20      exposures caused increased production of NO by hepatocytes and enhanced protein synthesis
21      (Laskin et al., 1994; 1996).
22           The O3-associated effects shown in the liver are most likely mediated by inflammatory
23      cytokines (e.g., TNF alpha) or other cytotoxic mediators (e.g., hydroxyl radicals) released by
24      activated macrophages in the lungs (Vincent et al., 1996; Laskin et al., 1998; Laskin and Laskin,
25      2001). Except for the earlier work on xenobiotic metabolism, the responses occurred after very
26      high acute  O3 exposures.
27
28      Spleen and Thymus
29           Ozone has been shown to primarily affect T-cell mediated systemic immunity.  New studies
30      evaluating  O3-induced effects on spleen and thymus have not been completed.
31

        December 2005                           AX5-80      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX5.3.6   Ozone Effects on Cutaneous and Ocular Tissues
 2           Ground-level smog exposure not only affects various organ systems, when inhaled, but may
 3      potentially have direct effects on exposed skin and eyes. Several new studies have examined the
 4      effects of O3 on skin.  Ozone can have an oxidizing effect on the outermost layer of the skin,
 5      called the stratum corneum (SC) where it may compromise skin barrier function and possibly
 6      induce an epidermal proinflammatory response (e.g., Weber et al., 2001; Cotovio et al., 2001;
 7      Thiele, 2001); however, these cutaneous effects of O3 are found only at very high concentrations
 8      used in experimental studies and have not been shown at more relevant ambient- or near-ambient
 9      levels (< 0.5 ppm) of O3 exposure.  The lack of ambient-level O3 effects on the skin is most likely
10      due to a well-developed defense against oxidative stress, utilizing nonenzymatic (e.g., vitamin C
11      and E, glutathione, uric acid, a-tocopherol) and enzymatic (e.g., superoxide dismutase, catalase,
12      glutathione reductase and peroxidase) antioxidants found in many living organisms (Cross et al.,
13      1998).
14           Effects of ground-level smog on the eyes have also been reported, but generally are
15      attributed to related photochemical oxidants like peroxyacetyl nitrate (Vyskocil et al., 1998) or
16      possibly to atmospheric O3 precursors or reaction products like  aldehydes. Ocular tissues also
17      have similar antioxidant protective function as in the skin, but are not well studied (Mucke, 1996;
18      Rose etal., 1998).
19           The cutaneous effects of O3 exposure were first reported by Thiele et al. (1997a,b,c,d).
20      Hairless mice (SKH-1) were exposed to 1, 5, and 10 ppm O3 for 2 h or to 1 ppm O3 for 2 h on six
21      consecutive days and skin layers from the epidermis and dermis were separated for analysis of
22      antioxidants and lipid peroxidation products.  Decreased antioxidant levels (a-tocopherol;
23      ascorbic acid) were found in the upper epidermis and increased malondialdehyde (MDA), a lipid
24      peroxidation product, was found in both the upper and lower epidermal layers. Effects were dose
25      dependent and became significant (p < 0.05) for a-tocopherol and MDA after single  exposures to
26      >1.0 and >5.0 ppm O3, respectively. Repeated O3 exposures caused significantly higher MDA
27      concentrations. The  effects of MDA accumulation could be prevented by enriching the skin with
28      vitamin E.
29           In the SKH-1 hairless mouse model, Weber et al. (1999, 2000, 2001) demonstrated that
30      0.8 to 10 ppm O3 exposure for 2 h depletes the low molecular weight antioxidants (e.g.,
31      a-tocopherol, vitamin C, glutathione, uric acid) in the SC in a dose-response manner and causes

        December 2005                          AX5-81      DRAFT-DO NOT QUOTE OR CITE

-------
 1      increased MDA. The effects were statistically significant (p < 0.05) at > 1 ppm O3 and > 5 ppm
 2      for antioxidant depletion and increased MDA, respectively.  Reactive aldehydes were found in the
 3      epidermis at the highest O3 concentration tested, though this level of exposure is of limited
 4      relevance.  Preexposure to O3 followed by low-dose ultraviolet (UV) radiation (0.33 MED)
 5      decreased the significance level to 0.5 ppm O3, probably through combined oxidative stress on the
 6      SC (Valacchi et al., 2000). Stress-inducible proteins (e.g., heme oxygenase-1) and other heat
 7      shock proteins (e.g., HSP27 and HSP70) were found in deeper cellular layers of the epidermis
 8      after 2 h of exposure to 8.0 ppm O3 (Valacchi et al., 2002).  Prolonged exposure to lower
 9      concentrations of O3  (0.8 ppm) for 6 h also induces cellular stress responses that included the
10      formation of HNE  protein adducts, HSP27, and heme-oxygenase-1  in the deeper cellular layers of
11      the skin that continued for up to 18 h after O3 exposure, followed by repair processes (Valacchi
12      et al., 2003).
13
14
15      AX5.4    INTERACTIONS OF OZONE WITH OTHER CO-OCCURRING
16                POLLUTANTS
17           Ozone is part of a complex mixture of air pollutants with a composition and pattern that
18      varies geographically and temporally (by hour of the day, day of the week, and season). Health
19      effects caused by the complex mixture are undoubtedly different (either subtly or significantly)
20      from the additive effects of a few of the hundreds of compounds present. The only disciplinary
21      approach that can evaluate a "real-world" complex mixture is epidemiology (Chapter 7).
22      However, because  of the difficulty in evaluation of causative factors and quantitative
23      relationships in epidemiology studies, it is useful to consider animal toxicological studies of
24      mixtures. Such studies can be divided into three categories: (1) ambient air mixtures,
25      (2) laboratory-generated complex mixtures (e.g., gasoline combustion mixtures having
26      ultraviolet-irradiation, other reaction mixtures with O3 and several other components), and
27      (3) binary mixtures.  In most cases, experimental designs in the first two classes did not have an
28      O3-only group, making it difficult to impossible to discern the influence of O3. The more recent
29      mixture studies that are discussed here typically have been with NO2, sulfuric acid (H2SO4), or
30      ammonium sulfate ([NH4]2SO4).
       December 2005                           AX5-82      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Interpreting the mixture studies in terms of real-world risk is difficult because laboratory
 2      exposure patterns do not always represent real-world exposure patterns. For example, in the real
 3      world, NO2 often peaks before O3 peaks, with a mixture occurring between the peaks, but most
 4      laboratory exposures used mixtures only. Also, most studies of O3 and NO2 mixtures used
 5      ambient levels of O3 and levels of NO2 high above ambient. As shall be seen, all interaction
 6      possibilities have occurred, depending upon the composition of the mixture, the endpoint
 7      examined, and the exposure regimen.  In some cases, no interaction was found.  Most often,
 8      additivity (the effects of the mixture are equal to the sum of the effects of the individual
 9      components) or synergism (the effects of the mixture are greater than the sum of the effects of the
10      individual components) was observed. Antagonism (the effects of the mixture are less than the
11      sum of the individual components) was rarely found.
12
13      AX5.4.1   Ozone and Nitrogen Oxides
14           The most commonly studied copollutant in binary mixtures with O3 is NO2. New studies
15      evaluating the effects of combined O3/NO2 exposures are listed in Table AX5-9.
16           Recent work has demonstrated that chronic exposures of rats to 0.8 ppm O3 and 14.4 ppm
17      NO2 for 6 h/day caused the rats to develop insufficiency and severe weight loss (Farman et al.,
18      1997). Half of these animals died after 55 to 78 days of exposure due to severe fibrosis.
19      Biochemical analysis of lung tissue demonstrated increased total lung collagen and elastin, with
20      loss of mature collagen, indicating there was breakdown and remodeling of the lung parenchyma.
21      In follow-up rat studies, Farman et al. (1999) reported a sequence of events starting with
22      increasing inflammatory and mild fibrotic changes for the first 3 weeks of exposure to 0.8 ppm O3
23      and 14.4 ppm NO2,  stabilized or even reduced changes after 4 to 6 weeks, and severe increases
24      over 7 to 9 weeks of exposure.  The authors speculated that repair processes occurring during the
25      middle 4 to 6 weeks of exposure become overwhelmed, leading to progressive fibrosis after 7 to 8
26      weeks of exposure.  When combined exposures to 0.8 ppm O3  and 14.4 ppm NO2 for 6 h/day,
27      7 days/week were extended for 90 days, Farman et al. (1999) found that the lesion extended far
28      into the acinus, but the extent of tissue involvement was the same after 7, 78, and 90 days of
29      exposure. At the end of exposure, in situ hybridization for procollagen types I and  III
30      demonstrated high levels  of messenger RNA within central acini in the lungs from the combined
31      exposure group, but not in lungs from the rats exposed to O3 and NO2 alone.

        December 2005                           AX5-83      DRAFT-DO NOT QUOTE OR CITE

-------
O
O
i
{_} '
CD
>-l
to
O
O
v\










X
oo




O
i
H
6
l>
-^H
0
H
O
c
O
s
0
&
O
H
W

Table AX5-9.
Concentration
O3 NO2

ppm fig/m3 Hg/m3 ppm Duration
MORPHOLOGY
0.8 1,568 27,072 14.4 6 h/day ,
7 days/week
for 90 days


0.3 588 2,256 1.2 Continuous for
3 days


BIOCHEMISTRY (cont'd)

Interactions of Ozone With Nitrogen Dioxide



Species Endpoints" Interaction

Rat, M Morphometry of lung syngeristic; more
(S-D) parenchyma; DNA probes for peripheral centriacinar
10-12 weeks procollagen; in situ lesion, but same after
old mRNA hybridization 7, 78, and 90 days of
exposure.
Rat, M DNA single strand breaks; None; effect due to O3
(S-D) polyADPR synthetase of
3 months old AMs; total cells, protein, and
LDHinBALF

0.8 1,568 27,072 14.4 6 h/day, 7 days/week Rat, M (S- Lung hydroxyproline, Synergistic; fibrosis
for 9 weeks




BAL = Bronchoalveolar lavage.
PG = Prostaglandin.
G-6-PD = Glucose-6-phosphate dehydrogenase.
GOT = GSH-disulfide transhydrogenase.
GSHPX = GSH peroxidase.










D) 10-12 hydrooxypyridinium, DNA, after 7-8 weeks of
weeks old and protein content of whole exposure
lung; morphology and
labeling index

SOD = Superoxide dismutase.
DR = Disulfide reductase.





Reference

Farman et al.
(1999)



Bermudez et al.
(1999,
Bermudez
(2001)

Farman et al.
(1997)






NADPH-CR = Reduced nicotinamide adenine dinucleotide phosphate-cytochrome c reductase.
GSH = Glutathione.
6-PG-D = 6-phosphogluconate dehydrogenase.























-------
 1           Bermudez et al. (1999) reported that a 3-day exposure to 0.3 ppm O3, and the combined
 2      exposure of O3 and 1.2 ppm NO2, caused significant DNA single-strand breaks in the alveolar
 3      macrophages of Sprague-Dawley rats.  No changes were caused by NO2-only exposure.  In a
 4      follow-up study, Bermudez (2001) showed that the same exposures stimulated the activity of
 5      polyADPR synthetase, an indicator of response to lung cellular DNA repair caused by oxidant-
 6      induced lung injury.
 7           Other published reports (Ishii et al., 2000b; Weller et al., 2000) indicate that the laboratory
 8      animal model of progressive pulmonary fibrosis, utilizing long-term, combined O3 (0.4 to
 9      0.8 ppm) and high-level NO2 (7 to 14 ppm) exposure, causes an initial acute pulmonary
10      inflammation, followed by adaptation and repair, and eventually causing pulmonary fibrosis after
11      6 to 13 weeks of exposure.
12
13      AX5.4.2   Ozone and Other Copollutants
14           Although the bulk of the toxicological database for binary mixtures of O3 involves NO2 or
15      acidic sulfate and nitrate aerosols (see Section 5.4.3), a few studies have examined responses
16      to combinations of O3 with other single pollutants, such as formaldehyde (HCHO); or with
17      surrogates of pollutants treated as a single pollutant, such as tobacco smoke. New studies
18      evaluating coexposures of O3 with acid aerosols and particle mixtures are listed in Table AX5-10.
19
20      Interactive effects of ozone and formaldehyde.
21           A study by Cassee and Feron (1994) focused on biochemical and histopathological changes
22      in the nasal respiratory epithelium of rats exposed 8 h/day for 3 days to 0.4 ppm O3 and 3.6 ppm
23      HCHO, alone and combined. No interactive effects were found, however, despite the high levels
24      of HCHO when compared to typical ambient levels of 1 to  10 ppb (e.g., Rehle et al., 2001).  In a
25      follow-up to their previous combined exposure study in rats, Mautz (2003) elaborated on changes
26      in breathing pattern and epithelial cell proliferation attributed to O3 and HCHO. Rats were
27      exposed to 0.6 ppm O3 and 10 ppm HCHO alone and in combination for 3 h with  exercise at two
28      times resting ventilation. Even with exercise, HCHO does  not penetrate to the lower respiratory
29      tract to interact with O3, and  does not alter breathing patterns to modify local O3 dose.
30      Parenchymal injury was, therefore, due to O3 alone. In the nasal transitional epithelium  and in the
31      trachea, however, combined exposure produced additive effects due to the increased volume of

        December 2005                          AX5-85      DRAFT-DO NOT QUOTE OR CITE

-------
December 2005
Table AX5-10. Interactions Of Ozone With Particles
Concentration
03
ppm
Hg/m3
PM
mg/m3a

Durationb
Species
Endpoints0 Interaction
Reference
SULFURIC ACID
X
oo
O\
0.1
0.2
0.1
0.3
0.6
0.1
0.3
0.6
0.6
196
392
196
588
1,176
196
588
1,176
1,176
0.02-0.15
(0.4 - 0.8 urn)
0.50 (0.3 urn)
0.125(0.3 urn)
0.50 (0.3 urn)
0.125(0.3 urn)
0.5 (0.06 and
0.3 umMMD)
23. 5 h/day or
intermittent 12
h/day for up to
90 days
3h
3h
4 h/day for
2 days
Rat
S-D
male
Rabbit NZW
male
Rabbit NZW
male
Rat
Morphology No interaction
Biochemistry
AM intracellular pH Antagonism
homeostasis and H+
extrusion
Airway responsiveness Antagonism
(in vitro bronchial rings
+ ACh)
Morphology: volume Synergism: ultrafme + O3, but
percentage of total not fine
Last and
Pinkerton
(1997)
Chen et al.
(1995)
El-Fawal et al.
(1995)
Kimmel et al.
(1997)
                                                                        parenchyma containing
                                                                        injured alveolar septae;
                                                                        bromode oxyuridine      Synergism: fine + O3
                                                                        cell labeling index in
                                                                        the periacinar region
PARTICLE MIXTURES
H ai
O
o
2
o
H 0.16
O 0.30
g 0.59
S
W
&
O
H
W
196



314
588
1,156






Diesel PM (NIST
#2975) reacted with
O3 for 48 h

0.05 - 0.22 mg/m3
ammonium bisulfate
0.03 -0.10 mg/m3 C
0.11- 0.39 pmNO2
0.02 -0.11 mg/m3
HNO3
(0.3 umMMAD)


24 h (IT)



4 h/day,
3 days/week
for 4 weeks






Rat
S-D


Rat
F344N
male






Inflammation



breathing pattern,
morphology, lavagable
protein, and clearance






Synergism



Complex interactions, but
possible loss of typical
attenuation seen with O3 only
exposure, reflecting persistence
of inflammation




Madden et al.
(2000)


Mautz et al.
(2001)








-------
December 2005




X
oo
O
H
6
O
0
H
O
O
0
O
H
W
Table AX5-10 (cont'd). Interactions Of Ozone With
Concentration
O3 PM
ppm fig/m3 mg/m3a
PARTICLE MIXTURES (cont'd)
0.2 392 0.07and 0.14 mg/m3
ammonium bisulfate
(0.45 urn MMMD);
0.05 and 0.10
mg/m3 carbon
0.2 392 0.50 mg/m3
ammonium bisulfate
(0.45 urn MMMD)
and elemental
carbon)
0.3 588 0.063 to 1.57 mg/m3
CAPs (Boston) + ip
OVA sensitization
0.4 784 0.20 and 0.50
mg/m3 fine, H2O2-
coated carbon
(0.26 urn MMMD)
0.5 980 Endotoxin (IN)
100 ug 24 h and
48 h after the 3rd
O3 exposure
0.5 980 OVA (IN)
50 ul (1%)



Durationb

4 h/day,
3 days/week
for 4 weeks

4 h/day,
3 days/week
for 4 weeks;
nose only

5h
4 h/day for 1 or
5 days
8 h/day for
3 days
8 h/day for
1 day or 3
consecutive
days



Species

Rat
F344
male
22-24 months
old
Rat
F344N-NIA
22-24 months
old

Mouse
BALB/c
Rat
S-D
300 g
Rat
F344
10-12 weeks
old
Rat
Brown
Norway 10-12
weeks old



Endpoints0

BAL protein and
albumin; plasma
hydroxylase and
fibronectin

DNA labeling of
dividing lung epithelial
and interstitial cells by
5-bromo-2-
deoxyuridine
Airway function
Inflammation
Nasal morphology
Nasal morphology



Particles
Interaction

Questionable interaction

Synergism

Interaction: increased RL and
airway responsiveness
Synergism for effect on day 5
Synergism: increased
intraepithelial mucosubstances
and mucous cell metaplasia
Synergism: increased
intraepithelial mucosubstances
and mucous all metaplasia




Reference

Bolarin et al.
(1997)

Kleinman
et al. (2000)

Kobzik et al.
(2001)
Kleinman
etal. (1999)
Fanucchi et al.
(1998)
Wagner et al.
(2001a,b)
Wagner et al.
(2002)




-------
o

O

to
O
O
Table AX5-10 (cont'd).  Interactions Of Ozone With Particles
X
oo
oo
Concentration
O3 PM
ppm Hg/m3 mg/m3a
PARTICLE MIXTURES (cont'd)
0.8 1,600 0.5mg. 1.5 rag or
5 mg of PM from
Ottawa Canada
(EHC-93)
1 1,960 O.llmg/m3 ultra
fine carbon ( 25 nm
CMD) + endotoxin
OH))





1 1,960 Endotoxin (37.5
EU) for 10 minutes

1 1,960 Endotoxin (IN) 0, 2,
or 20 ug in 120 uL



Duration* Species Endpoints0

2, 4, and Rat Inflammation
7 days after
IT instillation

6 h Rat Inflammation
F344
male,
10 weeks and
22 months old;
Mouse
TSK
male, 14-
17 months old
4, 20, or 24 h Mouse Inflammation
C57BL/6J
8 weeks old
8 h, repeated Rat Lung morphometric
after 24 h F344 analysis and
inflammation


Interaction

Interaction: increased TNF-oc in
BALF


Interaction: increased PMNs and
ROS release from BALF cells for
old rats and mice primed with
endotoxin; depressed in young
rats




Synergism: increased BALF
protein and PMNs

Synergism: increased BALF
PMNs and mucin glycoprotein;
increased intraepithelial
mucosubstances and mucingene
mRNA
Reference

Ulrich et al.
(2002)


Elder et al.
(2000a,b)







Johnston et al.
(2000, 2002)

Wagner et al.
(2003)



       aVMD = Volume median diameter.
       og = Geometric standard deviation.
       MMAD = Mass median aerodynamic diameter.
       CMD = Count median diameter.
       PM = Paniculate Matter.
       OVA= Ovalbumin.
                     AED = Aerodynamic diameter.
                    bUnless indicated otherwise, whole-body exposures used.
                    °BAL = Bronchoalveolar lavage.
                     AM = Alveolar macrophage.
                     IN = Intranasal
                     IT = Intratracheal.

-------
 1     toxicants during exercise. No other combined pollutant studies have been published in the peer-
 2     reviewed literature, although two studies compared the respiratory effects of O3 to HCHO.
 3     Nielson et al. (1999) compared upper airway sensory irritation caused by HCHO concentrations
 4     up to 4 ppm to the lower airway irritation caused by O3. Using BALB/c mice, they continuously
 5     measured fB, Vt, expiratory flow, T;, Te, and respiratory patterns during acute, 30-min exposures.
 6     The NOEL for HCHO was 0.3 ppm, compared to 1.0 ppm for O3.
 7
 8     Interactive effects of ozone and tobacco smoke.
 9           Wu et al. (1997)  also reported that inhalation of cigarette  smoke evokes a transient
10     bronchoconstrictive effect in anesthetized guinea pigs.  To examine whether O3 increases airway
11     responsiveness to cigarette smoke, effects of smoke inhalation challenge on total pulmonary
12     resistance (RL) and dynamic lung compliance (Cdyn) were compared before and after acute
13     exposure to 1.5 ppm O3for 1 h.  Before O3 exposure, inhalation of two breaths of cigarette smoke
14     (7 ml) at a low concentration (33%) induced a mild and reproducible bronchoconstriction that
15     slowly developed and reached its peak (ARL = 67 ± 19%, ACdyn = -29 ± 6%) after a delay of
16     > 1 min. After exposure to O3, the same cigarette smoke inhalation challenge evoked an intense
17     bronchoconstriction that occurred more rapidly, reaching its peak (ARL = 620 ± 224%, ACdyn =
18     -35 ± 7%) within 20 s, and was sustained for > 2 min.  By contrast, sham exposure to room air
19     did not alter the bronchomotor response to cigarette smoke challenge.  Pretreatment with selective
20     antagonists of neurokinin type 1 and 2 receptors completely blocked the enhanced airway
21     responsiveness. The authors concluded that O3 exposure induced airway hyperresponsiveness to
22     inhaled cigarette smoke, which resulted primarily from the bronchoconstrictive effect of
23     endogenous tachykinins.
24           To determine the effects of aged and diluted sidestream cigarette smoke (ADSS) as a
25     surrogate of environmental tobacco smoke (ETS) on O3-induced lung injury, Yu et al. (2002)
26     exposed male B6C3F1 mice to (1) filtered air, (2) ADSS, (3) O3, or (4) ADSS followed by O3
27     (ADSS/O3). Exposure to 30 mg/m3 ADSS, 6 h/day for 3 days, followed by exposure to 0.5 ppm
28     O3 for 24 h was associated with a significant increase in the number of cells recovered by B AL
29     compared with exposure to ADSS  alone or O3 alone. The proportion of neutrophils and
30     lymphocytes, as well as total protein level in BAL, also was significantly elevated following the
31     combined exposure when compared with all other groups. Within the centriacinar regions of the

       December 2005                          AX5-89      DRAFT-DO NOT QUOTE OR CITE

-------
 1      lungs, the percentage of proliferating cells identified by bromodeoxyuridine (BrdU) labeling was
 2      unchanged from control following exposure to ADSS alone, but was significantly elevated
 3      following exposure to O3 (280% of control), and further augmented in a statistically significant
 4      manner in mice exposed to ADSS/O3 (402% of control).  Following exposure to O3 alone or
 5      combined with ADSS, the ability of AMs to release interleukin (IL)-6 under lipopolysaccharide
 6      (LPS) stimulation was significantly decreased, while exposure to ADSS alone or ADSS/O3
 7      caused a significantly increased release of TNF-a from AMs under LPS stimulation.  The authors
 8      concluded that ADSS exposure enhances the sensitivity of animals to O3-induced lung injury.
 9           Toxicological studies with components of ETS (e.g., nicotine receptor agonists, acrolein,
10      and oxidants) have shown that the vagal bronchopulmonary C-fibers are stimulated by acute
11      exposures that initiate both central and local responses (Bonham et al., 2001; Mutoh et al., 2000).
12      The central responses (e.g., tachypnea, cough, bronchoconstriction, increased mucous secretion)
13      are more protective of the lung; however, local responses may include increased sensitization of
14      the C-fibers to other irritants, including O3.
15
16      AX5.4.3   Complex (Multicomponent) Mixtures Containing Ozone
17           Ambient pollution in most areas is a  complex mix of more than  two chemicals, and a
18      number of new studies have examined the  effects of exposure to multicomponent atmospheres
19      containing O3. Some of these studies attempted to simulate photochemical reaction products
20      occurring under actual atmospheric conditions.  However, the results  of these studies are often
21      difficult to interpret because of chemical interactions between the components, as well as the
22      resultant production of variable amounts of numerous secondary reaction products, and a lack of
23      precise control over the ultimate composition of the exposure environment.  In addition, the role
24      of O3 in the observed biological responses is often obscure.
25           A recent attempt has been made to examine multicomponent mixtures resulting from the
26      reaction of O3 with unsaturated hydrocarbons [e.g., isoprene (C5H8) and terpene (C10H16)],
27      producing HCHO, formic acid, acetone, acrolein, acetic acid, and other oxidation products, many
28      of which are strong airway irritants. Wilkins et al. (2001) evaluated sensory irritation by
29      measuring mean fB in the mouse bioassay and found a 50% reduction after 30 min of exposure to
30      reaction products of O3 and isoprene.  The mixture at this time period contained < 0.2 ppm O3, so
31      the authors attributed the observed effects  to the oxidation products.  Clausen et al. (2001), using

        December 2005                          AX5-90      DRAFT-DO NOT QUOTE OR CITE

-------
 1      the same mouse model, evaluated the reaction products of O3 and limonene. A 33% reduction in
 2      mean fB was produced after 30 min of exposure to the mixture containing < 0.3 ppm O3, again
 3      implicating the effects of strong irritant products.
 4           Pollutant mixtures containing acid aerosols comprise another type of commonly examined
 5      exposure atmosphere. Most of these mixtures included acidic sulfate aerosols as the copollutant.
 6      Peak (1-h) ambient levels of sulfuric acid (H2SO4) are estimated at 75 |ig/m3, with longer (12-h)
 7      averages about one-third of this concentration.
 8           More recent studies found some differences in airway responses to inhaled acid particle-O3
 9      mixtures that may have been partly due to airway dosimetry. Various physical and chemical
10      mechanisms may be responsible (see Schlesinger, 1995). For example, physical adsorption or
11      absorption of O3 or its reaction products on a particle could result in transport to more sensitive
12      sites, or to sites where O3, by itself, would not normally be reactive.  Chemical reactions on the
13      surface of particles can form secondary products that are more lexicologically active, or chemical
14      characteristics of the particle may change the residence time or reactivity of oxidation products at
15      the site of deposition. Chen et al. (1995) and El-Fawal et al. (1995) tested this hypothesis on
16      rabbit pulmonary macrophages and on airway reactivity, respectively. Male New Zealand white
17      rabbits were exposed for 3 h to 125 |ig/m3 H2SO4, 0.1, 0.3, or 0.6 ppm O3, and to combinations of
18      O3 and H2SO4.  Decreased pH following exposure to acid aerosol was correlated with phagocytic
19      activity and capacity of harvested macrophages; however, exposure to the mixture did not  show
20      this relationship (Chen et al., 1995). Responsiveness of harvested bronchial rings to acetylcholine
21      was increased following O3 exposure, but the combination of O3 and H2SO4 resulted in
22      antagonism (El-Fawal et al., 1995).
23           Churg et al. (1996) demonstrated increased uptake of asbestos or TiO2 into rat tracheal
24      explant cultures in response to 10 min O3 (up to 1.0 ppm) pre-exposure.  These data suggest that
25      low concentrations of O3 may increase the penetration of some types of PM into epithelial  cells.
26      More recently, Madden et al. (2000) demonstrated a greater potency for ozonized diesel PM to
27      induce prostaglandin E2 production from human epithelial cell cultures, suggesting that O3 can
28      modify the biological activity of PM derived from diesel exhaust (see details below).
29           Vincent et al. (1997) and Adamson et al. (1999) exposed rats to 0.8 ppm O3 in combination
30      with 5 or 50 mg/m3 of resuspended urban particles for 4 h. Although PM alone caused no  change
31      in cell proliferation (3H-thymidine labeling), co-exposure to either concentration of resuspended

        December 2005                          AX5-91      DRAFT-DO NOT QUOTE OR CITE

-------
 1      PM with O3 greatly potentiated the proliferative effects of exposure to O3 alone.  These
 2      interactive changes occurred in epithelial cells of the terminal bronchioles and the alveolar ducts.
 3      Kimmel et al. (1997) examined the effect of acute coexposure to O3 and fine or ultrafine H2SO4
 4      aerosols on rat lung morphology.  They determined morphometrically that alveolar septal volume
 5      was increased in animals coexposed to O3 and ultrafine, but not fine, H2SO4. Interestingly, cell
 6      labeling, an index of proliferative cell changes,  was increased only in animals co-exposed to fine
 7      H2SO4 and O3, as compared to animals exposed to O3 alone.  Last and Pinkerton (1997) found that
 8      subchronic exposure to acid aerosols (20 to 150 |ig/m3 H2SO4) had no interactive effect on the
 9      biochemical and morphometric changes produced by either intermittent or continuous O3
10      exposure (0.12 to 0.2 ppm). Thus, the interactive effects of O3 and acid aerosol coexposure in the
11      lung disappeared during the long-term exposure. Sindhu et al. (1998) observed an increase in rat
12      lung putrescine levels after repeated, combined exposures to O3 and a nitric acid vapor.
13           Other studies have examined interactions  between carbon particles and O3.  Creutzenberg
14      et al.  (1995) treated rats with a high concentration of carbon particles by intratracheal instillation
15      followed by either a 7-day or 60-day exposure to 0.5  ppm O3. The phagocytotic capacity  and
16      chemotactic migration capability of alveolar macrophages was impaired by carbon black, but was
17      stimulated by O3. Kleinman et al. (1999) examined the effects of O3 plus fine, H2SO4-coated,
18      carbon particles (MMAD = 0.26 jim) for 1 or 5 days. They found that the inflammatory response
19      with the O3-particle mixture was greater after 5  days  (4 h/day) than after day 1. This contrasted
20      with O3 exposure alone (0.4 ppm), which caused marked inflammation on acute exposure, but no
21      inflammation after 5 consecutive days of exposure.
22           Kleinman et al. (2000) examined the effects of  a mixture of elemental carbon particles, O3,
23      and ammonium bisulfate on rat lung collagen content and macrophage activity. Decreases in lung
24      collagen, and increases in macrophage respiratory burst and phagocytosis were observed relative
25      to other pollutant combinations.  Mautz et al. (2001)  used a similar mixture (i.e.,  elemental carbon
26      particles, O3, ammonium bisulfate, but with NO2 also) and exposure regimen as Kleinman et al.
27      (2000). There were decreases in pulmonary macrophage Fc-receptor binding and phagocytosis
28      and increases in acid phosphatase staining. Bronchoalveolar epithelial permeability and cell
29      proliferation were increased. Altered breathing-patterns also were observed, with some
30      adaptations occurring.
        December 2005                          AX5-92      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Bolarin et al. (1997) exposed rats to 50 or 100 |ig/m3 carbon particles in combination with
 2      ammonium bisulfate and O3. Despite 4 weeks of exposure, they observed no changes in protein
 3      concentration in lavage fluid or blood prolyl 4-hydroxylase, an enzyme involved in collagen
 4      metabolism.  Slight decreases in plasma fibronectin were present in animals exposed to the
 5      combined pollutants versus O3 alone. Thus, the potential for adverse effects in the lungs of
 6      animals challenged with a combined exposure to particles and gaseous pollutants is dependent on
 7      numerous factors, including the gaseous co-pollutant, concentration, and time.
 8           In a complex series of studies, Oberdorster et al. examined the interaction of several
 9      pulmonary oxidative stress pollutants.  Elder et al. (2000a,b) reported the results of combined
10      exposure to ultrafme carbon particles (100 |ig/m3) and O3 (1 ppm) in young and old Fischer 344
11      rats that were pretreated with aerosolized endotoxin. In old rats, exposure to carbon and O3
12      produced an interaction that resulted in a greater influx in neutrophils than that produced by either
13      agent alone.  This interaction was not seen in young rats.  Oxidant release from lavage fluid cells
14      also was assessed and the combination of endotoxin, carbon particles, and O3 produced an
15      increase in oxidant release in old rats.  This mixture produced the opposite response in the cells
16      recovered from the lungs of the young rats, indicating that the lungs of the aged animals
17      underwent greater oxidative stress in response to a complex pollutant mix of particles, O3, and a
18      biogenic agent. Johnston et al. (2000; 2002) reported the results of combined exposure to O3 (1
19      and 2.5 ppm for 4, 20, or 24 h) and low-dose endotoxin, or to O3 and endotoxin separately, in
20      newborn and adult C57BL/6J mice.  In the first study, adult (8 week old) mice showed greater
21      sensitivity to O3 than newborn (36 h old) mice on the basis of mRNAs encoding for various
22      chemokines and cytokines. In contrast, adult and newborn mice responded similarly 2 h after
23      endotoxin exposure (10 ng for 10 min), suggesting that age differences in O3-generated
24      inflammation is secondary to epithelial cell injury.  In the second study, 8 week old mice exposed
25      to O3 (1 ppm for 24 h) followed by endotoxin (37.5 EU for 10 min) showed increased
26      responsiveness over either exposure alone, on the basis of increased expression of chemokine and
27      cytokine messages and increased BAL fluid levels  of protein and PMNs.
28           Fanucchi et al. (1998) and Wagner et al. (2001a,b) examined the synergistic effect of
29      co-exposure to O3 and endotoxin on the nasal transitional epithelium of rats that also was
30      mediated, in part, by neutrophils.  Fisher 344 rats intranasally instilled with endotoxin and
31      exposed to 0.5 ppm O3, 8 h per day, for 3 days developed mucous cell metaplasia in the nasal

        December 2005                          AX5-93     DRAFT-DO NOT QUOTE OR CITE

-------
 1      transitional epithelium, an area normally devoid of mucous cells; whereas, intratracheal
 2      instillation of endotoxin (20 jig) caused mucous cell metaplasia rapidly in the respiratory
 3      epithelium of the conducting airways. A synergistic increase of intraepithelial mucosubstances
 4      and morphological evidence of mucous cell metaplasia were found in rat maxilloturbinates upon
 5      exposure to both O3 and endotoxin, compared to each pollutant alone.  A similar response was
 6      reported in OVA-sensitized Brown Norway rats exposed to 0.5 ppm O3, 8 h/day for 3 days
 7      (Wagner et al., 2002), indicating that coexposure to O3 and inflammatory biogenic substances like
 8      allergens (e.g., OVA) or bacterial  endotoxin can augment epithelial and inflammatory responses
 9      in rat nasal passages.
10           In follow-up studies, Wagner et al. (2003) reported that coexposure of rats to O3 and
11      endotoxin also enhanced epithelial and neutrophilic inflammatory responses in the pulmonary
12      airways. Fisher 344 rats were intranasally instilled with endotoxin and exposed to 1.0 ppm O3 for
13      8 h, which was repeated 24 h later. Three days after the last exposure, BALF was analyzed for
14      inflammatory cells and secreted mucosub stances (mucin SAC), and lung tissue was processed for
15      morphometric analysis.  Endotoxin instillation alone caused a dose-dependent increase in BALF
16      neutrophils that was further increased 2-fold in O3-exposed rats given 20 jig endotoxin, the
17      highest dose. Mucin glycoprotein SAC also was increased in the BALF at this dose, but not at
18      lower endotoxin doses. Ozone exposure alone did not cause mucus hypersecretion, but it did
19      potentiate mucus secretion in rats  given both 2 and 20 jig endotoxin and increased intraepithelial
20      mucosub stances 2-fold, which was further substantiated by significant increases in mucin gene
21      (rMucSAC) mRNA levels in the conducting airways.
22           The effect of O3 modifying the biological potency of PM (diesel PM and carbon black) was
23      examined by Madden et al. (2000) in rats. Reaction of NIST Standard Reference Material # 2975
24      diesel PM with 0.1 ppm O3 for 48 h increased the potency (compared to unexposed or air-exposed
25      diesel PM) to induce neutrophil influx, total protein, and LDH in lung lavage fluid in response to
26      intratracheal instillation.  Exposure of the diesel PM to high, non-ambient O3 concentration (1.0
27      ppm) attenuated the increased potency, suggesting destruction of the bioactive reaction products.
28      Unlike the diesel particles, carbon black particles exposed to 0.1 ppm O3 did not exhibit an
29      increase in biological potency, which suggested that the reaction of organic components of the
30      diesel PM with O3 were responsible for the increased potency. Reaction of particle components
31      with O3 was ascertained by chemical  determination of specific classes of organic compounds.

        December 2005                          AX5-94       DRAFT-DO NOT QUOTE OR CITE

-------
 1           Ulrich et al. (2002) investigated the effect of ambient PM from Ottawa Canada (EHC-93)
 2      on O3-induced inflammation. Male Wistar rats were exposed to 0.8 ppm O3 for 8 h and allowed
 3      to recover before intratracheal instillation of 0.5, 1.5, and 5 mg of EHC-93 in 0.3 ml of saline.
 4      The high concentrations of PM used were sufficient to induce pulmonary inflammation, which
 5      was not exacerbated by pre-exposure to O3. Rats from the combined exposure group did have
 6      higher and more persistent lung lavage protein and albumin levels, as well as increased plasma
 7      fibrinogen levels when compared to PM exposure alone.
 8           The interaction of PM and O3 was further examined in a murine model of ovalbumin
 9      (OVA)-induced  asthma.  Kobzik et al. (2001) investigated whether coexposure to inhaled,
10      concentrated ambient particles (CAPs) from Boston, MA and to O3 could exacerbate asthma-like
11      symptoms. On days 7 and 14 of life, half of the BALB/c mice used in this study were  sensitized
12      by intraperitoneal (ip) injection of OVA and then exposed to OVA aerosol on three successive
13      days to create the asthma phenotype. The other half received the ip OVA, but were exposed to  a
14      phosphate-buffered saline aerosol (controls).  The mice were further subdivided (n > 61/group)
15      and exposed for  5 h to CAPs, ranging from  63 to 1,569 |ig/m3, 0.3 ppm O3, CAPs + O3, or to
16      filtered air. Pulmonary resistance and airway responsiveness to an aerosolized MCh challenge
17      were measured after exposures.  A small, statistically significant increase in pulmonary resistance
18      and airway responsiveness, respectively, was found in both normal and asthmatic mice
19      immediately after exposure to CAPs alone and to CAPs + O3, but not to O3 alone or to filtered air.
20      By 24 h after exposure, the responses returned to baseline levels. There were no significant
21      increases in airway inflammation after any of the pollutant exposures. In this well-designed study
22      of a small-animal model of asthma, O3 and CAPs did not appear to be synergistic. In further
23      analysis of the data using specific elemental groupings of the CAPs, the acutely increased
24      pulmonary resistance was found to be associated with the AISi fraction of PM.  Thus, some
25      components of concentrated PM25 may affect airway caliber in sensitized animals, but the results
26      are difficult to extrapolate to people with asthma.
27           Six unique animal studies have examined the adverse cardiopulmonary effects of complex
28      mixtures in urban and rural environments of Italy (Gulisano  et al., 1997), Spain (Lorz and Lopez,
29      1997), and Mexico (Vanda et al., 1998; Calderon-Garciduefias et al., 2001; Moss et al., 2001).
30      Five of these studies have taken advantage of the differences in pollutant mixtures of urban and
31      rural environments to report primarily morphological changes in the nasopharynx and  lower

        December 2005                          AX5-95      DRAFT-DO NOT QUOTE OR CITE

-------
 1      respiratory tract (Gulisano et al., 1997; Lorz and Lopez, 1997; Calderon-Garciduefias et al.,
 2      200la) of lambs, pigeons, and dogs, respectively, after natural, continuous exposures to ambient
 3      pollution. Each study has provided evidence that animals living in urban air pollutants have
 4      greater pulmonary changes than that would occur in a rural and presumably cleaner, environment.
 5      The study by Moss et al. (2001) examined the nasal and lung tissue of rats exposed (23 h/day) to
 6      Mexico City air for up to 7 weeks and compared them to controls similarly exposed to filtered air.
 7      No inflammatory or epithelial lesions were found using quantitative morphological techniques;
 8      however, the concentrations of pollutants were low.
 9           The ambient air in urban areas, particularly in Southwestern Mexico City, is a complex
10      mixture of particles and gases, including high concentrations of O3 and aldehydes that previously
11      have been shown to cause airway inflammation and epithelial lesions in laboratory animals
12      (Harkema et al., 1994, 1997a,b).
13
14
15      AX5.6   EFFECTS OF  OTHER PHOTOCHEMICAL OXIDANTS
16           Complex atmospheric physical and chemical processes involving two classes of precursor
17      pollutants, volatile organic compounds and nitrogen oxides (NOX), lead to the formation of O3 and
18      other photochemical oxidants found in ambient air, such as peroxyacyl nitrates, nitric acid
19      (HNO3), and sulfuric acid, and to the formation of other compounds, such as PM, formaldehyde
20      (HCHO), and other carbonyl  compounds (see Chapter 2).  Peroxyacetyl nitrate (PAN) and
21      peroxypropionyl nitrate (PPN) are the most abundant of the non-ozone oxidants in ambient air of
22      industrialized areas, other than the inorganic nitrogenous oxidants such as NO2, and possibly
23      HNO3.
24           Tropospheric reactions  between  O3 and hydrocarbons (e.g., d-limonene) produce epoxides,
25      hydroperoxides, and peroxides. Hydrogen peroxide (H2O2) presumably constitutes the majority
26      of the measured ambient hydroperoxides (< 5 ppb), although a small amount of organic
27      hydroperoxides (ROOH) also may be formed. On the basis of equilibrium calculations and
28      limited data, Friedlander and Yeh (1998) estimated that atmospheric aerosols can carry as high as
29      1 mM of H2O2 and organic hydroperoxides (e.g., hydroxymethylhydroperoxide) in the associated
30      water. High concentrations of liquid phase H2O2 (50 jiM to 1 mM) are known to  induce in vitro
31      cell and tissue damage, and recent in vivo studies indicate that 10 and 20 ppb of inhaled H2O2

        December 2005                         AX5-96      DRAFT-DO NOT QUOTE OR CITE

-------
 1     vapor can penetrate the lower lung where it causes inflammation (Morio et al., 2001).  It is likely
 2     that hygroscopic components of PM transport ambient H2O2 into the lower lung and induce tissue
 3     injury as well. Exposure of rats to a H2O2-fine particle mixture (215 or 429 |ig/m3 ammonium
 4     sulfate) resulted in increased neutrophil influx, and production of inflammatory mediators by
 5     alveolar macrophages (Morio et al., 2001).  Hygroscopic secondary organic aerosols (SOA)
 6     generated by the O3/hydrocarbon reactions and their co-occurrence with H2O2 also provides
 7     another possible mechanism whereby H2O2 can be transported into the lower respiratory tract
 8     (e.g., Friedlander and Yeh, 1998).
 9          Therefore, acute toxicity of PAN is much lower than O3 and it is unlikely that present
10     ambient PAN levels would affect pulmonary function responses to O3 (Vyskocil et al., 1998).
11     Cytogenetic studies indicate that PAN is not a potent mutagen, clastogen, or DNA damaging
12     agent in mammalian cells in vivo or in vitro at concentrations several orders of magnitude higher
13     than the generally encountered ambient air levels in most cities (Vyskocil et al., 1998; Kligerman
14     et al., 1995; Heddle et al., 1993).  Some studies suggest that PAN may be a weak bacterial
15     mutagen at concentrations much higher than exist in present urban atmospheres (DeMarini et al.,
16     2000; Kleindienst et al., 1990).
17
       December 2005                          AX5-97     DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Adamson, I. Y. R.; Vincent, R.; Bjarnason, S. G. (1999) Cell injury and interstitial inflammation in rat lung after
  3             inhalation of ozone and urban particulates. Am. J. Respir. Cell Mol. Biol. 20: 1067-1072.
  4       Aizawa, H.; Koto, H.; Nakano, H.; Inoue, H.; Matsumoto, K.; Takata, S.; Shigyo, M.; Hara, N. (1997) The effects of
  5             a specific tachykinin receptor antagonist FK-224 on ozone-induced airway hyperresponsiveness and
  6             inflammation. Respirology 2: 261-265.
  7       Aizawa, H.; Shigyo, M.; Nakano, H.; Matsumoto, K.; Inoue, H.; Hara, N. (1999a) Effect of the Chinese herbal
  8             medicine, Bakumondo-to, on airway hyperresponsiveness induced by ozone exposure in guinea-pigs.
  9             Respirology 4: 349-354.
10       Aizawa, H.; Shigyo, M.; Matsumoto, K.; Inoue, H.; Koto, H.; Hara, N. (1999b) PACAP reverses airway
11             hyperresponsiveness induced by ozone exposure in guinea pigs. Respiration 66: 538-542.
12       Alfaro, M. F.; Putney, L.; Tarkington, B. K.; Hatch, G. E.; Hyde, D. M.; Schelegle, E. S. (2004) Effect of rapid
13             shallow breathing on the distribution of 18O-labeled ozone reaction product in the respiratory tract of the rat.
14             Inhalation Toxicol. 16: 77-85.
15       Anderson, S. D. (1996) Challenge tests to assess airway hyperresponsiveness and efficacy of drugs used in the
16             treatment of asthma. J. Aerosol Med. 9: 95-109.
17       Anderson, S. D.; Daviskas, E. (2000) The mechanism of exercise-induced asthma is ... J. Allergy Clin. Immunol. 106:
18             453-459.
19       Arito, H.; Takahashi, M.; Iwasaki, T.; Uchiyama, I. (1997) Age-related changes in ventilatory and heart rate
20             responses to acute ozone exposure in the conscious rat. Ind.  Health 35: 78-86.
21       Arsalane, K.; Gosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
22             synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
23             13:60-68.
24       Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S.  (1996) Longitudinal distribution of ozone absorption in the
25             lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
26       Atwal, O. S.; Pemsingh, R. S. (1981) Morphology of microvascular changes and endothelial regeneration in
27             experimental ozone-induced parathyroiditis: III. some pathologic considerations. Am. J. Pathol. 102:  297-307.
28       Atwal, O. S.; Wilson, T. (1974) Parathyroid gland changes following ozone inhalation: a morphologic study.
29             Arch. Environ. Health 28: 91-100.
30       Atwal, O. S.; Samagh, B. S.; Bhatnagar, M. K. (1975)  A possible autoimmune parathyroiditis following ozone
31             inhalation: II. a histopathologic, ultrastructural,  and immunofluorescent study. Am. J. Pathol. 79: 53-68.
32       Avila-Costa, M. R.; Colin-Barenque, L.; Fortoul, T. L; Machado-Salas, J. P.; Espinosa-Villanueva, J.;
33             Rugerio-Vargas, C.; Rivas-Arancibia, S. (1999) Memory deterioration in an oxidative stress model and its
34             correlation with cytological changes on rat hippocampus CA1. Neurosci. Lett. 270: 107-109.
35       Avital, A.; Springer, C.; Bar-Yishay, E.; Godfrey, S. (1995a) Adenosine, methacholine, and exercise challenges in
36             children with asthma or paediatric chronic obstructive pulmonary disease. Thorax 50: 511-516.
37       Avital, A.; Picard, E.; Uwyyed, K.; Springer, C. (1995b) Comparison of adenosine 5'-monophosphate and
3 8             methacoline for the differentiation of asthma from chronic airway diseases with the use of the auscultative
39             method in very young children. J. Pediatr. 127:  438-440.
40       Bassett, D.; Elbon-Copp, C.; Otterbein, S.; Barraclough-Mitchell, H.; DeLorme, M.; Yang, H. (2001) Inflammatory
41             cell availability affects ozone-induced lung damage. J. Toxicol. Environ. Health A 64: 547-565.
42       Becker,  S.; Quay, J.; Koren, H.  S. (1991) Effect of ozone on immunoglobulin production by humanB cells  in vitro. J.
43             Toxicol. Environ. Health 34: 353-366.
44       Bermudez, E.  (2001) Detection of poly(ADP-ribose) synthetase activity in alveolar macrophages of rats exposed to
45             nitrogen dioxide and ozone. Inhalation Toxicol. 13: 69-84.
46       Bermudez, E.; Ferng,  S. F.; Castro, C. E.; Mustafa, M. G. (1999) DNA strand breaks caused by exposure to ozone
47             and nitrogen dioxide. Environ. Res. 81: 72-80.
48       Bhalla, D. K. (1996) Alteration of alveolar macrophage chemotaxis, cell adhesion, and cell adhesion molecules
49             following ozone exposure of rats. J. Cell. Physiol. 169: 429-438.
50       Bhalla, D. K. (2002) Interactive effects of cigarette smoke and ozone in the induction of lung injury.  Toxicol. Sci. 65:
51             1-3.
52       Bhalla, D. K.; Gupta, S. K. (2000) Lung injury, inflammation, and  inflammatory stimuli in rats exposed to ozone.
53             J. Toxicol. Environ.  Health 59: 211-228.
54       Bhalla, D. K.; Hoffman, L. (1997) Time course of airway epithelial and inflammatory changes in rats exposed to
55             moderate levels of ozone. Inhalation Toxicol. 9: 829-842.
         December 2005                               AX5-98       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Bhalla, D. K.; Young, C. (1992) Effects of acute exposure to O3 on rats: sequence of epithelial and inflammatory
  2             changes in the distal airways. Inhalation Toxicol. 4: 17-31.
  3       Bhalla, D. K.; Gupta, S. K.; Reinhart, P. G. (1999) Alteration of epithelial integrity, alkaline phosphatase activity, and
  4             fibronectin expression in lungs of rats exposed to ozone. J. Toxicol. Environ. Health A 56: 329-343.
  5       Bhalla, D. K.; Reinhart, P. G.; Bai, C.; Gupta, S. K. (2002) Amelioration of ozone-induced lung injury by anti-tumor
  6             necrosis factor-a. Toxicol.  Sci. 69: 400-408.
  7       Bignami, G.; Musi, B.; Dell'Omo, G.; Laviola, G.; Alleva, E. (1994) Limited effects of ozone exposure during
  8             pregnancy on physical and neurobehavioral development of CD-I mice. Toxicol. Appl. Pharmacol.
  9             129:264-271.
10       Bimonte, H. A.; Nelson, M. E.; Granholm, A. C. (2003) Age-related deficits as working memory load increases:
11             relationships with growth factors. Neurobiol. Aging 24: 37-48.
12       Bolarin, D. M.; Bhalla, D. K.; Kleinman, M.  T. (1997) Effects of repeated exposures of geriatric rats to ozone and
13             particle-containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation
14             Toxicol. 9: 423-434.
15       Bonham, A. C.; Chen, C. Y.; Mutoh, T.; Joad, J. P. (2001) Lung C-fiber CNS reflex: role in the respiratory
16             consequences of extended environmental tobacco smoke  exposure ini young guinea pigs. Environ. Health
17             Perspect. 109(suppl. 4): 573-578.
18       Bornholdt, J.; Dybdahl, M.; Vogel, U.; Hansen, M.; Loft, S.; Wallin, H. (2002) Inhalation of ozone inducesDNA
19             strand breaks and inflammation in mice. Mutat. Res. 520: 63-71.
20       Brannan, J. D.; Koskela, H.; Anderson, S. D.; Chew, N. (1998) Responsiveness to Mannitol in asthmatic subjects
21             with exercise-  and hyperventiliation-induced asthma. Am. J. Respir. Crit. Care Med. 158: 1120-1126.
22       Bridges, J. P.; Davis,  H. W.; Demodarasamy, M.; Kuroki, Y.; Howies, G.; Hui, D. Y.; McCormack, F. X. (2000)
23             Pulmonary surfactant proteins A and D are potent endogenous inhibitors of lipid peroxidation and oxidative
24             cellular injury. J. Biol. Chem. 275: 38848-38855.
25       Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A.  (2003) Lung hyperpermeability, Clara-cell
26             secretory potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhalation Toxicol. 15:
27             1209-1230.
28       Calderon-Garciduenas, L.; Valencia-Salazar, G.; Rodriguez-Alcaraz, A.; Gambling, T. M.; Garcia, R.; Osnaya, N.;
29             Villarreal-Calderon, A.; Devlin, R. B.; Carson, J. L. (2001)  Ultrastructural  nasal pathology in children
30             chronically and sequentially exposed to air pollutants. Am. J. Respir. Cell Mol. Biol. 24: 132-138.
31       Campos-Bedolla, P.; Vargas, M. H.; Montano, L. M. (2002) Effect of acute ozone exposure on pregnant rat uterus
32             contractile responses. Reprod. Toxicol. 16: 269-273.
33       Cassee, F. R.; Feron,  V. J. (1994) Biochemical and histopathological changes in nasal epithelium of rats after 3-day
34             intermittent exposure to formaldehyde and ozone alone or in combination. Toxicol. Lett. 72: 257-268.
35       Catalano, P. J.; Rogus, J.; Ryan, L. M. (1995a) Consequences of prolonged inhalation of ozone onF344/Nrats:
36             collaborative studies. Part X: robust composite scores based on median polish analysis. Cambridge, MA:
37             Health Effects Institute; research report no. 65.
38       Catalano, P. J.; Chang, L.-Y. L.; Harkema, J. R.; Kaden, D. A.; Last, J. A.; Mellick, P. W.; Parks, W. C.; Pinkerton,
39             K. E.; Radhakrishnamurthy, B.; Ryan, L. M.; Szarek, J. L. (1995b) Consequences of prolonged inhalation of
40             ozone on F344/N rats: collaborative studies. Part XI: Integrative summary.  Cambridge, MA: Health Effects
41             Institute; research report no. 65.
42       Chang, L.-Y.; Stockstill, B. L.; Menache, M. G.; Mercer, R. R.;  Crapo, J. D. (1995) Consequences of prolonged
43             inhalation of ozone on F344/N rats: collaborative studies. Part VIII: morphometric analysis of structural
44             alterations in alveolar regions. Cambridge, MA: Health Effects Institute; pp. 3-39; research report no. 65.
45       Chang, M. M.-J.; Wu, R.; Plopper, C. G.; Hyde, D. M. (1998) IL-8 is one of the major chemokines produced by
46             monkey airway epithelium after ozone-induced injury. Am. J. Physiol. 275: L524-L532.
47       Cheek, J. M.; Buckpitt, A. R.; Li, C.; Tarkington, B. K.; Plopper, C. G. (1994) Ozone injury to alveolar epithelium in
48             vitro does not reflect loss of antioxidant defenses. Toxicol. Appl. Pharmacol. 125: 59-69.
49       Cheek, J. M.; McDonald, R. J.; Rapalyea, L.; Tarkington, B. K.; Hyde, D. M. (1995) Neutrophils enhance removal of
50             ozone-injured alveolar epithelial cells in vitro. Am. J. Physiol. 269: L527-L535.
51       Chen, L. C.; Qu, Q.; Amdur, M. O.; Schlesinger, R. B. (1995) Alteration of pulmonary macrophage intracellular pH
52             following inhalation exposure to sulfuric acid/ozone mixtures. Exp. Lung Res. 21: 113-128.
53       Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
54             contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure  in infant
55             primates. J. Appl. Physiol.  94: 819-827.
         December 2005                                AX5-99       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Cheng, P.-W.; Boat, T. F.; Shaikh, S.; Wang, O.-L.; Hu, P.-C.; Costa, D. L. (1995) Differential effects of ozone on
  2             lung epithelial lining fluid volume and protein content. Exp. Lung Res. 21: 351-365.
  3       Cho, H. Y.; Hotchkiss, J. A.; Harkema, J. R. (1999a) Inflammatory and epithelial responses during the development
  4             of ozone-induced mucous cell metaplasia in the nasal epithelium of rats. Toxicol. Sci. 51:  135-145.
  5       Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (1999b) Effects of pre-existing rhinitis on ozone-induced
  6             mucous cell metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 158: 92-102.
  7       Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (2000) Neutrophil-dependent and neutrophil-independent
  8             alterations in the nasal epithelium of ozone-exposed rats. Am. J. Respir. Crit. Care Med. 162: 629-636.
  9       Cho, H.-Y.; Zhang, L.-Y.; Kleeberger, S. R. (2001) Ozone-induced lung inflammation and hyperreactivity are
10             mediated via tumor necrosis factor-a receptors. Am.  J. Physiol. 280: L537-L546.
11       Churg, A.; Brauer, M; Keeling, B. (1996) Ozone enhances  the uptake of mineral particles by tracheobronchial
12             epithelial cells in organ culture. Am. J. Respir. Crit. Care Med. 153: 1230-1233.
13       Clausen, P. A.; Wilkins, C. K.; Wolkoff, P.; Nielsen, G.  D.  (2001) Chemical and biological evaluation of a reaction
14             mixture of R-(+)-limonene/ozone: formation of strong airway irritants. Environ. Int. 26: 511-522.
15       Cohen, M. D.; Zelikoff, J. T.; Qu, Q.; Schlesinger, R. B. (1996) Effects of ozone upon macrophage-interferon
16             interactions. Toxicology  114: 243-252.
17       Cohen, M. D.; Zelikoff, J. T.; Chen, L.-C.; Schlesinger, R. B.  (1998) Immunotoxicologic effects  of inhaled
18             chromium:  role of particle solubility and co-exposure to ozone. Toxicol. Appl. Pharmacol. 152: 30-40.
19       Cohen, M. D.; Sisco, M.; Li, Y.; Zelikoff, J. T.; Schlesinger, R. B. (2001) Ozone-induced modulation of
20             cell-mediated immune responses in the lungs. Toxicol.  Appl. Pharmacol. 171: 71-84.
21       Cohen, M. D.; Sisco, M.; Baker, K.; Li, Y.; Lawrence, D.; VanLoveren, H.; Zelikoff, J. T.; Schlesinger, R. B. (2002)
22             Effects of inhaled ozone on pulmonary immune cells critical to antibacterial responses in situ. Inhalation
23             Toxicol. 14: 599-619.
24       Colin-Barenque, L.; Avila-Costa, M. R.; Fortoul, T.; Rugerio-Vargas, C.; Machado-Salas, J. P.; Espinosa-Villanueva,
25             J.; Rivas-Arancibia, S. (1999) Morphologic alteration of the olfactory bulb after acute ozone exposure in rats.
26             Neurosci. Lett. 274: 1-4.
27       Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
28             reactive absorption-mediated cytotoxicity in vitro. Am. J. Physiol.:  10.1152/ajplung.00397.2003.
29       Cotovio, J.; Onno, L.; Justine, P.; Lamure,  S.; Catroux, P. (2001) Generation of oxidative stress in human cutaneous
30             models following in vitro ozone exposure. Toxicol. in Vitro 15: 357-362.
31       Cottet-Emard, J.-M.; Dalmaz, Y.; Pequignot, J.; Peyrin, L.;  Pequignot, J.-M. (1997) Long-term exposure to ozone
32             alters peripheral and central catecholamine activity in rats. Pfluegers Arch. 433: 744-749.
33       Creutzenberg, O.;  Bellmann, B.; Klingebiel, R.; Heinrich, U.; Muhle, H. (1995) Phagocytosis and chemotaxis of rat
34             alveolar macrophages after a combined or separate exposure to ozone and carbon black. Exp. Toxicol. Pathol.
35             47:202-206.
36       Cross, C. E.; VanDer Vliet, A.; Louie,  S.; Thiele, J. J.; Halliwell, B. (1998) Oxidative stress and antioxidants at
37             biosurfaces: plants, skin, and respiratory tract surfaces. Environ. Health Perspect. 106(suppl. 5): 1241-1251.
38       Custodio-Ramirez, V.; Paz, C. (1997) Ozone produces functional deficits in the rat visual pathway.
39             Electroencephalogr. Clin. Neurophysiol. 104: 269-273.
40       Daly, C.; Fox, K.; Henein, M. (2002) Natriuretic peptides in the diagnosis of heart disease-first amongst equals?
41             Int. J. Cardiol. 84: 107-113.
42       DeLorme, M. P.; Yang, H.; Elbon-Copp, C.; Gao, X.; Barraclough-Mitchell, H.;  Bassett, D. J. P. (2002)
43             Hyperresponsive airways correlate with lung tissue inflammatory cell changes in ozone-exposed rats. J.
44             Toxicol. Environ. Health Part A 65: 1453-1470.
45       DeMarini, D. M.;  Shelton, M. L.; Kohan, M. J.; Hudgens, E. E.; Kleindienst, T. E.; Ball, L. M.; Walsh, D.; de Boer, J.
46             G.; Lewis-Bevan, L.; Rabinowitz, J. R.; Claxton, L. D.; Lewtas, J. (2000)  Mutagenicity in lung of Big Blue(R)
47             mice and induction of tandem-base substitutions in Salmonella by the air pollutant peroxyacetyl nitrate (PAN):
48             predicted formation of intrastrand cross-links. Mutat. Res.  457: 41-55.
49       Delacourt, C.; Benoist, M. R.; Waernessyckle, S. (2001) Relationship between bronchial responsiveness and clinical
50             evolution in infants who wheeze. A four-year prospective study. Am. J. Respir. Crit. Care Med.
51             164:1382-1386.
52       Delaunois, A.; Segura, P.; Montano, L.  M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998) Comparison of
53             ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol. Appl. Pharmacol.
54             150:58-67.
55       DeH'Omo, G.; Fiore, M.; Petruzzi, S.; Alleva, E.; Bignami, G. (1995a) Neurobehavioral development of CD-I mice
56             after combined gestational and postnatal exposure to ozone. Arch. Toxicol. 69: 608-616.
         December 2005                               AX5-100      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Dell'Omo, G.; Wolfer, D.; Alleva, E.; Lipp, H.-P. (1995b) Developmental exposure to ozone induces subtle changes
  2             in swimming navigation of adult mice. Toxicol. Lett. 81: 91-99.
  3       Depuydt, P.; Joos, G. F.; Pauwels, R. A. (1999) Ambient ozone concentrations induce airway hyperresponsiveness in
  4             some rat strains. Eur. Respir. J. 14: 125-131.
  5       Dong, W.; Selgrade, M. K.; Gilmour,M. I.; Lange, R. W.; Park, P.; Luster, M. I.; Kari, F. W. (1998) Altered alveolar
  6             macrophage function in calorie-restricted rats. Am. J. Respir. Cell Mol. Biol. 19: 462-469
  7       Dorado-Martinez, C.; Parades-Carbajal, C.; Mascher, D.; Borgonio-Perez, G.; Rivas-Arancibia, S. (2001) Effects of
  8             different ozone doses on memory, motor activity and lipid peroxidation levels, in rats. Int. J. Neurosci.
  9             108: 149-161.
10       Dormans, J. A. M. A.; Boere, A. J. F.; vanLoveren, H.; Rombout, P. J. A.; Marra, M.; vanBree, L. (1996)
11             Age-related toxicity in rat lungs following acute and repeated ozone exposure. Inhalation Toxicol. 8: 903-925.
12       Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
13             time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
14       Driscoll, K. E.; Simpson, L.; Carter, J.; Hassenbein, D.; Leikauf, G. D. (1993) Ozone inhalation stimulates expression
15             of a neutrophil chemotactic protein, macrophage inflammatory protein 2. Toxicol. Appl. Pharmacol. 119:
16             306-309.
17       Dye, J. A.; Madden, M. C.; Richards, J. H.; Lehmann, J. R.; Devlin, R. B.; Costa, D. L. (1999) Ozone effects on
18             airway responsiveness, lung injury, and inflammation. Comparative rat strain and in vivo/in vitro
19             investigations. Inhalation Toxicol. 11: 1015-1040.
20       El-Fawal, H. A. N.; McGovern, T.; Schlesinger, R. B. (1995) Nonspecific bronchial responsiveness assessed in vitro
21             following acute inhalation exposure to ozone and ozone/sulfuric acid mixtures. Exp. Lung Res. 21: 129-139.
22       Elder, A. C. P.; Gelein, R.; Finkelstein J. N.; Cox, C.; Oberdorster, G. (2000a) Endotoxin priming affects the lung
23             response to ultrafine particles and ozone in young and old rats. In: Phalen, R. F., ed. Inhalation toxicology:
24             proceedings of the third colloquium on paniculate air pollution and human health (first special issue); June,
25             1999; Durham, NC. Inhalation Toxicol. 12(suppl. 1): 85-98.
26       Elder, A. C. P.; Gelein, R.; Finkelstein, J. N.; Cox, C.; Oberdorster, G. (2000b) Pulmonary inflammatory response to
27             inhaled ultrafine particles is modified by age, ozone exposure, and bacterial toxin. In: Grant, L. D., ed.
28             PM2000: paniculate matter and health. Inhalation Toxicol. 12(suppl. 4): 227-246.
29       Elsayed, N. M. (2001) Diet restriction modulates lung response and survivability of rats exposed to ozone.
30             Toxicology 159: 171-182.
31       Evans, M. J.; Fanucchi, M. V.; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.;  Schelegle, E. S.;
32             Gershwhin, L. J.; Miller, L. A.; Hyde, D. M.; Sannes, P. L.; Plopper, C. G. (2003) Atypical development of the
33             trachea! basement membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am. J. Physiol.
34             285:L931-L939.
35       Fakhrzadeh, L.; Laskin, J. D.; Laskin, D. L. (2002) Deficiency in inducible nitric oxide synthase protects mice from
36             ozone-induced lung inflammation and tissue injury. Am. J. Respir. Cell Mol. Biol. 26: 413-419.
37       Fanucchi, M. V.; Hotchkiss, J. A.; Harkema, J. R. (1998) Endotoxin potentiates ozone-induced mucous cell
38             metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 152: 1-9.
39       Fanucchi, M. V; Wong, V J.; Hinds, D.; Tarkington, B. K.; Van Winkle, L. S.; Evans, M. J.; Plopper, C. G. (2000)
40             Repeated episodes of exposure to ozone alters postnatal development of distal conducting airways in infant
41             rhesus monkeys. Am. J. Respir. Crit. Care Med. 161: A615.
42       Farman, C. A.; Pinkerton, K. E.; Rajini, P.; Witschi, H.;  Last, J. A. (1997) Evolution of lung lesions in rats exposed to
43             mixtures of ozone and nitrogen dioxide. Inhalation Toxicol. 9: 647-677'.
44       Farman, C. A.; Watkins, K.; Van Hoozen, B.; Last, J. A.; Witschi, H.; Pinkerton, K. E. (1999) Centriacinar
45             remodeling and sustained procollagen gene expression after exposure to ozone and nitrogen dioxide. Am. J.
46             Respir. Cell Mol. Biol. 20: 303-311.
47       Ferng, S.-F.; Castro, C. E.; Afifi, A. A.; Bermudez, E.; Mustafa, M. G. (1997) Ozone-induced DNA strand breaks in
48             guinea pig tracheobronchial epithelial cells. J. Toxicol. Environ. Health 51: 353-367.
49       Folkerts, G.; Busse, W. W.; Nijkamp, F. P.; Sorkness, R.; Gern, J. E. (1998) Virus-induced airway
50             hyperresponsiveness and asthma. Am. J. Respir. Crit. Care Med. 157: 1708-1720.
51       Foster, W. M.; Freed, A. N. (1999) Regional clearance of solute from peripheral airway epithelia: recovery after
52             sublobar exposure to ozone. J. Appl. Physiol. 86:  641-646.
53       Frampton, M. W.; Pryor, W. A.;  Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999) Aldehydes (nonanal and
54             hexanal) in rat and human bronchoalveolar lavage fluid after ozone exposure. Cambridge, MA: Health Effects
55             Institute; research report no. 90. Available: www.healtheffects.org/Pubs/Frampton-C.pdf [2000, February 9].
         December 2005                               AX5-101       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Friedlander, S. K.; Yeh, E. K. (1998) The submicron atmospheric aerosol as a carrier of reactive chemical species:
  2             case of peroxides. Appl. Occup. Environ. Hyg. 13: 416-420.
  3       Garssen, I; Van Bree, L.; Van Der Vliet, H.; Van Loveren, H. (1997) Ozone-induced impairment of pulmonary type
  4             IV hypersensitivity and airway hyperresponsiveness in mice. Inhalation Toxicol. 9: 581-599.
  5       Gohil, K.; Cross, C. E.; Last, J. A. (2003) Ozone-induced disruptions of lung transcriptomes. Biochem. Biophys. Res.
  6             Commun. 305: 719-728.
  7       Goldsmith, C.-A. W.; Ning, Y.-Y.; Qin,  G.; Imrich, A.; Lawrence, J.; Murthy, G. G., K.; Catalano, P. J.; Kobzik, L.
  8             (2002) Combined air pollution particle and ozone exposure increases airway responsiveness in mice. Inhalation
  9             Toxicol. 14: 325-347.
10       Gonzalez-Pina, R.; Paz, C. (1997) Brain monoamine changes in rats after short periods of ozone exposure.
11             Neurochem. Res. 22: 63-66.
12       Guerrero, A. L.; Dorado-Martinez, C.; Rodriguez, A.; Pedroza-Rios, K.; Borgonio-Perez, G.; Rivas-Arancibia, S.
13             (1999) Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. NeuroReport 10:
14             1689-1692.
15       Gulisano, M; Marceddu, S.; Barbara, A.; Pacini, A.; Buiatti, E.; Martini,  A.; Pacini, P. (1997) Damage to the
16             nasopharyngeal mucosa induced by current levels of urban air pollution: a field study in lambs. Eur. Respir. J.
17             10:567-572.
18       Gtinther, T.; Hollriegl, V; Vormann, J. (1993) Perinatal development of iron and antioxidant defence systems.
19             J. Trace Elem. Electrolytes Health Dis. 7: 47-52.
20       Gupta, S. K.; Reinhart, P. G.; Bhalla, D. K. (1998) Enhancement of fibronectin expression in rat lung by ozone and an
21             inflammatory stimulus. Am. J. Physiol. 275: L330-L335.
22       Haddad, E.-B.; Liu, S. F.; Salmon, M.; Robichaud, A.; Barnes, P. J.; Chung, K. F. (1995) Expression of inducible
23             nitric oxide  synthase mRNA in Brown Norway rats exposed to ozone: effect of dexamethasone. Eur. J.
24             Pharmacol.  Environ. Toxicol. Pharmacol. Sect. 293: 287-290.
25       Haddad, E.-B.; Salmon, M.; Koto, H.; Barnes, P. J.; Adcock, I.; Chung, K. F. (1996) Ozone induction of
26             cytokine-induced neutrophil chemoattractant (CINC) and nuclear factor-kappa b in rat lung: inhibition by
27             corticosteroids. FEES Lett. 379: 265-268.
28       Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.;  Holian, A. (1998) Potential involvement of
29             4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274: L8-L16.
30       Hanna, L. M.; Frank, R.; Scherer, P. W.  (1989) Absorption of soluble gases and vapors in the respiratory system.
31             In: Chang, H. K.; Paiva, M., eds. Respiratory physiology: an analytical approach. New York, NY:  Marcel
32             Dekker, Inc.; pp. 277-316. (Lenfant, C., ed. Lung biology in health and disease: v.  40).
33       Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W.  C. (1994) Consequences of prolonged
34             inhalation of ozone on F344/N rats: collaborative studies. Part VII: effects on the nasal mucociliary apparatus.
3 5             Cambridge, MA: Health Effects Institute; research report no. 65.
36       Harkema, J. R.; Catalano, P. J.; Hotchkiss, J. A.  (1997a) Consequences of prolonged inhalation of ozone on F344/N
37             rats: collaborative studies. Part XII. Atrophy of bone in nasal turbinates. Cambridge, MA: Health Effects
3 8             Institute; research report no. 65.
39       Harkema, J. R.; Hotchkiss, J. A.; Griffith, W. C. (1997b) Mucous cell metaplasia in rat nasal epithelium after a
40             20-month exposure to ozone: a morphometric study of epithelial differentiation. Am. J. Respir.  Cell Mol. Biol.
41             16:521-530.
42       Harkema, J. R.; Hotchkiss, J. A.; Barr, E. B.; Bennett, C. B.; Gallup, M.; Lee, J.  K.; Basbaum, C. (1999) Long-lasting
43             effects of chronic ozone exposure on rat nasal epithelium. Am. J. Respir.  Cell Mol. Biol. 20: 517-529.
44       Kara, R.; Paz, C. (1993) Effects of ozone exposure during pregnancy on ontogeny of sleep in rats. Neurosci. Lett.
45             164:67-70.
46       Hawgood, S.; Poulain, F. R. (2001) The pulmonary collectins and surfactant metabolism.  Annu. Rev. Physiol.
47             63:495-519.
48       Hawgood, S.; Ochs, M.; Jung, A.; Akiyama, J.; Allen, L.; Brown, C.; Edmondson, J.; Levitt, S.; Carlson, E.;
49             Gillespie, A. M.; Villar, A.; Epstein, C. J.; Poulain,  F. R. (2002) Sequential targeted deficiency of SP-A and -D
50             leads to progressive alveolar lipoproteinosis and emphysema. Am.  J. Physiol. 283:  L1002-L1010.
51       Heddle, J. A.; Shepson, P. B.; Gingerich, J. D.; So, K. W. (1993) Mutagenicity of peroxyacetyl nitrate (PAN) in vivo:
52             tests for somatic mutations and chromosomal aberrations. Environ. Mol. Mutagen.  21: 58-66.
53       Herbert, R. A.; Hailey, J. R.; Grumbein,  S.; Chou, B. J.; Sills, R. C.; Haseman, J. K.; Goehl, T.; Miller, R. A.;
54             Roycroft, J. H.; Boorman, G. A. (1996) Two-year and lifetime toxicity and carcinogenicity studies of ozone in
55             B6C3F1 mice. Toxicol. Pathol. 24: 539-548.
         December 2005                               AX5-102      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Highfill, J. W.; Watkinson, W. P. (1996) Ozone toxicity in the rat. II. Modeling changes due to ambient temperatures
  2             and duration. J. Appl. Physiol. 80: 1811-1818.
  3       Hoffer, E.; Baum, Y.; Tabak, A.; Frevert, C. (1999) Adhesion molecules of blood polymorphonuclear leukocytes and
  4             alveolar macrophages in rats: modulation by exposure to ozone. Hum. Exp. Toxicol. 18: 547-551.
  5       Holt, P. G.; Macaubas, C.; Stumbles, P. A.; Sly, P. D. (1999) The role of allergy in the development of asthma.
  6             Nature (London, U. K.) 402(suppl. 25): B12-B16.
  7       Hotchkiss, J. A.; Harkema, J. R.; Johnson, N. F. (1997) Kinetics of nasal epithelial cell loss and proliferation in F344
  8             rats following a single exposure to 0.5 ppm ozone. Toxicol. Appl. Pharmacol. 143: 75-82.
  9       Hotchkiss, J. A.; Hilaski, R.; Cho, H.; Regan, K.;  Spencer, P.; Slack, K.; Harkema, J. R.  (1998) Fluticasone
10             propionate attenuates ozone-induced rhinitis and mucous cell metaplasia in rat nasal airway epithelium. Am. J.
11             Respir. Cell Mol. Biol. 18: 91-99.
12       Huffman, L. J.; Judy, D. J.; Brumbaugh, K.; Frazer, D. G.; Reynolds, J. S.; McKinney, W. G.; Goldsmith, W. T.
13             (2001) Hyperthyroidism increases the risk of ozone-induced lung toxicity in rats. Toxicol. Appl. Pharmacol.
14             173: 18-26.
15       Huitron-Resendiz, S.; Custodio-Ramirez, V.; Escalante-Membrillo, C.; Gonzalez-Pina, R.; Paz, C. (1994) Sleep
16             alterations and brain regional changes of serotonin and its metabolite in rats exposed to ozone. Neurosci. Lett.
17             177: 119-122.
18       Hyde, D. M; Miller, L. A.; McDonald, R. J.; Stovall, M. Y.; Wong, V.; Pinkerton, K. E.; Wegner, C. D.; Rothlein,
19             R.; Plopper, C. G. (1999) Neutrophils enhance clearance of necrotic epithelial cells in ozone-induced lung
20             injury in rhesus monkeys. Am. J. Physiol. 277: LI 190-L1198.
21       Igarashi, A.; lijima, H.; Tamura, G.; Shirato, K. (1998) Tazanolast inhibits ozone-induced airway
22             hyperresponsiveness in guinea pigs. Am. J. Respir. Crit. Care Med.  157: 1531-1535.
23       lijima, M. K.; Kobayashi, T.; Kamada, H.; Shimojo, N. (2001) Exposure to ozone aggravates nasal allergy-like
24             symptoms in guinea pigs. Toxicol. Lett. 123: 77-85.
25       Inoue, H.; Aizawa, H.; Nakano, H.; Matsumoto, K.; Kuwano, K.; Nadel, J. A.; Kara, N. (2000) Nitric oxide synthase
26             inhibitors attenuate ozone-induced airway inflammation in guinea pigs: possible role of interleukin-8. Am. J.
27             Respir. Crit.  Care Med. 161: 249-256.
28       Ishii, Y.; Yang, H.; Sakamoto, T.; Nomura, A.; Hasegawa, S.; Hirata, F.; Bassett, D. J. P. (1997) Rat alveolar
29             macrophage cytokine production and regulation of neutrophil recruitment following acute ozone exposure.
30             Toxicol. Appl. Pharmacol. 147: 214-223.
31       Ishii, Y.; Shirato, M.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Ohtsuka, M.; Sagai, M.; Hasegawa,  S. (1998) Cloning
32             of rat eotaxin: ozone inhalation increases mRNA and protein expression in lungs of brown Norway rats. Am. J.
33             Physiol. 274: L171-L176.
34       Ishii, Y.; Hashimoto, K.; Hirano, K.; Morishima, Y.; Mochizuki, M.; Masuyama, K.; Nomura, A.; Sakamoto, T.;
35             Uchida, Y.; Sagai, M.; Sekizawa, K. (2000a) Ebselen decreases ozone-induced pulmonary inflammation in
36             rats. Lung 178: 225-234.
37       Ishii, Y.; Hirano, K.; Morishima, Y.; Masuyama, K.; Goto, Y.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Sagai, M.;
3 8             Sekizawa, K. (2000b) Early molecular and cellular events of oxidant-induced pulmonary fibrosis in rats.
39             Toxicol. Appl. Pharmacol. 167: 173-181.
40       Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
41             conscious rats. Ind. Health 36: 57-60.
42       Jang, A.-S.; Choi, I.-S.; Koh, Y.-I.; Park, C.-S.; Lee, J.-S.  (2002) The relationship between alveolar epithelial
43             proliferation and airway obstruction after ozone exposure. Allergy 57: 737-740.
44       Jimba, M.; Skornik, W. A.; Killingsworth, C. R.; Long, N. C.; Brain, J. D.; Shore, S. A. (1995) Role of C fibers in
45             physiological responses to ozone in rats. J.  Appl. Physiol. 78: 1757-1763.
46       Joad, J. P.; Kott, K. S.; Bonham, A. C. (1998) Exposing guinea pigs to ozone for 1 wk enhances responsiveness of
47             rapidly adapting receptors. J.  Appl. Physiol. 84: 1190-1197.
48       Joad, J. P.; Brie, J. M.; Weir, A. J.; Putney, L.; Hyde, D. M.; Postlewait, E. M.; Plopper, C. G. (2000) Effect of
49             respiratory pattern on ozone injury to the airways of isolated rat lungs. Toxicol. Appl. Pharmacol.  169: 26-32.
50       Johnston, C. J.; Stripp, B. R.; Reynolds, S. D.; Avissar, N. E.; Reed, C. K.; Finkelstein, J. N. (1999a) Inflammatory
51             and antioxidant gene expression in C57BL/6J mice after lethal and sublethal ozone exposures. Exp. Lung Res.
52             25:81-97.
53       Johnston, C. J.; Finkelstein, J. N.; Oberdorster, G.; Reynolds, S. D.; Stripp, B. R. (1999b) Clara cell secretory
54             protein-deficient mice differ from wild-type mice in inflammatory chemokine expression to oxygen and ozone,
55             but not to endotoxin. Exp. Lung Res.  25: 7-21.
         December 2005                              AX5-103       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Johnston, C. I; Reed, C. K.; Avissar, N. E.; Gelein, R.; Finkelstein, J. N. (2000a) Antioxidant and inflammatory
  2             response after acute nitrogen dioxide and ozone exposures in C57B1/6 mice. Inhalation Toxicol. 12: 187-203.
  3       Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2000b) Newborn mice differ from adult mice in
  4             chemokine and cytokine expression to ozone, but not to endotoxin. Inhalation Toxicol. 12: 205-224.
  5       Johnston, C. J.; Oberdorster, G.; Finkelstein, J. N. (2001) Recovery from oxidant-mediated lung injury: response of
  6             metallothionein, MIP-2, and MCP-1 to nitrogen dioxide, oxygen, and ozone exposures. Inhalation Toxicol. 13:
  7             689-702.
  8       Johnston, C. J.; Oberdorster, G.; Gelein, R.; Finkelstein, J. N. (2002) Endotoxin potentiates ozone-induced pulmonary
  9             chemokine and inflammatory responses. Exp. Lung Res. 28: 419-433.
10       Kafoury, R. M; Pryor, W. A.; Squadrito, G. L.; Salgo, M. G.; Zou, X.; Friedman, M. (1999) Induction of
11             inflammatory mediators in human airway epithelial cells by lipid ozonation products. Am. J. Respir. Crit. Care
12             Med.  160: 1934-1942.
13       Kenyon, N. J.; Van Der Vliet, A.; Schock, B.C.; Okamoto, T.; McGrew, G. M.; Last, J. A. (2002) Susceptibility to
14             ozone-induced acute lung injury in iNOS-deficient mice. Am. J. Physiol. 282: L540-L545.
15       Kim, M. Y.; Son, J. W.; Cho, M. H.; Choi, C. S.; Chae, C. H.; Lee, M. H. (2001) Oviductural carcinoma in B6C3F1
16             female mice exposed to 0.5 ppm ozone. Vet. Hum.  Toxicol. 43: 370-372.
17       Kimmel, T. A.; Chen, L. C.; Bosland, M. C.; Nadziejko, C. (1997) Influence of acid aerosol droplet size on structural
18             changes in the rat lung caused by acute exposure to sulfuric acid and ozone. Toxicol. Appl. Pharmacol. 144:
19             348-355.
20       Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y.; Longphre, M.; Harkema, J.; Jedlicka, A.; Eleff, S. M.; DiSilvestre, D.;
21             Holroyd, K. J. (1997) Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
22             Nat. Genet. 17: 475-478.
23       Kleeberger, S. R.; Reddy,  S.; Zhang, L.-Y.; Jedlicka, A. E. (2000)  Genetic susceptibility to ozone-induced lung
24             hyperpermeability:  role of toll-like receptor 4. Am. J. Respir. Cell Mol. Biol. 22: 620-627.
25       Kleeberger, S. R.; Reddy,  S. P.; Zhang, L.-Y.; Cho,  H.-Y.; Jedlicka, A. E. (2001a) Toll-like receptor 4 mediates
26             ozone-induced murine lung hyperpermeability via inducible nitric oxide synthase. Am. J. Physiol. 280:
27             L326-L333.
28       Kleeberger, S. R.; Ohtsuka, Y.; Ahang, L.-Y.; Longphre, M. (200Ib) Airway responses to chronic ozone exposure are
29             partially mediated through mast cells. J. Appl. Physiol. 90: 713-723.
30       Kleindienst, T. E.; Shepson, P. B.; Smith, D. F.; Hudgens, E. E.; Nero, C. M.; Cupitt, L. T.; Bufalini, J. J.; Claxton, L.
31             D. (1990)  Comparison of mutagenic activities of several peroxyacyl nitrates. Environ. Mol. Mutagen.
32             16:70-80.
33       Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
34             alone and  in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
35       Kleinman, M. T.; Bufalino, C.; Rasmussen, R.; Hyde, D.;  Bhalla, D. K.; Mautz, W. J. (2000) Toxicity of chemical
36             components of ambient fine paniculate matter (PM2 5) inhaled by aged rats. J. Appl. Toxicol. 20: 357-364.
37       Kligerman, A. D.; Mottus, K.; Erexson, G. L. (1995) Cytogenetic analyses of the in vitro and in vivo responses of
38             murine cells to peroxyacetyl nitrate (PAN). Mutat. Res. 341: 199-206.
39       Kobzik, L.; Goldsmith, C.-A. W.; Ning, Y. Y.; Qin, G.; Morgan, B.; Imrich, A.; Lawrence J.; Murthy, G. G. K.;
40             Catalano, P.  J. (2001) Effects of combined ozone and air pollution particle exposure in mice. Boston, MA:
41             Health Effects Institute; research report no. 106. Available: http://www.healtheffects.org/Pubs/Kobzik.pdf
42             [27 January 2003].
43       Kodavanti, U. P.; Hatch, G. E.; Starcher, B.; Giri, S. N.; Winsett, D.; Costa, D. L. (1995) Ozone-induced pulmonary
44             functional, pathological, and biochemical changes in normal and vitamin C-deficient guinea pigs. Fundam.
45             Appl. Toxicol. 24: 154-164.
46       Koike, E.; Kobayashi, T.; Nelson, D. J.; Me William, A. S.; Holt, P. G. (1998) Effect of ozone exposure on alveolar
47             macrophage-mediated immunosuppressive activity in rats. Toxicol.  Sci. 41: 217-223.
48       Koike, E.; Kobayashi, T.; Murakami, M.; McWilliam, A.  S.; Holt,  P. G. (1999) Mechanisms of ozone-induced
49             inhibitory  effect of bronchoalveolar lavage fluid on alveolar macrophage-mediated immunosuppressive
50             activity in rats. J. Leukoc. Biol. 66: 75-82.
51       Koike, E.; Kobayashi, T.;  Shimojo, N. (2001) Ozone exposure enhances expression of cell-surface molecules
52             associated with antigen-presenting activity on bronchoalveolar lavage cells in rats. Toxicol. Sci. 63: 115-124.
53       Koto, H.; Aizawa, H.; Takata, S.; Inoue, H.; Kara, N. (1995) An important role of tachykinins in ozone-induced
54             airway hyperresponsiveness. Am. J. Respir. Crit. Care Med. 151: 1763-1769.
         December 2005                              AX5-104       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Koto, H.; Salmon, M; Haddad el-B.; Huang, T.-J.; Zagorski, I; Chung, K. F. (1997) Role of cytokine-induced
  2             neutrophil chemoattractant (CINC) in ozone-induced airway inflammation and hyperresponsiveness. Am. J.
  3             Respir. Crit. Care Med. 156: 234-239.
  4       Koyama, Y.; Hayaishi, O. (1994) Modulation by prostaglandins of activity of sleep-related neurons in the
  5             preoptic/anterior hypothalamic areas in rats. Brain Res. Bull. 33: 367-372.
  6       Kudo, M.; Nishikawa, M.; Ikeda, H.; Okubo, T. (1996) Involvement of superoxide anions in ozone-induced airway
  7             hyperresponsiveness in unanesthetized guinea pigs. Environ. Toxicol. Pharmacol. 2: 25-30.
  8       Larson, S. D.; Schelegle, E. S.; Walby, W. F.; Gershwin, L. J.; Fanuccihi, M. V.; Evans, M. J.; Joad, J. P.;
  9             Tarkington, B. K.; Hyde, D. M.; Plopper, C. G. (2004) Postnatal remodeling of the neural components of the
10             epithelial-mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and
11             allergen. Toxicol. Appl. Pharmacol. 194: 211-220.
12       Laskin, D. L.; Laskin, J. D. (2001) Role of macrophages and inflammatory mediators in chemically induced toxicity.
13             Toxicology (Ireland) 160: 111-118.
14       Laskin, D. L.; Pendino,  K. J.; Punjabi, C. J.; del Valle, M. R.; Laskin, J. D. (1994) Pulmonary and hepatic effects of
15             inhaled ozone in rats. Environ. Health Perspect. 102(suppl. 10): 61-64.
16       Laskin, J. D.; Heck, D. E.; Laskin, D. L. (1996) Nitric oxide production in the lung and liver following inhalation of
17             the pulmonary irritant ozone.  In: Snyder, R.; Kocsis, J. J.; Sipes, I. G.; Kalf, G. F.; Jollow, D. J.; Greim, H.;
18             Monks, T. J.; Witmer, C. M.,  eds. Biological Reactive Intermediates V: Basic Mechanistic Research in
19             Toxicology and Human Risk Assessment: proceedings of the Fifth International  Symposium; January 1995;
20             Munich, Germany. Adv. Exp. Med. Biol. 387: 141-146.
21       Laskin, D. L.; Heck, D. E.; Laskin, J. D. (1998a) Role of inflammatory cytokines and nitric oxide in hepatic and
22             pulmonary toxicity. Toxicol. Lett. 102-103: 289-293.
23       Laskin, D. L.; Sunil, V.; Guo, Y.; Heck, D. E.; Laskin, J. D. (1998b) Increased nitric oxide synthase in the lung after
24             ozone inhalation is associated with activation of NF-KB. Environ. Health Perspect. 106(suppl. 5):  1175-1178.
25       Laskin, D. L.; Fakhrzadeh, L.; Heck, D. E.; Gerecke, D.; Laskin, J. D. (2002) Upregulation of phosphoinositide
26             3-kinase and protein kinase B in alveolar macrophages following ozone inhalation. Role of NF-KB and
27             STAT-1 in ozone-induced nitric oxide production and toxicity. Mol. Cell. Biochem. 234-235: 91-98.
28       Last, J. A.; Pinkerton, K. E. (1997) Chronic exposure of rats to ozone and sulfuric acid aerosol: biochemical and
29             structural responses. Toxicology 116: 133-146.
30       Lavnikova, N.; Prokhorova, S.; Lakhotia, A. V.; Gordon, R.; Laskin,  D. L. (1998) Distinct inflammatory responses of
31             adherent vascular lung neutrophils to  pulmonary irritants. J. Inflammation 48: 56-66
32       Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective differences
33             in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung and induction
34             in rats. Drug Metab. Dispos. 26: 396-400.
35       Lemos, M.; Lichtenfels, A. J. F. C.; Amaro, E., Jr.; Macchione, M.; Martins, M. A.; King, M.; Bohm,  G. M.; Saldiva,
36             P. H. N. (1994) Quantitative pathology of nasal passages in rats exposed to urban levels of air pollution.
37             Environ. Res. 66: 87-95.
38       Long, N. C.; Suh, J.; Morrow, J. D.;  Schiestl, R. H.; KrishnaMurthy, G.  G.; Brain, J. D.; Frei, B. (2001) Ozone
3 9             causes lipid peroxidation but little antioxidant depletion in exercising and nonexercising hamsters. J. Appl.
40             Physiol. 91:  1694-1700.
41       Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1996) Mast cells contribute to O3-induced epithelial
42             damage and proliferation in nasal and bronchial airways of mice. J. Appl. Physiol. 80: 1322-1330.
43       Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1999) Ozone-induced pulmonary inflammation and
44             epithelial proliferation are partially mediated by PAF. J. Appl.  Physiol. 86: 341-349.
45       Lorz, C.; Lopez, J. (1997) Incidence of air pollution in the pulmonary surfactant system of the pigeon  (Columbia
46             livid). Anat. Rec. 249: 206-212.
47       Lum, H.; Tyler, W. S.; Hyde, D. M.; Plopper, C. G. (1983) Morphometry of in situ and lavaged pulmonary alveolar
48             macrophages from control and ozone-exposed rats. Exp. Lung Res. 5: 61-78.
49       Madden, M.  C.; Richards, J. H.; Dailey, L. A.; Hatch, G. E.; Ohio, A. J. (2000) Effect of ozone on diesel exhaust
50             particle toxicity in rat lung. Toxicol. Appl. Pharmacol. 168: 140-148.
51       Mango, G. W.; Johnston, C. J.; Reynolds,  S. D.; Finkelstein, J. N.; Plopper, C. G.;  Stripp, B. R. (1998) Clara cell
52             secretory protein deficiency increases oxidant stress response in conducting airways. Am. J. Physiol. 275:
53             L348-L356.
54       Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Shimazu, N.; Iwasaki, H.; Ikezawa, K. (1995) Importance
55             of impairment of the airway epithelium for ozone-induced airway hyperresponsiveness in guinea pigs. Jpn. J.
56             Pharmacol. 67: 375-382.
         December 2005                              AX5-105      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Matsubara, S.; Kikkawa, H.; Kaminuma, O.; Ikezawa, K. (1997a) Angiotensin-converting enzyme inhibitors can
  2             potentiate ozone-induced airway hyperresponsiveness. Eur. J. Pharmacol. 337: 259-265.
  3       Matsubara, S.; Fushimi, K.; Kaminuma, O.; Kikkawa, H.; Ikezawa, K.; Naito, K. (1997b) Prevention of
  4             ozone-induced airway hyperresponsiveness and epithelial injury by phosphodiesterase inhibitors in guinea
  5             pigs. Environ. Toxicol. Pharmacol. 3: 201-209.
  6       Matsumoto, K.; Aizawa, H.; Inoue, H.; Koto, H.; Nakano, H.; Hara, N. (1999) Role of neutrophil elastase in
  7             ozone-induced airway responses in guinea-pigs. Eur. Respir. J.  14: 1088-1094.
  8       Mautz, W. J. (2003) Exercising animal models in inhalation toxicology: interactions with ozone and formaldehyde.
  9             Environ.  Res. 92: 14-26.
10       Mautz, W. J.; Kleinman, M. T.; Bhalla, D. K.; Phalen, R. F. (2001) Respiratory tract responses to repeated inhalation
11             of an oxidant and acid gas-particle air pollutant mixture. Toxicol. Sci. 61: 331-341.
12       McGraw, D. W.; Forbes, S. L.; Mak, J. C. W.; Witte, D. P.; Carrigan, P. E.; Leikauf, G. D.; Liggett, S. B. (2000)
13             Transgenic overexpression of beta(2)-adrenergic receptors in airway epithelial cells decreases
14             bronchoconstriction. Am. J. Physiol. 279: L379-L389.
15       McKinney, W. J.; Jaskot, R. H.; Richards, J. H.; Costa, D. L.; Dreher,  K. L. (1998) Cytokine mediation of
16             ozone-induced pulmonary adaptation. Am. J. Respir. Cell. Mol. Biol. 18:  696-705.
17       Miller, L. A.; Barnett, N. L.; Sheppard, D.; Hyde, D. M. (2001) Expression of the P6 integrin subunit is associated
18             with sites of neutrophil influx in lung epithelium. J. Histochem. Cytochem. 49:  41-48.
19       Morio, L. A.; Hooper, K. A.; Brittingham, J.; Li, T.-H.; Gordon, R. E.; Turpin, B. J.; Laskin, D. L. (2001) Tissue
20             injury following inhalation of fine paniculate matter and hydrogen perioxide is  associated with altered
21             production of inflammatory mediators  and antioxidants by  alveolar macrophages. Toxicol. Appl. Pharmacol.
22             177: 188-199.
23       Moss, O. R.; Gross, E. A.; James, R. A.; Janszen, D. B.; Ross, P. W.; Roberts, K. C.; Howard, A. M.; Harkema, J. R.;
24             Calderon-Garciduenas, L.; Morgan, K. T. (2001) Respiratory tract toxicity in rats exposed to Mexico City air.
25             Cambridge, MA: Health Effects Institute; research report no.  100. Available:
26             http://www.healtheffects.org/pubs-research.htm [15 May, 2003].
27       Mticke, W. (1996) The environment and the eye. Topics of ophthalmic toxicology. Leban. Med. J. 44: 146-150.
28       Murphy, D. J. (2002) Assessment of respiratory function in safety pharmacology. Fundam. Clin. Pharmacol.
29             16:  183-196.
30       Mutoh, T.; Joad, J. P.; Bonham, A.  C. (2000) Chronic passive cigarette smoke exposure augments bronchopulmonary
31             C-fibre inputs to  nucleus tractus solitarii neurones and reflex output in young guinea-pigs. J. Physiol. (London)
32             523:223-233.
33       Nakano, H.; Aizawa, H.; Matsumoto, K.; Fukuyama, S.; Inoue, H.; Hara, N. (2000) Cyclooxygenase-2 participates in
34             the late phase of airway hyperresponsiveness after ozone exposure in guinea pigs. Eur. J. Pharmacol.
35             403:267-275.
36       Neuhaus-Steinmetz, U.; Uffhausen, F.; Herz,  U.; Renz, H. (2000) Priming of allergic immune responses by repeated
37             ozone exposure in mice. Am. J.  Respir. Cell Mol. Biol. 23: 228-233.
38       Nichols, B. G.; Woods,  J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
39             factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci. 60:
40             356-362.
41       Nielsen, G. D.; Hougaard, K. S.; Larsen, S. T.; Hammer, M.;  Wolkoff, P.; Clausen, P.  A.; Wilkins, C. K.; Alarie, Y.
42             (1999) Acute airway effects of formaldehyde and ozone inBALB/c mice. Hum. Exp. Toxicol. 18:  400-409.
43       Nino-Cabrera, H. G.; Colin-Barenque,  L.; Avila-Costa, M. R.; Espinosa-Villanueva, J.; Fortoul, T. I.;
44             Rivas-Arancibia, S. (2002) Differences between hippocampus and cerebral cortex in aged rats in an oxidative
45             stress model. Int. J. Neurosci. 112: 373-381.
46       Noviski, N.; Brewer, J. P.; Skornik,  W. A.; Galli, S. J.; Drazen, J. M.; Martin, T. R. (1999) Mast cell activation is not
47             required for induction of airway hyperresponsiveness by ozone in mice. J. Appl. Physiol. 86: 202-210.
48       O'Connor, G. T.; Sparrow, D.;  Weiss, S. T. (1995) A prospective longitudinal study of methacholine airway
49             responsiveness as a predictor of pulmonary-function decline: the Normative Aging Study. Am. J. Respir. Crit.
50             Care Med. 152: 87-92.
51       Costing, R. S.; Van Iwaarden,  J. F.;  Van Bree, L.; Verhoef, J.; Van Golde, L. M. G.; Haagsman, H. P. (1992)
52             Exposure of surfactant protein A to ozone in vitro and in vivo impairs its interactions with alveolar cells. Am.
53             J. Physiol. 262: L63-L68.
54       Paige, R. C.; Royce, F. H.; Plopper,  C. G.; Buckpitt, A. R. (2000a) Long-term exposure to ozone increases acute
55             pulmonary centriacinar injury by 1-nitronaphthalene: I. Region-specific enzyme activity. J. Pharmacol. Exp.
56             Ther. 295: 934-941.
         December 2005                              AX5-106       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Paige, R. C.; Wong, V.; Plopper, C. G. (2000b) Long-term exposure to ozone increases acute pulmonary centriacinar
  2             injury by l-nitronaphthalene: II. Quantitative histopathology. J. Pharmacol. Exp. Ther. 295: 942-950.
  3       Palmer, L. J.; Burton, P. R.; Faux, J. A. (2000) Independent inheritance of serum immunoglobulin E concentrations
  4             and airway responsiveness. Am. J. Respir. Crit. Care Med. 161:  1836-1842.
  5       Palmer, L. J.; Rye, P. J.; Gibson, N. A. (2001) Airway responsiveness in early infancy predicts asthma, lung function,
  6             and respiratory symptoms by school age. Am. J. Respir. Crit. Care Med. 163: 37-42.
  7       Paquette, N. C.; Tankersley, C. G.; Zhang, L.-Y.; Kleeberger, S. R. (1994) Repeated subacute ozone exposure of
  8             inbred mice: airway inflammation and ventilation. Exp. Lung Res. 20:  579-594.
  9       Paz, C. (1997) Some consequences of ozone exposure on health. Arch. Med. Res. 28: 163-170.
10       Paz, C.; Bazan-Perkins, B. (1992) Sleep-wake disorganization in cats exposed to ozone. Neurosci. Lett. 140: 270-272.
11       Paz, C.; Huitron-Resendiz, S. (1996) The effects of ozone exposure on the sleep-wake cycle and serotonin contents in
12             the pons of the rat.  Neurosci. Lett. 204: 49-52.
13       Pearson, A. C.; Bhalla, D. K. (1997) Effects of ozone onmacrophage adhesion in vitro and epithelial and
14             inflammatory responses in vivo: the role of cytokines. J. Toxicol. Environ. Health 50: 143-157.
15       Peden, D. B.; Dailey, L. (1995) Modulation of mast cell functions by in vitro  ozone exposure. Am. J. Physiol.
16             268: L902-L910.
17       Pendino,  K. J.; Shuler, R.  L.; Laskin, J. D.; Laskin, D. L. (1994) Enhanced production of interleukin-1, tumor
18             necrosis factor-a, and fibronectin by  rat lung phagocytes following inhalation of a pulmonary irritant. Am. J.
19             Respir. Cell Mol. Biol. 11: 279-286.
20       Pendino,  K. J.; Meidhof, T. M.; Heck, D. E.; Laskin, J. D.; Laskin, D. L. (1995) Inhibition of macrophages with
21             gadolinium chloride abrogates ozone-induced pulmonary injury and inflammatory mediator production. Am. J.
22             Respir. Cell Mol. Biol. 13: 125-132.
23       Pendino,  K. J.; Gardner, C. R.; Shuler, R. L.; Laskin, J. D.; Durham, S. K.; Barton, D. S.; Ohnishi, S. T.; Ohnishi, T.;
24             Laskin, D. L. (1996) Inhibition of ozone-induced nitric oxide synthase expression in the lung by endotoxin.
25             Am. J. Respir. Cell Mol. Biol. 14: 516-525.
26       Petruzzi,  S.; Fiore, M.; DeH'Omo, G.; Bignami, G.; Alleva, E. (1995) Medium and long-term behavioral effects in
27             mice of extended gestational exposure to ozone. Neurotoxicol. Teratol. 17: 463-470.
28       Petruzzi,  S.; De Acetis, L.; Chiarotti, F.; Sorace, A.; Alleva, E. (1999) Limited changes in handedness and morphine
29             reactivity in CD-I mice after pre- and postnatal ozone exposure. Acta Neurobiol. Exp. 59:  115-122.
30       Pinkerton, K. E.; Menache, M. G.; Plopper,  C. G. (1995) Consequences of prolonged inhalation of ozone on F344/N
31             rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary acinus, and lung
32             antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98; research report no. 65.
33             Available from: NTIS, Springfield, VA; PB95-261996.
34       Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
3 5             ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
36             epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
37             research report no.  65.
38       Plopper, C. G.; Fanucchi,  M. V. (2000) Do urban environmental pollutants exacerbate childhood lung diseases?
39             Environ. HealthPerspect. 108: A252-A253.
40       Plopper, C. G.; Chu, F.-P.; Haselton, C. J.; Peake, J.; Wu, J.; Pinkerton, K. E. (1994) Dose-dependent tolerance to
41             ozone: I. tracheobronchial epithelial reorganization in rats after 20 months' exposure. Am. J. Pathol.
42             144:404-420.
43       Plopper, C. G.; Hatch, G. E.;  Wong, V.;  Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.; Buckpitt,
44             A. R. (1998) Relationship of inhaled  ozone concentration to acute tracheobronchial epithelial injury,
45             site-specific ozone  dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
46             19:387-399.
47       Polosa, R.; Holgate, S. T.  (1997) Adenosine bronchoprovocation: a promising marker of allergic inflammation in
48             asthma? Thorax 52: 919-923.
49       Postlethwait, E. M.; Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
50             estimated surface concentrations and lining layer effects. Am. J. Physiol. 274: L1006-L1016.
51       Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir,  A. J.; Putney, L. F.; Wong, V. J.;
52             Velsor, L. W.; Plopper, C. G. (2000)  Three-dimensional mapping of ozone-induced acute cytotoxicity in
53             tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. Cell Mol. Biol. 22: 191-199.
54       Prows, D. R.; Shertzer, H. G.; Daly, M. J.; Sidman, C. L.; Leikauf, G. D. (1997) Genetic analysis of ozone-induced
55             acute lung injury in sensitive and  resistant strains of mice. Nat. Genet.  17: 471-474.
         December 2005                               AX5-107      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Prows, D. R.; Daly, M. I; Shertzer, H. G.; Leikauf, G. D. (1999) Ozone-induced acute lung injury: genetic analysis of
  2             F2 mice generated from A/J and C57BL/6J strains. Am. J. Physiol. 277: L372-L380.
  3       Pryor, W. A.; Squadrito, G. L.; Friedman, M. (1995) A new mechanism for the toxicity of ozone. Toxicol. Lett.
  4             82/83:287-293.
  5       Pryor, W. A.; Bermudez, E.; Cueto, R.; Squadrito, G. L.  (1996) Detection of aldehydes in bronchoalveolar lavage of
  6             rats exposed to ozone. Fundam. Appl. Toxicol. 34: 148-156.
  7       Rehle, D.; Leleux, D.; Erdelyi, M.; Tittel, F.; Fraser, M.; Friedfeld, S.; et al. (2001) Ambient formaldehyde detection
  8             with a laser spectrometer based on difference-frequency generation in PPLN. Appl. Phys. B: Lasers Opt.
  9             72: 947-952.
10       Reinhart, P. G.; Gupta, S. K.; Bhalla, D. K. (1999)  Attenuation of ozone-induced lung injury by interleukin-10.
11             Toxicol. Lett. 110:35-42.
12       Rivas-Arancibia, S.; Vazquez-Sandoval, R.; Gonzalez-Kladiano, D.; Schneider-Rivas, S.; Lechuga-Guerrero, A.
13             (1998) Effects of ozone exposure in rats on memory and levels of brain and pulmonary superoxide dismutase.
14             Environ.  Res. 76:  33-39.
15       Rivas-Arancibia, S.; Dorado-Martinez, C.; Borgonio-Perez, G.; Hiriart-Urdanivia, M.; Verdugo-Diaz, L.;
16             Duran-Vazquez, A.; Colin-Baranque, L.; Avila-Costa, M. R. (2000) Effects of taurine on ozone-induced
17             memory deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ. Res. 82: 7-17.
18       Rivas-Arancibia, S.; Dorado-Martinez, C.; Colin-Barenque, L.; Kendrick, K. M.; De la Riva, C.; Guevara-Guzman,
19             R. (2003) Effect of acute ozone exposure on locomotor behavior and striatal function. Pharmacol. Biochem.
20             Behav. 74: 891-900.
21       Rivas-Manzano, P.; Paz, C. (1999) Cerebellar morphological alterations in rats induced by prenatal ozone exposure.
22             Neurosci. Lett. 276: 37-40.
23       Rose, R. C.; Richer,  S. P.; Bode, A. M.  (1998) Ocular oxidants and antioxidant protection. Proc. Soc. Exp. Biol. Med.
24             217:397-407.
25       Savov, J. D.; Whitehead, G.  S.; Wang, J.; Liao, G.; Usuka, J.; Peltz, G.; Foster, W. M.; Schwartz, D. A. (2004)
26             Ozone-induced acute pulmonary injury in inbred mouse strains. Am. J. Respir. Cell Mol. Biol. 31: 69-11.
27       Schelegle, E. S.; Alfaro, M. F.; Putney, L.; Stovall, M.; Tyler, N.; Hyde, D. M. (2001) Effect of C-fiber-mediated,
28             ozone-induced rapid shallow breathing on airway epithelial injury in rats. J. Appl. Physiol. 91: 1611-1618.
29       Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V.; Van Winkle, L. S.;  Gerriets,  J. E.; Walby, W. F.;
30             Mitchell, V; Tarkington, B. K.; Wong, V J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu, R.;
31             Evans, M. J.; Hyde, D. M.; Plopper, C. G. (2003a) Repeated episodes of ozone inhalation amplifies the effects
32             of allergen sensitization and inhalation on airway  immune and structural development in Rhesus monkeys.
33             Toxicol. Appl. Pharmacol. 191: 74-85.
34       Schelegle, E. S.; Walby,  W. F.; Alfaro,  M. F.; Wong, V. J.; Putney, L.; Stovall, M. Y.; Sterner-Kock, A.; Hyde,
35             D. M.; Plopper, C. G. (2003b) Repeated episodes of ozone inhalation attenuates airway injury/repair and
36             release of substance P, but not adaptation. Toxicol. Appl. Pharmacol. 186: 127-142.
3 7       Schlesinger, R. B. (1995) Interaction of gaseous and paniculate pollutants in the respiratory tract: mechanisms  and
38             modulators. Toxicology 105: 315-325.
39       Schlesinger, R. B.; Cohen, M. D.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache,  M. G.
40             (2002a) Ozone differentially modulates airway responsiveness in atopic versus nonatopic guinea pigs.
41             Inhalation Toxicol. 14:431-457.
42       Schlesinger, R. B.; Cohen, M.; Gordon, T.; Nadziejko, C.; Zelikoff, J.  T.; Sisco, M.; Regal,  J.  F.; Menache, M. G.
43             (2002b) Ozone-induced modulation of airway hyperresponsiveness in guinea pigs. Boston, MA: Health  Effects
44             Institute; research report no. 109.
45       Schwartz, D. A. (2002) TLR4 and LPS  hyporesponsiveness in humans. Int. J. Hyg. Environ. Health 205: 221-227.
46       Segura, P.; Montano, L. M.;  Bazan-Perkins, B.; Gustin, P.; Vargas, M. H. (1997) Ozone at high-pollution urban levels
47             causes airway hyperresponsiveness to substance P but not to other agonists. Environ. Toxicol. Pharmacol. 3:
48             91-95.
49       Sen, S.; Dulchavsky, S. A.; Dutta, S.  (1993) Effects of triiodothyronine (T3) supplementation upon ozone-induced
50             lung injury. Free Radic. Res. Commun. 18: 299-308.
51       Shore,  S. A.; Abraham, J. H.; Schwartzman, I. N.; Murthy, G. G.; Laporte, J. D. (2000) Ventilatory responses to
52             ozone are reduced in immature rats. J. Appl. Physiol. 88: 2023-2030.
53       Shore,  S. A.; Schwartzman, I. N.; Le Blanc, B.; Krishna  Murthy, G.  G.; Doerschuk, C. M. (2001) Tumor necrosis
54             factor receptor 2 contributes to ozone-induced airway hyperresponsiveness in mice. Am. J. Respir. Crit.  Care
55             Med. 164: 602-607.
         December 2005                              AX5-108       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Shore, S. A.; Johnston, R. A.; Schwartzman, I. N.; Chism, D.; Krishna Murthy, G. G. (2002) Ozone-induced airway
  2             hyperresponsiveness is reduced in immature mice. J. Appl. Physiol. 92: 1019-1028.
  3       Shore, S. A.; Rivera-Sanchez, Y. M; Schwartzman, I. N.; Johnston, R. A. (2003) Responses to ozone are increased in
  4             obese mice. J. Appl. Physiol. 95: 938-945.
  5       Sindhu, R. K.; Mautz, W. J.; Kikkawa, Y. (1998) Chronic exposure to ozone and nitric acid vapor results in increased
  6             levels of rat pulmonary putrescine. Arch. Toxicol. 72:  445-449.
  7       Slade, R.; Watkinson, W. P.; Hatch, G. E. (1997) Mouse strain differences in ozone dosimetry and body temperature
  8             changes. Am. J. Physiol. 272: L73-L77.
  9       Sommer, B.; Montano, L. M.; Chavez, J.; Gustin, P.; Vargas, M. H. (1998) Guinea pig lung resistance shows
10             circadian rhythmicity not influenced by ozone. Respir. Physiol. 113: 223-229.
11       Sommer, B.; Vargas, M. H.; Chavez, J.; Carbajal, V; Segura, P.; Montano, L. M. (2001) Differences between inhaled
12             and intravenous bronchial challenge to detect O3-induced hyperresponsiveness. J. Appl. Physiol.
13             91:2595-2601.
14       Sorace, A.; De Acetis, L.; Alleva, E.; Santucci, D. (2001) Prolonged exposure to low doses of ozone: short- and
15             long-term changes to behavioral performance in mice.  Environ. Res. 85: 122-134.
16       Spannhake, E. W. (1996) Down-regulation of canine airway mast cell function following exposure to ozone in vivo.
17             Exp. Lung Res. 22: 163-178.
18       Steerenberg, P. A.; Garssen, J.; van Bree, L.; van Loveren, H. (1996) Ozone alters T-helper cell mediated bronchial
19             hyperreactivity and resistance to bacterial infection. Exp. Toxicol. Pathol. 48: 497-499.
20       Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar to
21             nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
22       Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M. J.; Ellis, R.; Dahlback, M.; O'Byrne, P. M. (1994) Effect of
23             inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs. J.
24             Appl. Physiol. 77: 2578-2583.
25       Stevens, W. H. M.; VanderHeyden, C; Wattie, J.; Lane, C. G.; Smith, W.; O'Byrne, P. M. (1995a) Effect of a
26             leukotriene B4 receptor antagonist SC-53228 on ozone-induced airway hyperresponsiveness and inflammation
27             in dogs. Am. J. Respir. Crit.  Care Med.  152: 1443-1448.
28       Stevens, W. H. M.; Conlon, P. D.; O'Byrne, P. M. (1995b) Ozone-induced oxygen radical release from
29             bronchoalveolar lavage  cells and airway hyper-responsiveness in dogs. J. Physiol. (London) 486: 257-265.
30       Stick, S. M. (2002) Pulmonary physiology, airway responsiveness and asthma. Med. J. Aust. 177: S55-S56.
31       Stick, S. M.; Turnbull, S.; Chua, H. L.; Landau, L. L; LeSouef, P. N. (1990) Bronchial responsiveness to histamine in
32             infants and older children. Am. Rev. Respir. Dis. 142:  1143-1146.
33       Stockstill, B. L.; Chang, L.-Y.; Menache, M. G.; Mellick, P. W.; Mercer, R. R.; Crapo, J. D. (1995) Bronchiolarized
34             metaplasia and interstitial fibrosis in rat lungs chronically exposed to high ambient levels of ozone.  Toxicol.
35             Appl. Pharmacol.  134: 251-263.
36       Sun, J.; Chung, K. F. (1997) Interaction of ozone exposure with airway hyperresponsiveness and inflammation
3 7             induced by trimellitic anhydride in sensitized guinea pigs. J. Toxicol. Environ. Health 51: 77-87.
3 8       Sun, J.; Koto, H.; Chung, K. F. (1997) Interaction of ozone and allergen challenges  on bronchial responsiveness and
39             inflammation in sensitised guinea pigs. Int. Arch. Allergy Immunol. 112:  191-195.
40       Szarek, J. L.; Stewart, N. L.; Zhang, J. Z.; Webb, J. A.; Valentovic, M. A.; Catalano, P. (1995)  Contractile responses
41             and structure of small bronchi isolated from rats after 20 months' exposure to ozone. Fundam. Appl. Toxicol.
42             28: 199-208.
43       Takahashi, T.; Miura, M.; Katsumata, U.; Ichinose, M.; Kimura, K.; Inoue, H.; Takishima, T.; Shirato, K. (1993)
44             Involvement of superoxide in ozone-induced airway hyperresponsiveness in anesthetized cats. Am. Rev.
45             Respir. Dis. 148: 103-106.
46       Takahashi, M.; Kleeberger, S. R.; Croxton, T. L. (1995) Genetic control of susceptibility to ozone-induced changes in
47             mouse trachea! electrophysiology. Am. J. Physiol. 269: L6-L10.
48       Takata, S.; Aizawa, H.; Inoue,  H.; Koto, H.; Kara, N. (1995)  Ozone exposure suppresses epithelium-dependent
49             relaxation in feline airway. Lung 173: 47-56.
50       Takebayashi, T.; Abraham, J.;  Murthy, G. G. K.; Lilly, C.; Rodger, I.; Shore, S. A. (1998) Role of tachykinins in
51             airway responses to ozone in rats. J. Appl. Physiol. 85: 442-450.
52       Tankersley, C. G.; Kleeberger, S. R. (1994) Ozone-induced inflammation and altered ventilation in genetically
53             susceptible mice: a comparison of acute and subacute exposures. Toxicol Lett. 72: 279-289.
54       Tankersley, C. G.; Fitzgerald, R.  S.; Mitzner, W. A.;  Kleeberger, S. R. (1993) Hypercapnic ventilatory responses in
55             mice differentially susceptible to acute ozone exposure. J. Appl. Physiol. 75:  2613-2619.
         December 2005                              AX5-109       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Tesfaigzi, I; Hotchkiss, J. A.; Harkema, J. R. (1998) Expression of the Bcl-2 protein in nasal epithelia of F344/N rats
  2             during mucous cell metaplasia and remodeling. Am. J. Respir. Cell Mol. Biol. 18: 794-799.
  3       Thiele, J. J. (2001) Oxidative targets in the stratum corneum. A new basis for antioxidative strategies. Skin
  4             Pharmacol. Appl. SkinPhysiol.  14(suppl. 1): 87-91.
  5       Thiele, J. J.; Traber, M. G.; Podda, M; Tsang, K.; Cross, C. E.; Packer, L. (1997a) Ozone depletes tocopherols and
  6             tocotrienols topically applied to murine skin. FEES Lett. 401: 167-170.
  7       Thiele, J. J.; Podda, M.; Packer, L. (1997b) Tropospheric ozone:  an emerging environmental stress to skin. Biol.
  8             Chem. 378: 1299-1305.
  9       Thiele, J. J.; Traber, M. G.; Polefka, T. G.; Cross,  C. E.; Packer, L. (1997c) Ozone-exposure depletes vitamin E and
10             induces lipid peroxidation in murine stratum corneum. J. Invest. Dermatol. 108: 753-757.
11       Thiele, J. J.; Traber, M. G.; Tsang, K.;  Cross, C. E.; Packer, L. (1997d) In vivo exposure to ozone depletes vitamins C
12             and E and induces lipid peroxidation in epidermal layers of murine skin. Free Radical Biol. Med. 23:  385-391.
13       Tsai, J.-J.; Lin, Y.-C.; Kwan, Z.-H.; Kao, H.-L. (1998) Effects of ozone on ovalbumin sensitization in guinea pigs.
14             J. Microbiol. Immunol. Infect. 31: 225-232.
15       Tsukagoshi, H.; Haddad, E.-B.; Sun, J.; Barnes, P. J.; Chung, K.  F. (1995) Ozone-induced airway
16             hyperresponsiveness: role of superoxide anions, NEP, and BK receptors. J. Appl. Physiol. 78: 1015-1022.
17       Tyler, W. S.; Tyler, N. K.; Last, J. A.; Gillespie, M. J.; Barstow,  T. J. (1988) Comparison of daily and seasonal
18             exposures of young monkeys to ozone. Toxicology 50:  131-144.
19       Tyler, W. S.; Tyler, N. K.; Magliano, D. J.; Hinds, D. M.;  Tarkington, B.; Julian, M. D.; Hyde, D. M.; Plopper, C. G.;
20             Dungworth, D. L.  (1991) Effects of ozone inhalation on lungs of juvenile monkeys. Morphometry after a 12
21             month exposure and following a 6 month post-exposure period. In: Berglund, R. L.; Lawson, D. R.; McKee, D.
22             J., eds. Tropospheric ozone and the environment: papers from an international conference; March 1990; Los
23             Angeles, CA. Pittsburgh, PA: Air & Waste  Management Association; pp.  151-160. (A&WMA transactions
24             series no. TR-19).
25       U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
26             Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and
27             Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA;
28             PB87-142949.
29       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
3 0             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
31             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
32             www.epa.gov/ncea/ozone.htm.
33       Uhlson, C.; Harrison,  K.; Allen, C. B.; Ahmad, S.; White,  C. W.; Murphy, R. C. (2002) Oxidized phospholipids
34             derived from ozone-treated lung surfactant  extract reduce  macrophage and epithelial cell viability. Chem. Res.
35             Toxicol. 15: 896-906.
36       Ulrich, M. M. W.; Alink, G. M.; Kumarathasan, P.; Vincent, R.;  Boere, A. J.; Cassee, F. R. (2002) Health effects and
3 7             time course of paniculate matter on the cardiopulmonary system in rats with lung inflammation. J. Toxicol.
38             Environ. HealthPart A 65: 1571-1595.
39       Valacchi, G.; Bocci, V (2000) Studies on the biological effects of ozone: 11. Release of factors from human
40             endothelial cells. Mediators Inflammation 9: 271-276.
41       Valacchi, G.; Weber,  S. U.; Luu, C.; Cross, C. E.;  Packer,  L. (2000) Ozone potentiates vitamin E depletion by
42             ultraviolet radiation in the murine stratum corneum. FEES Lett. 466: 165-168.
43       Valacchi, G.; Van der Vliet, A.; Schock, B.C.; Okamoto, T.; Obermuller-Jevic, U.; Cross, C. E.; Packer, L.  (2002)
44             Ozone exposure activates oxidative stress responses in murine skin. Toxicology 179: 163-170.
45       Valacchi, G.; Pagnin,  E.; Okamoto, T.; Corbacho,  A. M.; Olano, E.; Davis, P. A.; Van der Vliet, A.; Packer, L.;
46             Cross, C. E. (2003) Induction of stress proteins and MMP-9 by 0.8 ppm of ozone in murine skin. Biochem.
47             Biophys. Res. Commun. 305: 741-746.
48       Van Bree, L.; Dormans, J. A. M. A.; Boere, A. J. F.; Rombout, P. J. A. (2001) Time study on development and repair
49             of lung injury following ozone exposure in  rats. Inhalation Toxicol. 13: 703-717.
50       Van Bree, L.; Dormans, J. A. M. A.; Koren, H. S.; Devlin, R. B.; Rombout, P. J. A. (2002) Attenuation and recovery
51             of pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhalation Toxicol.
52             14:883-900.
53       Van Hoof, I. H. J. M.; Van Bree, L.; Bast, A.  (1996) Changes in  receptor function by oxidative  stress in guinea pig
54             trachea! smooth muscle. Cent. Eur. J. Public Health 4(suppl.): 3-5.
55       Van Hoof, H. J. M.; Van Acker, F. A. A.; Voss, H.-P.; Van Bree, L.; Bast, A. (1997a) Acute exposure to ozone does
56             not influence neuroreceptor density and sensitivity in guinea pig lung. Toxicol. Lett. 90: 53-60.
         December 2005                              AX5-110      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Van Hoof, H. J. M; Voss, H.-P.; Kramer, K.; Boere, A. J. F.; Dormans, J. A. M. A.; Van Bree, L.; Bast, A. (1997b)
  2             Changes in neuroreceptor function of tracheal smooth muscle following acute ozone exposure of guinea pigs.
  3             Toxicology 120: 159-169.
  4       Vanda, B.; de Buen, N.; Jasso, R.; Valero, G.; Vargas, M. H.; Olmos, R.; Arreola, J. L.; Santillan, P.; Alonso, P.
  5             (1998) Inflammatory cells and ferruginous bodies inbronchoalveolar lavage in urban dogs. Acta Cytol.
  6             42: 939-944.
  7       Vargas, M. H.; Segura, P.; Campos, M. G.; Hong, E.; Montano, L. M. (1994) Effect of ozone exposure on
  8             antigen-induced airway hyperresponsiveness in guinea pigs. J. Toxicol. Environ. Health 42: 435-442.
  9       Vargas, M. H.; Romero, L.; Sommer, B.; Zamudio, P.; Gustin, P.; Montano, L. M. (1998) Chronic exposure to ozone
10             causes tolerance to airway hyperresponsiveness in guiea pigs: lack of SOD role. J. Appl. Physiol.
11             84:1749-1755.
12       Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994a) Ozone increases amino- and
13             carboxy-terminal atrial natriuretic factor prohormone peptides in lung, heart, and circulation. J. Biochem.
14             Toxicol.  9: 107-112.
15       Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994b) Increase in atrial natriuretic factor in
16             the lungs, heart, and circulatory system owing to ozone.  Chest 105: 1551-1554.
17       Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994c) Ozone increases atrial natriuretic
18             peptides in heart, lung and circulation of aged vs. adult animals. Gerontology (Basel) 40: 227-236.
19       Vesely, K. R.; Schelegle, E. S.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Hyde, D. M. (1999a) Breathing pattern
20             response and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am. J. Respir. Cell
21             Mol. Biol. 20: 699-709.
22       Vesely, K. R.; Hyde, D. M.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Schelegle, E. S. (1999b) Capsaicin-sensitive
23             C-fiber-mediated protective responses in ozone inhalation in rats. J. Appl. Physiol. 86: 951-962.
24       Vincent, R.;  Janzen, E. G.;  Chen, G.; Kumarathasan, P.; Haire,  D. L.; Guenette, J.; Chen, J. Z.; Bray, T. M. (1996)
25             Spin trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radical Res. 25: 475-488.
26       Vincent, R.;  Bjarnason, S. G.; Adamson, I. Y. R.; Hedgecock, C.; Kumarathasan, P.; Guenette, J.; Potvin, M.;
27             Goegan, P.; Bouthillier, L. (1997) Acute pulmonary toxicity of urban paniculate matter and ozone. Am. J.
28             Pathol. 151: 1563-1570.
29       Vyskocil, A.; Viau, C.; Lamy,  S. (1998) Peroxyacetyl nitrate: review of toxicity. Hum. Exp. Toxicol. 17: 212-220.
30       Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (200la) Effects of ozone and endotoxin coexposure on rat airway
31             epithelium: potentiation of toxicant-induced alterations. Environ. Health Perspect. 109(suppl. 4): 591-598.
32       Wagner, J. G.; VanDyken, S. J.; Hotchkiss, J. A.; Harkema, J.  R. (200Ib) Endotoxin enhancement of ozone-induced
33             mucous cell metaplasia is neutrophil-dependent in rat nasal epithelium. Toxicol. Sci. 60: 338-347.
34       Wagner, J. G.; Hotchkiss, J. A.; Harkema, J. R. (2002) Enhancement of nasal inflammatory and epithelial responses
3 5             after ozone and allergen coexposure in brown Norway rats. Toxicol.  Sci. 67: 284-294.
36       Wagner, J. G.; VanDyken, S. J.; Wierenga, J. R.; Hotchkiss, J. A.; Harkema, J. R. (2003) Ozone exposure enhances
37             endotoxin-induced mucous cell metaplasia in rat pulmonary airways. Toxicol. Sci. 74:  437-446.
38       Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
3 9             ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect. 110:
40             79-84.
41       Watkinson, W. P.; Wiester, M. J.; Highfill, J. W. (1995) Ozone toxicity in the rat. I. Effect of changes in ambient
42             temperature on extrapulmonary physiological parameters. J. Appl. Physiol. 78: 1108-1120.
43       Watkinson, W. P.; Campen, M. J.; Nolan, J. P.; Costa, D. L. (2001) Cardiovascular and systemic responses to inhaled
44             pollutants in rodents: effects of ozone and paniculate matter. Environ. Health Perspect. 109(suppl. 4): 539-546.
45       Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
46             responses to inhaled pollutants in healthy  and compromised rodents: modulation via interaction with
47             environmental factors. Environ. Res. 92: 35-47.
48       Watt, K. C.;  Plopper, C. G.; Weir, A. J.; Tarkington, B.; Buckpitt, A. R. (1998) Cytochrome P450 2E1 in rat
49             tracheobronchial airways: response to ozone  exposure. Toxicol. Appl. Pharmacol. 149: 195-202.
50       Wattiez, R.;  Noel-Georis, L; Cruyt, C.; Broeckaert, F.; Bernard, A.; Falmagne, P. (2003) Susceptibility to oxidative
51             stress: proteomic analysis of bronchoalveolar lavage from ozone-sensitive  and ozone-resistant strains of mice.
52             Proteomics 3: 658-665.
53       Weber, S. U.; Thiele, J. J.; Cross, C. E.; Packer,  L. (1999) Vitamin C, uric acid, and glutathione gradients in murine
54             stratum corneum and their susceptibility to ozone exposure. J. Invest. Dermatol. 113:1128-1132.
55       Weber, S. U.; Jothi, S.; Thiele, J. J. (2000) High-pressure liquid chromatography  analysis of ozone-induced depletion
56             of hydrophilic and lipophilic antioxidants in murine skin. Methods Enzymol. 319: 536-546.
         December 2005                               AX5-111       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Weber, S. U.; Han, N.; Packer, L. (2001) Ozone: an emerging oxidative stressor to skin. Curr. Probl. in Dermatol. 29:
  2             52-61.
  3       Weller, B. L.; Crapo, J. D.; Slot, I; Posthuma, G.; Plopper, C. G.; Pinkerton, K. E. (1997) Site- and cell-specific
  4             alteration of lung copper/zinc and manganese superoxide dismutases by chronic ozone exposure. Am. J.
  5             Respir. Cell Mol. Biol. 17: 552-560.
  6       Weller, B. L.; Witschi, H.; Pinkerton, K. E. (2000) Quantitation and localization of pulmonary manganese superoxide
  7             dismutase and tumor necrosis factor a following exposure to ozone and nitrogen dioxide. Toxicol. Sci.  54:
  8             452-461.
  9       Wells, C. A.; Ravasi, T.; Faulkner, G. J.; Carninci, P.; Okazaki, Y.; Hayashizaki, Y.; Sweet, M; Wainwright,  B. J.;
10             Hume, D. A. (2003) Genetic control of the innate immune response. BMC Immunol. 4: 5. Available:
11             http://www.biomedcentral.eom/1471-2172/4/5 [18 February, 2003]
12       Wiester, M. J.; Watkinson, W. P.; Costa, D.  L.; Crissman, K. M.; Richards, J. H.; Winsett, D. W.; Highfill, J. W.
13             (1996) Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters. J.
14             Appl. Physiol. 81: 1691-1700.
15       Wilkins, C. K.; Clausen, P. A.; Wolkoff, P.;  Larsen, S. T.; Hammer, M.; Larsen, K.; Hansen, V.; Nielsen, G. D.
16             (2001) Formation of strong irritants in mixtures of isoprene/ozone and isoprene/ozone/nitrogen dioxide.
17             Environ. Health Perspect. 109: 937-941.
18       Witschi, H.; Espiritu, L; Pinkerton, K. E.; Murphy, K.; Maronpot, R. R. (1999) Ozone carcinogenesis revisited.
19             Toxicol. Sci. 52: 162-167.
20       Wu, Z.-X.; Morton, R. F.; Lee, L.-Y. (1997) Role of tachykinins in ozone-induced airway hyperresponsiveness to
21             cigarette smoke in guinea pigs. J. Appl. Physiol. 83: 958-965.
22       Yamaguchi, S.; Nagai, H.; Tanaka, H.; Tsujimoto, M.; Tsuruoka, N. (1994) Time course study for antigen-induced
23             airway hyperreactivity and the effect of soluble IL-5 receptor. Life Sci. 54(Pharmacol. Letters): PL 471-475.
24       Yamauchi, T.; Shima, M.; Kuwaki, T.; Ando, M.; Ohmichi, M.; Fukuda, Y.; Adachi, M. (2002) Acute effects  of
25             ozone exposure on lung function in mice sensitized to ovalbumin. Toxicology (Ireland) 172: 69-78.
26       Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
27             enhances ozone-induced lung injury inB6C3Fl mice. Toxicol. Sci. 65:  99-106.
28       Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced airways
29             hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
30       Zhao, Q.; Simpson, L. G.; Driscoll, K. E.; Leikauf, G. D. (1998) Chemokine regulation of ozone-induced neutrophil
31             and monocyte inflammation. Am. J. Physiol. 274: L39-L46.
32
         December 2005                              AX5-112      DRAFT-DO NOT QUOTE OR CITE

-------
 i           6.  CONTROLLED HUMAN EXPOSURE STUDIES
 2                           OF OZONE AND RELATED
 3                        PHOTOCHEMICAL OXIDANTS
 4
 5
 6     6.1   INTRODUCTION
 7          In the previous chapter, results of ozone (O3) studies in laboratory animals and in vitro test
 8     systems were presented. The extrapolation of results from animal studies is one mechanism by
 9     which information on potential adverse human health effects from exposure to O3 is obtained.
10     More direct evidence of human health effects due to O3 exposure can be obtained through
11     controlled human exposure studies of volunteers or through field and epidemiologic studies of
12     populations exposed to ambient O3 (see Chapter 7). Controlled human exposure studies
13     typically use fixed concentrations of O3 under carefully regulated environmental conditions and
14     subject activity levels. This chapter discusses studies in which volunteers were exposed for up
15     to 8 h to between 0.08 to 0.75 ppm O3 while at rest or during varying intensities of exercise.
16          The majority of controlled human studies have investigated the effects of exposure to O3 in
17     young non-smoking healthy adults (18 to 35 years of age) performing continuous exercise (CE)
18     or intermittent exercise (IE). Varied combinations of O3 concentration, exercise routine, and
19     exposure duration have been used in these studies. The responses to ambient O3 concentrations
20     include  decreased inspiratory capacity; mild bronchoconstriction; rapid, shallow breathing
21     pattern during exercise; and symptoms of cough and pain on deep inspiration.  Reflex inhibition
22     of inspiration results in a decrease in forced vital capacity (F VC) and total lung capacity (TLC)
23     and, in combination with mild bronchoconstriction, contributes to a decrease in the forced
24     expiratory volume in  1 s (FEVj). In addition to physiological pulmonary responses and
25     respiratory symptoms, O3 has been shown to result in airway hyperresponsiveness and
26     inflammation.
27          The most salient observations from studies reviewed in the 1996 EPA Ozone Air Quality
28     Criteria Document or O3 AQCD (U.S. Environmental Protection Agency, 1996) were that:
29     (1) young healthy adults exposed to O3 concentrations > 0.08 ppm develop significant reversible,
30     transient decrements in pulmonary function if minute ventilation (VE ) or duration of exposure
31     are increased sufficiently, (2) children experience similar spirometric responses but lesser

       January 2005                             6-1       DRAFT-DO NOT QUOTE OR CITE

-------
 1      symptoms from O3 exposure relative to young adults, (3) O3-induced spirometric responses are
 2      decreased in the elderly relative to young adults, (4) there is a large degree of intersubject
 3      variability in physiologic and symptomatic responses to O3 but responses tend to be reproducible
 4      within a given individual over a period of several months, and (5) subjects exposed repeated to
 5      O3 over several days develop a tolerance to successive exposures, as demonstrated by an
 6      attenuation of responses, which is lost after about a week without exposure.
 7           There are several important limitations associated with these clinical studies: (1) the
 8      ability to study only short-term, acute effects; (2) difficulties in trying to link short-term effects
 9      with long-term consequences; (3) the use of a small number of volunteers that may not be
10      representative of the general population; and (4) the statistical limitations associated with the
11      small sample size. Sample size affects the power of a study, and having a small number of
12      samples causes a risk of Type II error, i.e., the incorrect conclusion that no difference exists
13      between treatments or groups when comparisons are not significantly different. This affects the
14      confidence in estimates of a minimum O3 concentration  at which some degree of pulmonary
15      impairment will occur in both the general population and susceptible subpopulations.  As a
16      result, the conclusions drawn from many of the studies cited in this chapter may underestimate
17      the presence of responses at low O3 concentrations and low activity levels.
18           Most of the scientific information summarized in this chapter comes from the literature
19      published since the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996).  In addition
20      to further study of physiological pulmonary responses and respiratory symptoms, much of this
21      literature has focused on mechanisms of inflammation and cellular responses to injury induced
22      by O3 inhalation. A more thorough discussion and review of this literature appears in Annex
23      AX6 of this document.  In summarizing the literature, effects are described if they  are
24      statistically  significant at a probability (p-value) of less than 0.05, otherwise trends are noted
25      as  such.
26           As spirometry typically improves in healthy young adults with exercise exposures to filter
27      air (FA), the term "O3-induced" is used herein and in the annex to designate effects that have
28      been corrected for responses during FA exposures.  For healthy adults, an O3-induced change in
29      lung function is the difference between the decrement experienced with O3 exposure and the
30      improvement observed with FA exposure. However, the distinction between an O3-induced
31      change and  a post- versus preexposure change is particularly important in individuals with

        January 2005                               6-2        DRAFT-DO NOT QUOTE OR CITE

-------
 1     respiratory disease who may experience exercise-induced decrements in pulmonary function
 2     during both FA and O3 exposures. Hence, in subjects with respiratory disease, exercise-induced
 3     responses could be mistaken for O3-induced responses in the absence of a correction for FA
 4     responses.
 5
 6
 7     6.2   PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
 8           IN HEALTHY SUBJECTS
 9     6.2.1   Introduction
10          As reviewed in the 1986 and 1996 O3 AQCD's (U.S. Environmental Protection Agency,
11     1986, 1996), 0.5 ppm is the lowest O3 concentration at which statistically significant reductions
12     in FVC  and FEVj have been reported in sedentary subjects. On average, young adults (n = 23;
13     mean age, 22 yrs) exposed at rest for 2 h to 0.5 ppm O3 had O3-induced decrements of-4% in
14     FVC and -7%  in FEVj (Folinsbee et al., 1978; Horvath et al., 1979). During exercise,
15     spirometric and symptoms responses are observed at lower O3 concentrations. For acute
16     exposures of 2  h or less to ^0.12 ppm O3, if VE is sufficiently increased by exercise, healthy
17     human subjects generally experience decreases in TLC, inspiratory capacity (1C), FVC, FEVb
18     mean forced expiratory flow from 25% to 75% of FVC (FEF25.75), and tidal volume (VT) and
19     increases in specific airways resistance (sRaw), breathing frequency (fB), and airway
20     responsiveness. These exposure also cause symptoms of cough, pain on deep inspiration,
21     shortness of breath, throat irritation, and wheezing.  With  exposures of 4- to 8-h in duration,
22     statistically significant pulmonary function and symptoms responses are observed at lower
23     O3 concentrations and lower VE than in shorter duration studies.
24
25     6.2.2   Acute Exposure for Up to 2 h
26          With heavy CE (VE = 89  L/min), an O3-induced decrement of 9.7% in FEVj has been
27     reported for healthy young adults (n = 17; age, 24 ±3 yrs) exposed for only 1 h to 0.12 ppm O3
28     (Gong et al., 1986). With moderate-to-heavy IE (15 min intervals of rest and exercise
29     [VE = 68 L/min]), McDonnell et al. (1983) reported a physiologically small, but significant,
30     O3-induced decrement of 3.4% in FEVj for young healthy adults (n = 22, age, 22 ±3 yrs)

       January 2005                             6-3        DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposed for 2 h to 0.12 ppm O3. Using the same 2 h exposure protocol, Linn et al. (1986) found
 2      no statistically significant spirometic responses at O3 concentrations of 0.16 ppm and lower.
 3      However, the subjects in the Linn et al. (1986) study were potentially exposed concurrently in
 4      Los Angeles to ambient O3 levels of between 0.12 and 0.16 ppm and were on average 3 yrs older
 5      than the subjects in the McDonnell et al. (1983) study. (The attenuating effects of increasing age
 6      and repeated O3 exposures are discussed in Sections 6.5.1 and 6.6, respectively.)  The disparities
 7      between the Linn et al. (1986) and McDonnell et al. (1983) studies demonstrate the difficulty in
 8      determining a no-effect-level for O3 based on relatively small study populations.
 9           Studies analyzing large data sets (>300 subjects) provide better predictive ability of acute
10      changes in FEVj at low levels of O3 and VE than possible via comparisons between smaller
11      studies. Such an analysis was performed by  McDonnell et al. (1997), who examined FEVj
12      responses in 485 healthy white males (18 to 36 years of age; subjects recruited from the area
13      around Chapel Hill, NC) exposed once for 2 h to O3 concentrations of up to 0.40 ppm at rest or
14      with IE. Decrements in FEVj were modeled by sigmoid-shaped curve as a function of subject
15      age, O3 concentration, VE, and duration of exposure.  Figure 6-1 illustrates the predicted
16      O3-induced decrements in FEVj for young healthy adults (20 yrs of age) exposed for up to 2 h
17      to O3 during moderate IE (VE = 30 L/min).  The responses illustrated for 0.1 ppm  in Figure 6-1
18      are approximately the same as responses predicted for an exposure to 0.3 ppm at rest. Although
19      not illustrated in the figure, the predicted FEVj decrements increase with VE . Regarding
20      applicability to the general population, the McDonnell et al. (1997) model has an apparent
21      limitation of considering only data for white males. However, two other large studies (n = 372;
22      18 to 35 yrs of age; subjects recruited from the area around Chapel Hill, NC) found no
23      significant gender nor race effects on spirometric responses to O3 exposure (Seal et al., 1993,
24      1996).
25           Ultman et al. (2004) recently reported pulmonary responses in 60 young heathy non-
26      smoking adults (32 M, 28 F) exposed to 0.25 ppm O3 for 1 h with CE at a target VE of 30 L/min.
27      Consistent with findings reported in the 1996 O3 criteria document, considerable intersubject
28      variability in FEVj decrements was reported by Ultman et al. (2004) with responses ranging
29      from -4 to 56%. One-third of the subjects had FEVj decrements of > 15% and 7% of the
30      subjects had decrements of > 40%.  It should be pointed out that the McDonnell et al. (1997)


        January 2005                              6-4        DRAFT-DO NOT QUOTE OR CITE

-------
                                                Time (h)

       Figure 6-1.  Predicted O3-induced decrements in FEVt as a function of exposure duration
                   and O3 concentration in young healthy adults (20 yrs of age) during moderate
                        •
                   IE (VE = 30 L/min). Predictions are for Model 1 coefficients in Table 3 of
                   McDonnell et al. (1997).
 1     model predicts only average responses.  In a more recent study, McDonnell et al. (1999) also
 2     reported a model predicting average symptom responses from O3 exposure. Unfortunately,
 3     neither of these papers (McDonnell et al., 1997, 1999) provide predictions of intersubject
 4     variability in response. (Section 6.4 of this Chapter discusses intersubject variability in response
 5     to O3 exposure).
 6
 7     6.2.3   Prolonged Ozone Exposures
 8          In the exposure range of 0.08 to 0.16 ppm O3, a number of studies using moderate quasi
 9     continuous exercise (QCE; 50 min exercise and 10 min rest per h) for 4 to 8 h have shown
10     significant responses under the following conditions: 0.16 ppm for 4 h with QCE at
11     VE  « 40 L/min (Folinsbee et al., 1994), 0.08  to 0.12 ppm for 6.6 h with QCE at VE « 35 to
12     40 L/min (Adams, 2002; Adams, 2003a; Folinsbee et al., 1988; Horstman et al.,  1990), and
13     0.12 ppm for 8 h of IE (30 min per h) at VE « 40 L/min (Hazucha et al., 1992). Symptoms and
14     spirometric responses increased with duration of exposure, O3 concentration, and total VE.
15     Airway resistance is only modestly affected with moderate or even heavy exercise combined
16     with O3 exposure (Folinsbee et al., 1978; McDonnell et al.,  1983; Seal et al., 1993).
       January 2005
6-5
DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.2.3.1   Effect of Exercise Ventilation Rate on FEVt Response to 6.6 h Ozone Exposure
 2          It is well established that response to O3 exposure is a function of VE in studies of 2 h or
 3      less in duration (See Section AX6.2.2).  It is reasonable to expect that response to a prolonged
 4      6.6-h O3 exposure is also function of VE, although quantitative analyzes are lacking. Data from
 5      five similar prolonged exposure studies are available for evaluation of FEVj responses as a
 6      function of exercise VE (Adams, 2000; Adams and Ollison, 1997; Folinsbee et al., 1988,  1994;
 7      Horstman et al., 1990). Each of these studies exposed similarly aged subjects (mean ages 22 to
 8      25 yrs) to 0.12 ppm O3 for 6.6 h. In total, ten sets of mean FEVj decrements were available for
 9      exercise VE ranging from 20 to 43 L/min, although no data were available for VE between 20
10      and 30 L/min (data illustrated in Figure AX6-2). As in 2 h exposure studies, FEVj decrements
11      are a function VE in prolonged 6.6-h exposure studies as demonstrated by a significant
12      correlation between these variables (Pearson, r = 0.95, p < 0.001; Spearman, r = 0.84, p < 0.01).
13
14      6.2.3.2   Exercise Ventilation Rate as a Function of Body/Lung  Size on FEVt Response
15              to 6.6 h Ozone Exposure
16          Based on the assumption that the total inhaled O3 dose (product of O3 concentration,
17      exposure duration,  and VE) is proportional to the lung size, exercise VE are typically selected to
18      be a multiple of body surface area (BSA) or FVC. Data from several recent studies do not
19      support the contention that VE should be normalized.  In an analysis of data from 485 young
20      adults, McDonnell  et al. (1997) found that any effect of BSA, height, or baseline FVC on percent
21      decrement  in FEVj was small to nonexistent.  This is consistent with Messineo and Adams
22      (1990), who compared pulmonary function responses in young adult women having small
23      (n = 14)  or large (n = 14) lung sizes (mean FVC of 3.74 and 5.11 L, respectively) and found no
24      significant group difference in FEVj decrements.  For 30 subjects exposed to 0.12 ppm O3 for
25      6.6 h, Adams (2000) also reported that FEVj responses were more closely related to VE than to
26      VE normalized to BSA. The O3 dosimetry study of Bush et al. (1996) suggested that
27      normalization of the O3 dose might more appropriately be a function of anatomic dead space.
28      Ozone penetrates deeper into the lungs  of individuals with larger conducting airway volumes,
29      however, FEVj responses in subjects exposed for 2 h to 0.25 ppm O3 do not appear to be
30      associated with O3 uptake (Ultman et al., 2004).

        January 2005                              6-6        DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.2.3.3   Comparison of 2 h IE to 6.6 h O3 Exposure Effects on Pulmonary Function
 2           Adams (2003b) examined whether prolonged 6.6-h QCE exposure to a relatively low O3
 3      concentration (0.08 ppm) and the 2-h IE exposure at a relatively high O3 concentration (0.30
 4      ppm) elicited consistent individual subject FEVj responses. Individual subject O3 exposure
 5      reproducibility was first examined via a regression plot of the postexposure FEVj response to the
 6      6.6-h chamber exposure as a function of postexposure FEVj response to the 2-h IE chamber
 7      exposure. The R2 of 0.40, although statistically significant, was substantially less than that
 8      observed in a comparison of individual FEVj response to the two 2-h IE exposures by chamber
 9      and face mask, respectively (R2 = 0.83). The Spearman rank order correlation for the chamber
10      6.6-h and chamber 2-h exposure comparison was also substantially less (0.49) than that obtained
11      for the two 2-h IE exposures (0.85). The primary reason for the greater variability in the
12      chamber 6.6-h exposure FEVj response as a function of that observed for the two 2-h IE
13      exposures is very likely related to the increased variability in response upon repeated exposure to
14      O3 concentrations lower than 0.18 ppm (R = 0.57, compared to a mean R of 0.82 at higher
15      concentrations) reported by McDonnell et al. (1985a). This rationale is supported by the lower R
16      (0.60) observed by Adams (2003b) for the FEVj responses found in 6.6 h chamber and face
17      mask exposures to 0.08 ppm O3, compared to an R of 0.91 observed for responses found at 0.30
18      ppm O3.
19
20      6.2.4  Triangular Ozone Exposures
21           To further explore the factors that determine responsiveness to O3, Hazucha et al. (1992)
22      designed a protocol to examine the effect of varying, rather than constant, O3 concentrations.
23      Subjects were exposed to an O3 level that increased linearly from 0 to 0.24 ppm for the first 4 h
24      and then decreased linearly from 0.24 to 0 ppm over the second 4 h of the 8 h exposure
25      (triangular concentration profile) and to a constant level exposure of 0.12 ppm O3 for 8 h.  While
26      total inhaled O3 doses for the constant and the triangular concentration profile were almost
27      identical, the FEVj response was dissimilar. For the constant 0.12 ppm O3 exposure, FEVj
28      declined -5% by the fifth hour and then remained at that level. With the triangular O3
29      concentration profile, there was minimal FEVj response over the first 3 h followed by a rapid
30      decrease in FEVj (-10.3%) over next 3 h. During the seventh and eighth hours, FEVj improved
       January 2005                              6-7        DRAFT-DO NOT QUOTE OR CITE

-------
 1     to -6.3% despite continued exposure to a lower O3 concentration (0.12 to 0.00 ppm, mean =
 2     0.06 ppm).
 3          More recently, Adams (2003a) used a less abrupt triangular O3 exposure profile at
 4     concentrations assumed to be typical  of outdoor ambient conditions (beginning at 0.03 ppm,
 5     increasing steadily to 0.15 ppm in the fourth hour and decreasing steadily to 0.05 ppm at 6.6 h
 6     (mean = 0.08 ppm).  Postexposure values for FEVj and symptoms were not significantly
 7     different between the 6.6 h triangular and a square-wave 0.08 ppm O3 exposure.  There was no
 8     evidence of FEVj response recovery with the triangular exposure as observed by Hazucha et al.
 9     (1992).  Rather, FEVj responses observed by Adams (2003a) for the triangular exposure seemed
10     to plateau during the last 2 h, i.e., -5.46% at 4.6 h, -6.27% at 5.6 h, and  -5.77% at 6.6 h.
11          With square-wave O3 exposures between 0.08 to 0.12 ppm, FEVj decrements may increase
12     with time of exposure (and O3 dose) or reach plateau (Horstman et al., 1990; McDonnell et al.,
13     1991). For the triangular exposures used by Hazucha et al. (1992)  and Adams (2003a), maximal
14     FEVj responses occurred 1 h to 2 h after peak O3 concentration and 1 h to 2 h before the
15     maximal O3 dose occurred (at the end of the O3 exposure).
16
17     6.2.5   Mechanisms of Pulmonary  Function Responses
18          Inhalation of O3 for several hours while physically  active elicits both subjective respiratory
19     tract symptoms and acute pathophysiologic changes.  The typical symptomatic response
20     consistently reported in studies is that of tracheobronchial airway irritation.  This is accompanied
21     by decrements in lung capacities and  volumes, bronchoconstriction, airway hyperresponsiveness,
22     airway inflammation, immune system activation, and epithelial injury. The severity of
23     symptoms and the magnitude of response depend on inhaled dose, O3 sensitivity of an individual
24     and the  extent of tolerance resulting from previous exposures.  The development of effects is
25     time- dependent during both exposure and recovery periods with considerable overlap of
26     evolving and receding effects. The time sequence, magnitude and the type of responses of this
27     complex series of events, both in terms of development and recovery, indicate that several
28     mechanisms,  activated at different time  of exposure must contribute to the overall lung function
29     response (U.S. Environmental Protection Agency, 1996).
30          Available information on recovery from O3 exposure indicates that an initial phase of
31     recovery proceeds relatively rapidly,  and some 40 to 65% of the acute spirometric and symptom

       January 2005                              6-8       DRAFT-DO NOT QUOTE OR CITE

-------
 1      response appears to occur within about 2 h (Folinsbee and Hazucha, 1989). Following a 2 h
 2      exposure to 0.4 ppm O3 with IE, Nightingale et al. (2000) observed a 13.5% decrement in FEVj.
 3      By 3 h postexposure, however, only a 2.7% FEVj decrement persisted as illustrated in
 4      Figure 6-2.  A similar postexposure recovery in FVC was also observed. Gerrity et al. (1993)
 5      suggested that for healthy young adults transient increases in mucus clearance (mediated by
 6      cholinergic receptors) due to O3 exposure may be coincident to pulmonary function responses,
 7      i.e., the transient increases in clearance and decrements in lung function return to baseline values
 8      within 2 to 3 h postexposure.  However, there is some indication that the spirometric responses,
 9      especially at higher O3 concentrations, are not fully recovered within 24 h (Folinsbee and
10      Horvath, 1986; Folinsbee et al., 1998). Small residual lung function effects are almost
11      completely resolved within 24 hours.  In  hyperresponsive individuals, the recovery takes longer,
12      as much as 48 hours, to return to baseline values.  Collectively, these observations suggest that
13      there is a rapid recovery of O3-induced spirometric responses and symptoms, which may occur
14      during resting exposure to O3 (Folinsbee  et al., 1977) or as O3 concentration is reduced during
15      exposure (Hazucha et al., 1992),  and a slower phase, which may take at least 24 h to complete
16      (Folinsbee and Hazucha, 2000).  Repeated exposure studies at higher concentrations typically
17      show that FEVj  response to O3 is enhanced on the second of several days of exposure (Table
18      AX6-8). This enhanced response suggests a residual effect of the previous exposure, about 22 h
19      earlier,  even though the preexposure spirometry may be the same as on the previous day. The
20      absence of the enhanced response with repeated exposure at lower O3 concentrations may be the
21      result of a more  complete recovery or less damage to pulmonary tissues (Folinsbee et al., 1994).
22          As the exposure to O3 progresses, airway inflammation begins to develop and the immune
23      response at both cellular and subcellular level is activated. Airway hyperreactivity develops
24      more slowly than pulmonary function effects, while neutrophilic inflammation of the airways
25      develops even more slowly and reaches the maximum 3 to 6 h postexposure. The cellular
26      responses (e.g., release of immuno-modulatory cytokines) appear to still be active as late as 20 h
27      postexposure (Torres et al.,  2000). Following cessation of exposure, the recovery in terms of
28      breathing pattern, pulmonary function and airway hyperreactivity progresses rapidly and is
29      almost complete within 4 to 6 hours in moderately responsive individuals.  More slowly
30      developing inflammatory and cellular changes may persist for up to 48 hours.
31

        January 2005                              6-9        DRAFT-DO NOT QUOTE OR CITE

-------
                                4.0-,
                            LU
                                2.0-
                                    Pre        0         2         4
                                         Time from Exposure (hours)
                       24
       Figure 6-2. Recovery of FEVt responses following a 2 h exposure to 0.4 ppm O3 with IE.
                   Immediately postexposure, FEVt was significantly (**p<0.001) decreased.
                   At 3 h postexposure, FEVt was at 97% of the preexposure value.
       Adapted from Nightingale et al. (2000).
 1     6.2.5.1   Pathophysiologic Mechanisms
 2     Breathing pattern changes
 3          Human studies consistently report that inhalation of O3 alters the breathing pattern without
 4     significantly affecting minute ventilation.  A progressive decrease in tidal volume and a
 5     "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
 6     exposure suggests a direct modulation of ventilatory control.  These changes parallel a response
 7     of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
 8     Bronchial C-fibers and rapidly adapting receptors appear to be the primary vagal afferents
 9     responsible for O3-induced changes in ventilatory rate and depth in both humans (Folinsbee and
10     Hazucha, 2000) and animals (Coleridge et al., 1993; Hazucha and Sant'Ambrogio, 1993;
11     Schelegle et al., 1993).
12          The potential modulation of breathing pattern by activation of sensory afferents located in
13     extrathoracic airways by O3 has not yet been studied in humans. Nasal  only O3 exposure of rats
14     produces changes in breathing pattern that are similar to changes observed in humans (Kleinman
15     etal., 1999).
       January 2005
6-10
DRAFT-DO NOT QUOTE OR CITE

-------
 1      Symptoms and lung function changes
 2           As discussed, in addition to changes in ventilatory control, O3 inhalation by humans will
 3      also induce a variety of symptoms, reduce vital capacity (VC) and related functional measures,
 4      and increase airway resistance.
 5           Schelegle et al. (2001) recently demonstrated that the reduction in VC due to O3 exposure
 6      is a reflex action and not a voluntary termination of inspiration as result of discomfort. They
 7      reported that O3-induced symptom responses (mediated in part by bronchial C-fibers) are
 8      substantially  reduced by inhaled topical anesthetic. However, the anesthetic had a minor and
 9      irregular effect on pulmonary function decrements and tachypnea.  Since respiratory symptom
10      response were largely abolished, these findings support reflex inhibition of VC due to
11      stimulation of both bronchial and pulmonary C-fibers.
12           The involvement of nociceptive bronchial C-fibers modulated by opioid receptors in
13      limiting maximal inspiration and eliciting subjective symptoms in humans was studied by
14      Passannante et al. (1998).  Sufentanil (an opioid agonist and analgesic) rapidly  reversed
15      O3-induced symptom responses and reduced spirometric decrements in "strong" responders. The
16      incomplete recovery in FEVj following sufentanil administration, however, suggests
17      involvement of non-opioid receptor modulated mechanisms as well. Interestingly, naloxone
18      (opioid receptor antagonist) had no significant effect on FEVj decrements in "weak" responders.
19      Plasma levels of p-endorphin (a potent pain suppressor) were not related with O3 responses.
20
21      Airway hyperreactivity
22           In addition to limitation of maximal inspiration and its effects on other spirometric
23      endpoints, activation of airway sensory afferents also plays a role in receptor-mediated
24      bronchoconstriction and an increase  in airway resistance.  Despite this common mechanism,
25      post-O3 pulmonary function changes and either early or late bronchial hyperresponsiveness
26      (BHR) to inhaled aerosolized methacholine or histamine are poorly correlated either in time or
27      magnitude. Fentanyl and indomethacin, the drugs that have been shown to attenuate O3-induced
28      lung function decrements in humans, did not prevent induction of BHR when administered to
29      guinea pigs prior to O3 exposure (Yeadon et al., 1992).  Neither does post-O3 BHR seem to be
30      related to airway baseline reactivity.  These findings imply that the mechanisms are either not
        January 2005                              6-11        DRAFT-DO NOT QUOTE OR CITE

-------
 1      related or are activated independently in time.  Animal studies (with limited support from human
 2      studies) have suggested that an early post-O3 BHR is, at least in part, vagally mediated (Freed,
 3      1996) and that stimulation of C-fibers can lead to increased responsiveness of bronchial smooth
 4      muscle independently of systemic and inflammatory changes which may be even absent (load
 5      et al.,  1996). In vitro study of isolated human bronchi have reported that O3-induced airway
 6      sensitization involves changes in smooth muscle excitation-contraction coupling (Marthan,
 7      1996). Characteristic O3-induced inflammatory airway neutrophilia which at one time was
 8      considered a leading BHR mechanism, has been found in a murine model, to be only
 9      coincidentally associated with BHR and there was no cause and effect relationship (Zhang et al.,
10      1995). However, this observation does not rule out involvement of other cells such as
11      eosinophils or T-helper cells in BHR modulation. There is some evidence that release of
12      inflammatory mediators by these cells can sustain BHR and bronchoconstriction. In vitro and
13      animal studies have also suggested that airway neutral endopeptidase activity can be a strong
14      modulator of BHR (Marthan et al., 1996; Yeadon et al., 1992). Late BHR observed in some
15      studies is plausibly due to a sustained damage of airway epithelium and continual release of
16      inflammatory mediators (Foster et al., 2000).  Thus, O3-induced BHR appears to be a product of
17      many mechanisms acting at different time periods and levels of the bronchial smooth muscle
18      signaling pathways (The effects ofO3 on BHR are described in Section 6.8}.
19
20      6.2.5.2   Mechanisms at a Cellular and Molecular Level
21           Stimulation of vagal  afferents by O3 and reactive products, the primary mechanism of lung
22      function impairment is enhanced and sustained by what can be considered in this context to be
23      secondary mechanisms activated at a cellular and molecular level.  The complexity of these
24      mechanisms is beyond the scope of this section and the reader is directed to Section 6.9 of this
25      chapter for greater detail. A comprehensive review by Mudway  and Kelly (2000) discusses the
26      cellular and molecular mechanisms of O3-induced pulmonary response in great detail.
27           Stimulation of bronchial C-fibers by O3 not only inhibits maximal inspiration but, through
28      local axon reflexes, induces neurogenic inflammation. This pathophysiologic process is
29      characterized by release of tachykinins and other proinflammatory neuropeptides.  Ozone
30      exposure has been shown to elevate C-fiber-associated tachykinin substance P in human
31      bronchial lavage fluid (Hazbun et al. 1993) and to deplete neuropeptides synthesized and

        January 2005                              6-12        DRAFT-DO NOT QUOTE OR CITE

-------
 1      released from C-fibers in human airway epithelium rich in substance P-immunoreactive axons.
 2      Substance P and other transmitters are known to induce granulocyte adhesion and subsequent
 3      transposition into the airways, increase vascular permeability and plasma protein extravasation,
 4      cause bronchoconstriction, and promote mucus secretion (Solway and Leff, 1991).  Although the
 5      initial pathways of neurogenic, antigen-induced, and generally immune-mediated inflammation
 6      are not the same, they eventually converge leading to further amplification of airway
 7      inflammatory processes by subsequent release of cytokines, eicosanoids, and other mediators.
 8      Significantly negative correlations between O3-induced leukotriene (LTC4/D4/E4) production and
 9      spirometric decrements (Hazucha et al., 1996), and an increased level of postexposure PGE2, a
10      mediator known to stimulate bronchial C-fibers, show that these mediators play an important
11      role in attenuation of lung function due to O3 exposure (Mohammed et al., 1993; Hazucha et al.,
12      1996). Moreover, because the density of bronchial C-fibers is much lower in the small than
13      large airways, the reported post O3 dysfunction of small airways assessed by decrement in
14      FEF25.75 (Weinman et al., 1995; Frank et al., 2001) may be due in part to inflammation.  Also,
15      because of the relative slowness of inflammatory responses as compared to reflex effects, O3-
16      triggered inflammatory mechanisms are unlikely to initially contribute to progressive lung
17      function reduction.  It is plausible, however, that when fully activated, they sustain and possibly
18      further aggravate already impaired lung function.  Indeed, a prolonged recovery of residual
19      spirometric decrements following the initial rapid improvement after exposure termination could
20      be due to slowly resolving airway inflammation. Bronchial biopsies performed 6 h postexposure
21      have shown that O3 caused a significant decrease in immunoreactivity to substance P in the
22      submucosa (Krishna et al., 1997). A strong negative correlation with FEVj also suggests that the
23      release of substance P may be a contributing mechanism to persistent post-O3
24      bronchoconstriction (Krishna et al., 1997). Persistent spirometry changes observed for up to
25      48 h postexposure could plausibly be  sustained by the inflammatory mediators, many of which
26      have bronchoconstrictive properties (Blomberg et al., 1999).
27
28
        January 2005                              6-13        DRAFT-DO NOT QUOTE OR CITE

-------
 1     6.3   SUBJECTS WITH PREEXISTING DISEASE
 2          Individuals with respiratory disease are of primary concern in evaluating the health effects
 3     of O3 because even a small change in function is likely to have more impact on a person with
 4     reduced reserve, i.e., O3-induced effects are superimposed on preexisting pulmonary impairment.
 5
 6     6.3.1   Subjects with Chronic Obstructive Pulmonary Disease
 7          For patients with COPD performing light to moderate IE, no decrements in pulmonary
 8     function were observed after  1- and 2-h exposures to < 0.30 ppm O3 (Kehrl et al., 1985; Linn
 9     et al., 1982a, 1983a; Solic et al., 1982) and only small decreases in forced expiratory volume
10     were observed for 3-h exposures of chronic bronchitics to 0.41 ppm O3 (Kulle et al., 1984).
11     More recently, Gong et al. (1997a) found no significant difference in response between age-
12     matched controls and COPD patients to a 4 h exposure to 0.24 ppm O3 with IE.  Although the
13     clinical significance is uncertain, small transient decreases in arterial blood oxygen saturation
14     have also been observed in some of these studies.
15
16     6.3.2   Subjects with Asthma
17          Based on studies reviewed in the 1996 criteria document (U.S. Environmental Protection
18     Agency, 1996), asthmatics appear to be at least as sensitive to acute effects of O3 as healthy
19     nonasthmatic subjects.
20          Several recent studies support a tendency for slightly increased spirometric responses in
21     mild asthmatics versus healthy subjects. Alexis et al. (2000) reported reductions in FVC (12%,
22     10%) and FEVj (13%, 11%) for 13 mild asthmatic and 9 healthy subjects, respectively, exposed
23     to 0.4 ppm  O3 for 2 h with IE ( VE = 30 L/min).  The FVC and FEVj responses were attenuated
24     by indomethacin in the healthy subjects but not the asthmatics. As assessed by the magnitude of
25     reductions in mid-flows (viz.  FEF25, FEF50, FEF60p, FEF75) following O3 exposure,  the small
26     airways tended to be more affected in asthmatics than healthy subjects.  In a larger study, Torres
27     et al. (1996) exposed 24 asthmatics, 12 allergic rhinitics, and 10 healthy subjects to 0.25 ppm O3
28     for 3 h with IE. The O3-induced FEVj decrements tended to be greater in the  diseased
29     populations (allergic rhinitis,  14.1%; asthmatics, 12.5%; healthy controls, 10.2%).
30     Scannell et al. (1996) exposed 18 asthmatics to 0.2 ppm O3 for 4 h with IE (VE -25 L/min/m2


       January 2005                             6-14       DRAFT-DO NOT QUOTE OR CITE

-------
 1     BSA). An O3-induced increase in sRaw tended to be greater in the asthmatics compared to 81
 2     healthy subjects who underwent similar experimental protocols (Aris et al., 1995; Balmes et al.,
 3     1996).
 4           Similar O3-induced spirometric responses are suggested by some studies. The Scannell
 5     et al. (1996) study of 18 asthmatics reported FEVj and FVC decrements that were similar to 81
 6     healthy subjects (Aris et al., 1995; Balmes et al., 1996).  Similar group decrements in FEVj and
 7     FVC were reported by Hiltermann et al. (1995), who exposed 6 asthmatics and 6 healthy
 8     subjects to 0.4 ppm O3 for 2 h with light IE. Basha et al. (1994) also reported similar spirometric
 9     responses between 5 asthmatic and 5 healthy subjects exposed to 0.2 ppm O3 for 6 h with IE.
10     The lack of significant differences in the Hilltermann et al. (1995) and Basha et al. (1994)
11     studies is not compelling given the extremely small sample sizes and corresponding lack of
12     statistical power.  The Basha et al. (1994) study was also confounded by the asthmatics having
13     an average preexposure FEVj that was about 430 mL lower (a 12% difference) on the O3-day
14     relative to the air-day. Hence, only the Scannell et al. (1996) study supports similar O3-induced
15     spirometric responses in asthmatics versus healthy subjects.
16           One study has reported that asthmatics tend to have smaller O3-induced FEVj decrements
17     relative healthy subjects (3% versus 8%, respectively) when exposed to 0.2 ppm  O3 for 2 h with
18     IE (Mudway et al., 2001). However, the asthmatics in the Mudway et al. (2001)  study also
19     tended to be older than the healthy subjects, which could partially  explain their lesser response.
20           In a longer exposure duration (7.6 h) study, Horstman et al. (1995) exposed 17 mild-to-
21     moderate asthmatics and 13 healthy controls to 0.16 ppm O3 or FA with QCE (VE «30 L/min).
22     The FEVj decrement observed in the asthmatics was significantly  greater than in the healthy
23     subjects (19% versus 10%, respectively). There was also tendency for a greater O3-induced
24     decrease in FEF25_75 in asthmatics relative to the healthy subjects (24% versus 15%,
25     respectively). A significant positive correlation in asthmatics was also reported between O3-
26     induced spirometric responses and baseline lung function, i.e., responses increased with severity
27     of disease.
28           With repeated O3 exposures asthmatics, like healthy subjects (see Section 6.6}  develop
29     tolerance. Gong et al. (1997b) exposed 10 asthmatics to 0.4 ppm O3, 3 h per day with IE
30     (VE -32 L/min), for 5 consecutive days. Symptom and spirometric responses were greatest on
31     the first (-35 % FEVj) and second (-34 % FEVj) exposure days, and progressively diminished

       January 2005                             6-15        DRAFT-DO NOT QUOTE OR CITE

-------
 1      toward baseline levels (-6 % FEVj) by the fifth exposure day. Similar to healthy subjects,
 2      asthmatics lost their tolerance 4 and 7 days later.
 3           Other studies have reported that asthmatics have a somewhat exaggerated inflammatory
 4      response to acute O3 exposure relative to healthy controls (e.g., McBride et al.,  1994; Basha
 5      et al., 1994; Peden et al.,  1995, 1997; Peden, 2001a; Scannell et al., 1996; Hiltermann et al.,
 6      1997, 1999; Michelson et al., 1999; Vagaggini et al., 1999; Newson et al., 2000; Holz et al.,
 7      2002) also (see Section 6.9). Inflammatory responses do not appear to be correlated with lung
 8      function responses in either asthmatic or healthy subjects (Balmes et al., 1996, 1997; Holz et al.,
 9      1999). This lack of correlations between inflammatory and spirometric responses may be due to
10      differences in the time kinetics of these responses (Stenfors et al., 2002). In addition, airway
11      responsiveness to inhaled allergens is increased by O3 exposure in subjects with allergic asthma
12      for up to 24 h (see Section 6.8).
13
14      6.3.3  Subjects with Allergic Rhinitis
15           Allergic rhinitis is a condition defined by inflammation of the nasal membranes. Nayak
16      (2003) recently reviewed the commonalities between asthma and allergic rhinitis. Clinically,
17      greater than 60% of asthmatics have allergic rhinitis and slightly less than 40% of allergic
18      rhinitics have asthma. Leukotrienes and histamine are well-recognized mediators of responses
19      (viz., inflammation, hyperresponsiveness, and bronchoconstriction) in both asthma and allergic
20      rhinitis. Although, rhinitis and asthma are distinguished as affecting the upper and lower
21      airways, respectively, it has been suggested that these diseases are manifestations of the same
22      disease entity.
23           Given the prevalence of concomitant asthma and rhinitis and their common response
24      mediators, it should be expected that allergic rhinitics might respond more similarly to
25      asthmatics than healthy individuals. Regarding spirometric responses, Torres et al. (1996)
26      provide the only data demonstrating a trend in support of this supposition.
27           Studies demonstrating the interaction between air pollutants and allergic processes in the
28      human nasal airways and rhinoconjunctival tissue have been reviewed by Peden (200Ib) and
29      Riediker et al. (2001), respectively. Ozone exposure of subjects with allergic rhinitis has been
30      shown to induce nasal inflammation and increase airway responsiveness to nonspecific
31      bronchoconstrictors.

        January 2005                             6-16        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Peden et al. (1995), who studied allergic asthmatics exposed to O3 found that O3 causes an
 2      increased response to nasal allergen challenge in addition to nasal inflammatory responses.
 3      Their data suggested that allergic subjects have an increased immediate response to allergen after
 4      O3 exposure.  In a follow-up study, Michelson et al. (1999) reported that 0.4 ppm O3 did not
 5      promote early-phase-response mediator release or enhance the response to allergen challenge in
 6      the nasal airways of mild, asymptomatic dust mite-sensitive asthmatic subjects.  Ozone did,
 7      however, promote an inflammatory cell influx, which helps induce a more significant late-phase
 8      response in this population.
 9           Torres et al. (1996) found that O3 causes an increased response to bronchial allergen
10      challenge in subjects with allergic rhinitis.  This study also measured responses in healthy
11      subjects and mildly allergic asthmatics (see Sections AX6.3.2 andAX6.8). All subjects were
12      exposed to 0.25 ppm O3 for 3 h with IE. Statistically significant O3-induced decrements in FEVj
13      occurred in rhinitics (14.1%), asthmatics (12.5%), and the healthy controls (10.2%), but these
14      responses did not differ statistically between groups. Methacholine responsiveness was
15      significantly increased in asthmatics, but not in subjects with allergic rhinitis.  Airway
16      responsiveness to an individual's historical allergen (either grass and birch pollen, house dust
17      mite, or animal dander) was significantly increased after O3 exposure when compared to FA
18      exposure.  The authors  concluded that subjects with allergic rhinitis, but without asthma, could
19      be at risk if a high O3 exposure is followed by a high dose of allergen.
20           Holz et al. (2002) extended the results of Torres et al. (1996) by demonstrating that
21      repeated daily exposure to lower concentrations of O3 (0.125 ppm for 4 days) causes an
22      increased response to bronchial allergen challenge in subjects with preexisting allergic airway
23      disease, with or without asthma.  These investigators observed no major difference in the pattern
24      of bronchial allergen response between asthmatics or rhinitics, except for  a 10-fold increase in
25      the dose of allergen required to elicit a similar response (> 20% decrease in FEVj) in the
26      asthmatic subjects. Early phase responses were more consistent in subjects with rhinitis and
27      late-phase responses were more pronounced in subjects with asthma. There also was a tendency
28      towards a greater effect of O3 in subjects with greater baseline response to specific allergens
29      (chosen on the basis of skin prick test and history, viz., grass, rye, birch, or alder pollen, house
30      dust mite, or animal dander). These data suggest that the presence of allergic bronchial
31      sensitization, but not a history of asthma, may be a key determinant of increased airway allergen

        Tanuary 2005                               6-17        DRAFT-DO NOT QUOTE OR CITE

-------
 1     responsiveness following exposure to O3 (for a more complete discussion of airway
 2     responsiveness) see Section AX6.8.
 3
 4     6.3.4   Subjects with Cardiovascular Disease
 5          Possibly due to the age of subjects studied, O3 exposure does not appear to result in
 6     significant pulmonary function impairment or evidence of cardiovascular strain in patients with
 7     cardiovascular disease relative to healthy controls. Gong et al. (1998) exposed 10 hypertensive
 8     and 6 healthy adult males, 41 to 78 years of age, to 0.3 ppm O3 for 3 h with IE at 30 L/min.  For
 9     all  subjects combined (no significant group differences), there was an O3-induced decrement of
10     7% in FEVj and an 70% increase in the alveolar-arterial oxygen tension gradient. The overall
11     results did not indicate any major acute cardiovascular effects of O3 in either the hypertensive or
12     normal subjects.
13
14
15     6.4  INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
16           OF RESPONSE
17          Analysis of factors that contribute to intersubj ect variability is important for the
18     understanding of individual responses, mechanisms of response, and health risks associated with
19     acute O3 exposures.  A large intersubj ect variability in response to O3 has been reported by
20     numerous investigators (Adams et al., 1981; Aris et al.,  1995; Folinsbee et al., 1978; Kulle et al.,
21     1985; McDonnell et al., 1983).  The magnitude of individual variability in FEVj response in 2 h
22     IE exposures increases at higher O3 concentrations (Kulle et al., 1985: McDonnell et al., 1983).
23     McDonnell (1996) examined the FEVj response data from three 6.6-h exposure studies
24     conducted at the EPA Health Effects Research Laboratory, and showed that the FEVj responses
25     in FA were small with most tightly grouped around zero. With increasing O3 concentrations
26     between 0.08 and 0.12 ppm, the mean response became asymmetrical with a few individuals
27     experiencing quite large decrements in FEVj  {Intersubject variability observed in O3 dosimetry
28     studies is discussed in Chapter 4.2).
29          As an example of the variation in spirometric responses to O3 exposure, Hazucha et al.
30     (2003) analyzed the distribution of O3 responsiveness in 240 subjects (18 to 60 years of age)
31     exposed to 0.42 ppm O3 (on 3 occasions) for  1.5 h with IE at VE = 20 L/min/m2 BSA.  Across all

       January 2005                             6-18       DRAFT-DO NOT QUOTE OR  CITE

-------
 1      ages, 18% of subjects were weak responders (< 5% FEVj decrement), 39% were moderate
 2      responders, and 43% were strong responders (> 15% FEVj decrement).  Younger subjects
 3      (< 35 years of age) were predominately strong responders, whereas, older subjects (> 35 years of
 4      age) were mainly weak responders. The influence of age on intersubject variability was also
 5      noted by Passannante et al. (1998) who found that subjects under 35 years of age were more like
 6      to be strong responders than older individuals. For repeated exposures, Hazucha et al. (2003)
 7      reported that the reproducibility of FEVj responses was related to the length of time between
 8      exposures. The Spearman correlation coefficient of 0.54 was found between responses for
 9      exposures separated by 105 days (median), whereas, a correlation coefficient of 0.85 was found
10      between responses for exposures separated by only 7 days (median).
11          The more reproducible the subject's response, the more precisely it indicates his/her
12      intrinsic responsiveness.  In 2 h IE O3 exposures, McDonnell et al. (1985b) found a relatively
13      poor FEVj reproducibility (R = 0.58) at the lowest concentration, 0.12 ppm, due, in part, to a
14      lack of specific O3 response or a uniformly small response in the majority of subjects. It was
15      concluded that for 2 h IE  O3 exposures equal to or greater than 0.18 ppm, the intersubject
16      differences in magnitude  of change in FVC and FEVj are quite reproducible over time (21 to
17      385 days; mean = 33 days) and are due primarily to differences in intrinsic responsiveness of
18      individual subjects to O3 exposure.
19          Intersubject variability, mechanisms of response, and health risks associated with acute O3
20      exposures are complicated by a poor association between various O3-induced responses. In a
21      retrospective study of 485 male subjects (ages 18 to 36 yrs) exposed to one of six O3
22      concentrations at one of three activity levels for 2 h, McDonnell et al. (1999) observed
23      significant, but low, Spearman rank order correlations between FEVj response and symptoms of
24      cough (R = 0.39), shortness of breath (R = 0.41), and pain on deep inspiration (R = 0.30). These
25      authors concluded that these responses are related mechanistically to some degree, but indicates
26      that there is not a single factor which is responsible for the observed individual differences in O3
27      responsiveness across the spectrum of symptom and lung function responses.
28          The effect of large intersubject variability on the ability to predict individual
29      responsiveness to O3 was demonstrated by McDonnell et al. (1993). These investigators
30      analyzed the data of 290 male subjects (18 to  32 years of age) who underwent repeat 2 h IE
31      exposures to one or more O3 concentrations ranging from 0.12 to 0.40 ppm. They attempted to

        January 2005                              6-19       DRAFT-DO NOT QUOTE OR CITE

-------
 1     identify personal characteristics (i.e., age, height, baseline pulmonary functions, presence of
 2     allergies, and past smoking history) that might predict individual differences in FEVj response.
 3     Only age contributed significantly to intersubject responsiveness (younger subjects were more
 4     responsive), accounting for just 4% of the observed variance. Interestingly, O3 concentration
 5     accounted for only 31% of the variance, strongly suggesting the importance of as yet undefined
 6     individual characteristics that determine FEVj responsiveness to O3. The authors concluded that
 7     much individual variability in FEVj response to O3 remains unexplained.
 8
 9
10     6.5  FACTORS MODIFYING RESPONSIVENESS TO OZONE
11     6.5.1   Influence of Age
12          Beyond the age of 18 to 20 yrs, spirometric and symptom responses to O3 exposure begin
13     to decline with increasing age. In healthy individuals, the rate of decline in O3 responsiveness
14     appears to be greater in younger (18 to 35 yrs) versus middle aged (35 to 55 yrs) individuals
15     (Passannante et al., 1998; Hazucha et al., 2003). Beyond this age (> 55 yrs), acute O3 exposure
16     elicits minimal spirometric changes.  An average FEVj decrement of-3% has been reported by
17     Gong et al. (1997a) for this older population under a "worst case" exposure scenario (0.24 ppm
18     O3 with 4 h IE). Although Gong et al. (1997a) and others have examined responses to O3
19     exposure in subjects of various ages, the exposure conditions differ between most studies so that
20     age effects remain uncertain.
21          Three recent studies, which analyzed large data sets (>240 subjects) of similarly exposed
22     subjects, show clearly discernable changes in FEVj responses to O3 as a function of age. Seal
23     et al. (1996) analyzed O3-induced spirometric responses in 371 young nonsmokers (18 to
24     35 years of age) exposed for 2.3 h during IE at a VE of 25 L/min/m2 BSA. On average, for the
25     same O3 concentration (C), the response of 25, 30,  and 35 year old individuals are predicted to
26     be 83, 65, and 48%, respectively, of the response in 20 year olds. For example, a 5.4%
27     decrement in FEVj is predicted for 20 year old exposed to 0.12 ppm O3 for 2.3 h IE (VE = 25
28     L/min/m2 BSA), whereas, a similarly exposed 35 yr old is predicted to have only a 2.6%
29     decrement.
30          McDonnell et al. (1997) examined FEVj responses in 485 healthy white males (18 to
31     36 years of age) exposed once for 2 h to an O3 concentration of 0.0, 0.12, 0.18, 0.24, 0.30, or

       January 2005                             6-20        DRAFT-DO NOT QUOTE OR CITE

-------
 1      0.40 ppm at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA).  For the same
 2      exposure conditions (C, VE, and duration), the average responses of 25, 30, and 35 year old
 3      individuals are predicted to be 69, 48, and 33%, respectively, of the response in 20 year olds.
 4      Hazucha et al. (2003) analyzed the distribution of O3 responsiveness in 240 subjects (18 to
 5      60 years of age) exposed to 0.42 ppm O3 for 1.5 h with IE at VE = 20 L/min/m2 BSA. In males,
 6      the FEVj responses of 25, 35, and 50 year olds are predicted to be 94, 83, and 50% ,
 7      respectively, of the average response in 20 year old males. In females,  the FEVj responses of 25,
 8      35, and 50 year olds are predicted to be 82, 46, and 18%, respectively, of the average response in
 9      20 year old females.
10          For subjects aged 18 to 36 yrs, McDonnell et al. (1999) recently reported that symptom
11      responses from  O3 exposure also decrease with increasing age.  Whether the same age-dependent
12      pattern of O3 sensitivity decline also holds for airway reactivity or inflammatory endpoints has
13      not been determined.
14
15      6.5.2    Gender and Hormonal Influences
16          Several studies have suggested that physiological differences between the genders may
17      predispose females to a greater susceptibility to O3. Lower plasma and nasal lavage fluid levels
18      of uric acid (the most prevalent antioxidant) in females relative to males may be a contributing
19      factor (Housley et al.,  1996). Consequently, reduced absorption of O3 in the upper airways may
20      promote its deeper penetration. Dosimetric measurements have shown that the absorption
21      distribution of O3 is independent of gender when absorption is normalized to anatomical dead
22      space (Bush et al., 1996). More recently, Ultman et al. (2004) reported that the whole lung
23      uptake fraction  of O3 was significantly greater in males (91.4%) than females (87.1%). But, this
24      increase in O3 uptake in the males was consistent with their larger VT and smaller fB relative to
25      the females. Furthermore, O3 uptake was not correlated with spirometric responses.  Thus, a
26      differential removal of O3 by uric acid seems to be minimal. In general, the physiologic
27      response of young healthy females to O3 exposure appears comparable to the response of young
28      males (Hazucha et al., 2003). Although, during the follicular phase of the  menstrual cycle, lung
29      function response to O3 is enhanced (Fox et al., 1993).
30
       January 2005                              6-21        DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
6.5.3   Racial, Ethnic and Socioeconomic Status Factors
     A few epidemiologic studies have implied that minorities are more responsive to O3 than
then Caucasians. However, this may be more of a consequence of the overall quality of health
care and SES then an innate sensitivity to oxidants (Gwynn and Thurston, 2001; Seal et al,
1996).  The paucity of data has prevented making any definitive conclusions on the influence of
race, ethnic or other related factors on the responsiveness to O3.

6.5.4   Influence of Physical Activity
     Any physical activity will increase minute ventilation and therefore the dose of inhaled O3.
Consequently, the intensity of physiological response following an acute exposure will be
strongly associated with minute ventilation (see Figure 6-3).
                                                Time (h)

       Figure 6-3.   Predicted O3-induced decrements in FEVt as a function of exposure duration
                    and level of IE (line labels are VE levels) in young healthy adults (20 yrs of
                    age) exposed to 0.3 ppm O3. The illustrated activity levels range from rest
                    (VE = 10 L/min) to moderate exercise (VE = 40 L/min).  Predictions are for
                    Model 1 coefficients in Table 3 of McDonnell et al. (1997).
       Source: Based on McDonnell et al. (1997).
       January 2005
                                         6-22
DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.5.5  Environmental Factors
 2           Since the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996) few human
 3      laboratory studies have examined the potential influence of environmental factors such as rural
 4      versus urban environment, passive cigarette smoke exposure, and bioactive admixtures such as
 5      endotoxin on healthy individual's pulmonary function changes due to O3.
 6           New controlled human exposure studies have confirmed that smokers are less responsive
 7      to O3 than nonsmokers. Spirometric and plethysmographic pulmonary function decline,
 8      nonspecific airway hyperreactivity, and inflammatory response of smokers to O3 were all weaker
 9      than those reported for nonsmokers. Although all these responses are intrinsically related, the
10      functional association between them, as in nonsmokers, has been weak.  Similarly, the time
11      course of development and recovery of these effects, as well their reproducibility, was not
12      different from  nonsmokers. Chronic airway inflammation with desensitization of bronchial
13      nerve endings  and an increased production of mucus may plausibly explain the pseudo-
14      protective effect of smoking (Frampton et al., 1997; Torres et al., 1997).
15           The effect on environment tobacco smoke (ETS) on O3 responses has received very little
16      attention.  In one study, preexposure of mice to sidestream cigarette smoke (ETS  surrogate),
17      elicited no immediate effects, but potentiated subsequent O3-induced inflammatory responses
18      (Yu et al., 2002) (See Chapter 5.4.2 for additional ETS details}.
19           The influence  of ambient temperature on pulmonary effects induced by O3 exposure in
20      humans has been studied infrequently under controlled laboratory conditions. Several
21      experimental human studies have reported additive effects of heat and O3 exposure (see U.S.
22      Environmental Protection Agency, 1986, 1996).  Foster et al. (2000) exposed 9 young healthy
23      subjects for  130 min (IE 10 min at 36 to 39 1/min) to filtered air and to ramp profile O3 at 22°
24      and 30 °C, 45-55% RH. The O3 exposure started at 0.12 ppm, reached the peak of 0.24 ppm
25      mid-way through and subsequently declined to 0.12 ppm at the end of exposure.  At the end of
26      exposure FEVj decreased significantly (p < 0.5) by -8% at 22°C and -6.5% at 30 °C. One day
27      (19 h) later, the decline of 2.3% from baseline was still significant (p < 0.05). FVC decrements
28      were smaller and significant only for the 22 °C condition immediately postexposure. There was
29      a decline in specific airway conductance (sGaw; p < 0.05) at 30°C but not  at 22 °C.  The
30      nonspecific bronchial responsiveness to methacloline assessed as PC50 sGaw was significantly
       January 2005                             6-23        DRAFT-DO NOT QUOTE OR CITE

-------
 1      (p < 0.05) higher one day following O3 exposure at both temperatures but more so at 30 °C.
 2      Thus, these findings suggest that elevated temperature may partially attenuate spirometric
 3      responses but enhance airway reactivity.
 4
 5      6.5.6    Oxidant-Antioxidant Balance
 6          The first line of defense against oxidative stress is antioxidant present in epithelial lining
 7      fluid (ELF) which scavenge free radicals and limit lipid peroxidation. Exposure to O3 depletes
 8      the antioxidant level in nasal ELF probably due to scrubbing of O3 (Mudway et al., 1999),
 9      however, the concentration and the activity of antioxidant enzymes either in ELF or plasma do
10      not appear to be related to O3 responsiveness (Avissar et al., 2000; Blomberg et al., 1999; Samet
11      et al., 2001).  Carefully controlled studies of dietary antioxidant supplementation have
12      demonstrated some protective effects of a-tocopherol and ascorbate on spirometric lung function
13      from O3 but not on the intensity  of subjective symptoms and inflammatory response including
14      cell recruitment, activation and a release of mediators (Samet et al., 2001; Trenga et al., 2001).
15      Dietary antioxidants have also afforded partial protection to asthmatics by attenuating post-
16      exposure bronchial hyperresponsiveness (Trenga et al., 2001). Animal studies (described in
17      Chapter 5.2.1.3) have also demonstrated the protective effects of ELF antioxidants during O3
18      exposures.
19
20      6.5.7    Genetic and Other Factors
21          Several recent studies (Bergamaschi et al., 2001) have reported that genetic polymorphism
22      of antioxidant enzymes may modulate pulmonary function and inflammatory response to  O3
23      challenge.  It appears that healthy carriers of NQO1 wild type in combination with GSTM1 null
24      genotype are more responsive to O3.  Adults with GSTM1 null only genotype did not show O3
25      hyperresponsiveness. In contrast, asthmatic children with GSTM1 null genotype (Romieu et al,
26      2004) were reported to  be more responsive to O3.  In general, the findings between studies are
27      inconsistent and additional, better controlled studies, are needed to clarify influence of genetic
28      polymorphism on O3 responsiveness.
29
30
        January 2005                              6-24       DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.6  REPEATED O3 EXPOSURE EFFECTS
 2          Based on studies reviewed here and in the previous O3 criteria documents (U.S.
 3      Environmental Protection Agency, 1986,  1996), several conclusions can be drawn about
 4      repeated 1 to 2-h O3 exposures. Repeated exposures to O3 can cause an enhanced (i.e., greater)
 5      pulmonary function response on the second day of exposure (see Tables AX6-8 andAX6-9for
 6      added detail).  This enhancement appears to be dependent on the interval between the exposures
 7      (24 h is associated with the greatest increase) and is absent with intervals > 3 days (Bedi et al.,
 8      1985; Folinsbee and Horvath, 1986; Schonfeld et al., 1989). An enhanced response also appears
 9      to depend to some extent on the magnitude of the initial response (Horvath et al.,  1981). Small
10      responses to the first O3 exposure are less likely to result in an enhanced response on the second
11      day of O3 exposure (Folinsbee et al., 1994). With continued daily exposures (i.e., beyond the
12      second day) there is an attenuation of pulmonary function responses, typically after 3 to 5 days
13      of repeated exposure.  This attenuated response persists for less than 1 week (Kulle et al., 1982;
14      Linn et al., 1982b) or as long as 2 weeks (Horvath et al., 1981). In temporal conjunction with
15      pulmonary function changes, symptoms induced by O3, such as cough, PDI, and chest
16      discomfort, are increased on the second exposure day and attenuated with repeated exposure
17      thereafter (Folinsbee et al., 1980, 1998; Foxcroft and Adams, 1986; Linn et al., 1982b).
18      O3-induced changes in airway responsiveness persist longer and attenuate more slowly than
19      pulmonary function and symptoms responses (Dimeo et al., 1981; Kulle et al., 1982), although
20      this has been studied only on a limited basis (Folinsbee et al., 1994).  In longer-duration (4 h to
21      6.6 h), lower-concentration studies that do not cause an enhanced second-day response, the
22      attenuation of response to O3 appears to proceed more rapidly (Folinsbee et al., 1994)  [Effects of
23      repeated exposures on inflammatory responses are discussed in Section 6.9.4).
24
25
26      6.7  EFFECTS  ON EXERCISE PERFORMANCE
27          The effects of acute O3 inhalation on endurance exercise performance have been examined
28      in numerous controlled laboratory studies. These studies were discussed in the 1996 O3 AQCD
29      (U.S. Environmental Protection Agency, 1996) and  can be divided into two categories: (1) those
30      that examined the effects of acute O3 inhalation on maximal oxygen uptake (VO2max) and
       January 2005                             6-25        DRAFT-DO NOT QUOTE OR CITE

-------
 1      (2) those that examined the effects of acute O3 inhalation on the ability to complete strenuous
 2      continuous exercise protocols of up to 1 h in duration.
 3           In brief, endurance exercise performance and VO2max may be limited by acute exposure to
 4      O3 (Adams and Schelegle, 1983; Schelegle and Adams, 1986; Gong et al., 1986; Foxcroft and
 5      Adams, 1986; Folinsbee et al., 1977; Linder et al., 1988). Gong et al. (1986) and Schelegle and
 6      Adams (1986) found that significant reductions in maximal endurance exercise performance may
 7      occur in well-conditioned athletes while they perform CE (VE  > 80 L/min) for 1 h at O3
 8      concentrations > 0.18 ppm. Reports from studies of exposure to O3 during high-intensity
 9      exercise indicate that breathing discomfort associated with maximal ventilation may be an
10      important factor in limiting exercise performance in some, but not all, subjects.
11
12
13      6.8  EFFECTS ON AIRWAY RESPONSIVENESS
14           Airway or bronchial hyperresponsiveness (BHR) refers to a condition in which the
15      propensity for the airways to bronchoconstrict due to a variety of stimuli becomes augmented.
16      Airway responsiveness is typically quantified by measuring the decrement in pulmonary
17      function (i.e., spirometry or plethysmography) following the inhalation of small amounts of an
18      aerosolized bronchoconstrictor agent (specific [antigen, allergen] or nonspecific [methacholine,
19      histamine]) or a measured stimulus (e.g., exercise, cold air).
20           Ozone exposure causes an increase in nonspecific airway responsiveness as indicated by a
21      reduction in the concentration of methacholine or histamine required to produce a given
22      reduction in FEVj or increase in SRaw.  Increased airway responsiveness is an important
23      consequence of exposure to O3 because its presence means that the airways are predisposed to
24      narrowing on inhalation of a variety of stimuli (e.g., specific allergens, SO2, cold air).
25           Ozone exposure of asthmatic subjects, who characteristically have increased airway
26      responsiveness at baseline, can cause further increases in responsiveness (Kreit et al., 1989).
27      Similar relative changes in airway responsiveness are seen in asthmatics  exposed to O3 despite
28      their markedly different baseline airway responsiveness. Several studies (Torres et al., 1996;
29      Kehrl et al., 1999; Molfino et al.,  1991) have been published suggesting an increase in specific
30      (i.e., allergen-induced) airway reactivity. An important aspect of increased airway
        January 2005                              6-26       DRAFT-DO NOT QUOTE OR CITE

-------
 1      responsiveness after O3 exposure is that this represents a plausible link between ambient O3
 2      exposure and increased hospital admissions for asthma.
 3           Changes in airway responsiveness after O3 exposure appear to be resolved more slowly
 4      than changes in FEVj or respiratory symptoms (Folinsbee and Hazucha, 2000).  Furthermore, in
 5      studies of repeated exposure to O3, changes in airway responsiveness tend to be somewhat less
 6      susceptible to attenuation with consecutive exposures than changes in FEVj (Dimeo et al., 1981;
 7      Folinsbee et al., 1994; Gong et al., 1997b; Kulle et al.,  1982). Increases in airway
 8      responsiveness do not appear to be strongly associated with decrements in lung function or
 9      increases in symptoms.
10           The mechanism of O3-induced increases in airway responsiveness is only partially
11      understood, but it appears to be associated with a number of cellular and biochemical changes in
12      airway tissue. Although inflammation could play a role in the increase in airway responsiveness,
13      cyclooxygenase inhibitors have not been effective at blocking the O3-induced influx of PMNs
14      into BAL fluid (Hazucha et al., 1996;  Ying et al., 1990).  Therefore, O3-induced airway
15      responsiveness may not be due to the presence of PMNs in the airway or to the release of
16      arachidonic acid metabolites.  Rather, it seems likely that the mechanism for this response is
17      multifactorial, possibly involving the presence of cytokines, prostanoids, or neuropeptides;
18      activation of macrophages, eosinophils, or mast cells; and epithelial damage that increases direct
19      access of mediators to the smooth muscle or receptors in the airways that are responsible for
20      reflex bronchoconstriction.
21
22
23      6.9  INFLAMMATION AND  HOST DEFENSE EFFECTS
24      6.9.1  Introduction
25           Short-term exposure of humans to O3 can cause acute inflammation and that long-term
26      exposure of laboratory animals results in a chronic inflammatory state (see Chapter 5).  The
27      relationship between repetitive bouts of acute inflammation in humans caused by O3 and the
28      development of chronic respiratory disease is unknown.
29           The presence of neutrophils (PMNs) in the lung has long been accepted as a hallmark of
30      inflammation and is an important indicator that O3 causes inflammation in the lungs. It is
31      apparent, however, that inflammation  within airway tissues may persist beyond the point that

        January 2005                              6-27        DRAFT-DO NOT QUOTE OR CITE

-------
 1     inflammatory cells are found in BAL fluid.  Soluble mediators of inflammation such as the
 2     cytokines (IL-6, IL-8) and arachidonic acid metabolites (e.g., PGE2, PGF2a, thromboxane, and
 3     leukotrienes [LTs] such as LTB4) have been measured in the BAL fluid of humans exposed to
 4     O3. In addition to their role in inflammation, many of these compounds have
 5     bronchoconstrictive properties and may be involved in increased airway responsiveness
 6     following O3 exposure.
 7           Some recent evidence suggests that changes in small airways function may provide a
 8     sensitive indicator of O3 exposure and effect, despite the fact that inherent variability in their
 9     measurement by standard spirometric approaches make their assessment difficult.  Observations
10     of increased functional responsiveness of these areas relative to the more central airways, and of
11     persistent effects following repeated exposure, may indicate that further investigation of
12     inflammatory processes in these regions is warranted.
13
14     6.9.2   Inflammatory Response in the Upper Respiratory Tract
15           The nasal passages constitute the primary portal for inspired air at rest and, therefore, the
16     first region of the  respiratory tract to come in contact with airborne pollutants. Nikasinovic et al.
17     (2003) recently reviewed the literature of laboratory-based nasal inflammatory studies published
18     since 1985.  Nasal lavage (NL) has provided a useful tool for assessing O3-induced inflammation
19     in the nasopharynx. Increased levels of PMNs in the NL fluid of humans exposed to 0.5 ppm
20     O3 at rest for 4 h has been reported (Graham et al., 1988; Bascom et al., 1990).
21           Graham and Koren (1990) compared inflammatory mediators present in both the NL and
22     BAL fluids of humans exposed to 0.4 ppm O3 for 2 h.  Similar increases in PMN were observed
23     in NL and BAL, suggesting a qualitative correlation between inflammatory changes in the lower
24     airways (BAL) and the upper respiratory tract (NL). Torres et al. (1997) compared NL and BAL
25     in smokers and nonsmokers exposed to 0.22 ppm O3 for 4 h. In contrast to Graham and Koren
26     (1990), they did not find a relationship between numbers or percentages of PMNs in the nose
27     and the lung, perhaps in part due to the variability observed in their NL recoveries. Albumin, a
28     marker of epithelial cell permeability, was increased 18 h later, but not immediately after
29     exposure, as seen by Bascom et al. (1990).
30           McBride et al. (1994) reported that asthmatic subjects were more sensitive than non-
31     asthmatics to upper airway inflammation at an O3 concentration (0.24 ppm (1.5 h)) that did not

       January 2005                             6-28        DRAFT-DO NOT QUOTE OR CITE

-------
 1      affect lung or nasal function or biochemical mediators. A significant increase in the number of
 2      PMNs in NL fluid was detected in the asthmatic subjects both immediately and 24 h after
 3      exposure. Peden et al. (1995) also found that O3 at a concentration of 0.4 ppm had a direct nasal
 4      inflammatory effect, and reported a priming effect on the response to nasal allergen challenge, as
 5      well. A subsequent study in dust mite-sensitive asthmatic subjects indicated that O3 at this
 6      concentration enhanced eosinophil influx in response to allergen, but did not promote early
 7      mediator release or enhance the nasal response to allergen (Michelson et al., 1999).  Similar to
 8      observations made in the lower airways, the presence of O3 molecular "targets" in nasal lining
 9      fluid is likely to provide some level of local protection against exposure. In a study of healthy
10      subjects exposed to 0.2 ppm O3 for 2 h, Mudway and colleagues (1999) observed a significant
11      depletion of uric acid in NL fluid at 1.5 h following exposure.
12
13      6.9.3   Inflammatory Response in the Lower Respiratory Tract
14           Seltzer et al. (1986) were the first to demonstrate that exposure of humans to O3 resulted in
15      inflammation in the lung. Bronchoalveolar lavage fluid (3 h post-exposure) from subjects
16      exposed to O3 contained increased PMNs as well as increased levels of PGE2, PGF2a, and TXB2
17      compared to fluid from air-exposed subjects. Koren et al. (1989a,b) described inflammatory
18      changes 18 h after O3 exposure.  In addition to an eightfold increase in PMNs, Koren et al.
19      reported a two-fold increase in BAL fluid protein,  albumin, and immunoglobulin G (IgG) levels,
20      suggestive of increased epithelial cell permeability. There was a 12-fold increase in IL-6 levels,
21      a two-fold increase in PGE2, and a two-fold increase in the complement component,  C3a.
22      Evidence for stimulation of fibrogenic processes in the lung was shown by significant increases
23      in coagulation components, Tissue Factor and Factor VII (McGee et al., 1990), urokinase
24      plasminogen activator and fibronectin (Koren et al., 1989a). Subsequent studies by Lang et al.,
25      (1998), using co-cultures of cells of the BEAS-2B  bronchial epithelial line and of the HFL-1
26      lung fibroblast line, provided additional information about O3-induced fibrogenic processes.
27      They demonstrated that steady-state mRNA levels of both alpha 1 and procollagens type I and
28      III in the fibroblasts were increased following O3 exposure and that this effect was mediated by
29      the O3-exposed epithelial cells. This group of studies demonstrated that exposure to O3 results in
30      an inflammatory reaction in the lung, as evidenced by increases in PMNs and proinflammatory
31      compounds. Furthermore, they demonstrated that  cells and mediators capable of damaging

        January 2005                              6-29       DRAFT-DO NOT QUOTE OR CITE

-------
 1      pulmonary tissue are increased after O3 exposure, and provided early suggestion of the potential
 2      importance of the epithelial cell-myofibroblast "axis" in modulating fibrotic and fibrinolytic
 3      processes in the airways.
 4           Isolated lavage of the mainstream bronchus using balloon catheters or BAL using small
 5      volumes of saline have been used to assess O3-induced changes in the large airways. Studies
 6      collecting lavage fluid from isolated airway segments after O3 exposure indicate increased
 7      neutrophils in the airways (Aris et al., 1993; Balmes et al., 1996; Scannell et al., 1996). Other
 8      evidence of airway neutrophil increase comes from studies in which the initial lavage fraction
 9      ("bronchial fraction") showed increased levels of neutrophils (Schelegle et al., 1991; Peden
10      et al., 1997; Balmes et al., 1996; Torres et al., 1997). Bronchial biopsies show increased PMNs
11      in airway tissue (Aris et al.,1993) and, in sputum collected after O3 exposure, neutrophil numbers
12      are elevated (Fahy et al., 1995).
13           Increased BAL protein, suggesting O3-induced changes in epithelial permeability (Koren
14      et al., 1989a, 1991 and Devlin et al., 1991) supports earlier work in which increased epithelial
15      permeability, as measured by increased clearance of radiolabled diethylene triamine pentaacetic
16      acid (99mTc-DTPA) from the lungs of humans exposed to O3, was demonstrated (Kehrl et al.,
17      1987).  In addition, Foster and Stetkiewicz (1996) have shown that increased permeability
18      persists for at least 18-20 h and the effect is greater at the lung apices than at the base.  In a study
19      of mild atopic asthmatics exposed to 0.2 ppm O3 for 2 h,  Newson, et al. (2000) observed a 2-fold
20      increase in the  percentage of PMNs present at 6 hours post exposure, with no change in markers
21      of increased permeability as assessed by sputum induction. By 24 h, the neutrophilia was seen
22      to subside while levels of albumin, total protein, myeloperoxidase, and eosinophil cationic
23      protein increased significantly. It was concluded that the transient PMN influx induced by acute
24      exposure of these asthmatic subjects was followed by plasma extravasation and the activation of
25      both PMNs and eosinophils within the airway tissues.  Such changes in permeability associated
26      with acute inflammation may provide better access  of inhaled antigens, particulates, and other
27      substances to the submucosal region.
28           Devlin et al. (1991) reported an inflammatory response in subjects exposed to 0.08 and
29      0.10 ppm O3 for 6.6 h. Increased numbers of PMNs and  levels of IL-6 were found at both
30      O3 concentrations, suggesting that lung inflammation from O3 can occur as a consequence of
31      prolonged exposure to ambient levels while exercising.  Interestingly, those individuals who had

        January 2005                               6-30        DRAFT-DO NOT QUOTE OR CITE

-------
 1      the largest increases in inflammatory mediators in this study did not necessarily have the largest
 2      decrements in pulmonary function, suggesting that separate mechanisms underlie these two
 3      responses. The absence of a relationship between spirometric responses and inflammatory cells
 4      and markers has been reported in several studies (Balmes et al., 1996; Schelegle et al., 1991;
 5      Torres et al.,  1997; Hazucha et al., 1996; Blomberg et al., 1999).  These observations relate
 6      largely to disparities in the times of onset and duration following single exposures.
 7           As indicated above, a variety of potent proinflammatory mediators have been reported to
 8      be released into the airway lumen following O3 exposure. Studies of human alveolar
 9      macrophages (AM) and  airway epithelial cells exposed to O3 in vitro suggest that most mediators
10      found in the BAL fluid of O3-exposed humans are produced by epithelial cells. Macrophages
11      exposed to O3 in vitro showed only small increases in PGE2 (Becker et al., 1991). In contrast,
12      airway epithelial cells exposed in vitro to O3 showed large concentration-dependent increases in
13      PGE2, TXB2, LTB4, LTC4, and LTD4 (McKinnon et al., 1993) and increases in IL-6,  IL-8, and
14      fibronectin at O3 concentrations as low as 0.1 ppm (Devlin et al.,  1994). Macrophages lavaged
15      from subjects exposed to 0.4 ppm (Koren et al., 1989a) showed changes in the rate of synthesis
16      of 123 different proteins, whereas AMs exposed to O3 in vitro showed changes in only six
17      proteins, suggesting that macrophage function was altered by mediators released from other
18      cells. Furthermore, recent evidence suggests that the release of mediators from AMs may be
19      modulated by the products of O3-induced oxidation of airway lining fluid components, such as
20      human surfactant protein A (Wang et al., 2002).
21           Although the release of mediators has been demonstrated to occur at exposure
22      concentrations and times that are minimally cytotoxic to airway cells, potentially detrimental
23      latent effects have been  demonstrated in the absence of cytotoxicity.  These include the
24      generation of DNA single strand breaks (Kozumbo et al., 1996) and the loss of cellular
25      replicative activity (Gabrielson et al., 1994) in bronchial epithelial cells exposed in vitro, and the
26      formation of protein and DNA adducts. A highly toxic aldehyde  formed during O3-induced lipid
27      peroxidation is 4-hydroxynonenal  (HNE). Healthy human subjects exposed to 0.4 ppm O3 for 1
28      h underwent BAL 6 h later. Analysis of lavaged alveolar macrophages by Western blot
29      indicated increased levels  of a 32-kDa HNE-protein adduct, as well as 72-kDa heat shock protein
30      and ferritin, in O3- versus air-exposed subjects (Hamilton et al., 1998).  In a recent study of
31      healthy subjects exposed to 0.1 ppm O3 for 2 h (Corradi et al., 2002), formation of 8-hydroxy-2'-

        January 2005                               6-31        DRAFT-DO NOT QUOTE OR CITE

-------
 1      deoxyguanosine (8-OHdG), a biomarker of reactive oxidant species (ROS)-DNA interaction,
 2      was measured in peripheral blood lymphocytes.  At 18 h post exposure, 8-OHdG was
 3      significantly increased in cells compared to pre-exposure levels, presumably linked to concurrent
 4      increases in chemical markers of ROS.  Of interest, the increase in 8-OHdG was only significant
 5      in a subgroup of subjects with the wild genotype for NAD(P)H:quinone oxidoreductase and the
 6      null genotype for glutathione-S-transferase Ml, suggesting that polymorphisms in redox
 7      enzymes may confer "susceptibility' to O3 in some individuals.  The generation of ROS
 8      following exposure to O3 has been shown to be associated with a wide range of responses.  In a
 9      recent study, ROS production by alveolar macrophages lavaged from  subjects exposed to
10      0.22 ppm for 4 h was assessed by flow cytometry (Voter et al., 2001).  Levels were found to be
11      significantly elevated 18 h post exposure and associated with several markers of increased
12      permeability. An in vitro study of human tracheal epithelial cells exposed to  O3 indicated that
13      generation of ROS resulted in decrease in synthesis of the bronchodilatory prostaglandin, PGE2,
14      as a result of inactivation of prostaglandin endoperoxide G/H synthase 2 (Alpert et al., 1997).
15      These and similar studies indicate that the responses to products of O3 exposure in the airways
16      encompass a broad range of both stimulatory and inhibitory activities, many of which may  be
17      modulated by susceptibility factors upstream in the exposure process, at the level of
18      compensating for the imposed oxidant stress.
19           The inflammatory responses to O3 exposure also have been studied in asthmatic subjects
20      (Basha et al., 1994; Scannell et al., 1996; Peden et al., 1997).  In these studies, asthmatics
21      showed significantly more neutrophils in the BAL (18 h post-exposure) than similarly exposed
22      healthy individuals.  In one of these studies (Peden et  al., 1997), which included only allergic
23      asthmatics who tested positive for Dematophagoides farinae  antigen, there was an eosinophilic
24      inflammation (2-fold increase), as well as neutrophilic inflammation (3-fold increase). In a
25      study of subjects with intermittent asthma that utilized a 2-fold higher concentration of O3 (0.4
26      ppm) for 2 h, increases in eosinophil cationic protein,  neutrophil elastase and  IL-8 were found to
27      be significantly increased 16 h post-exposure and comparable in induced sputum and BAL fluid
28      (Hiltermann et  al,  1999). In two  studies (Basha et al.,  1994; Scannell  et al., 1996), IL-8 was
29      significantly higher in post-O3 exposure BAL in  asthmatics compared to non-asthmatics,
30      suggesting a possible mediator for the increased  neutrophilic inflammation in those subjects.
31      In a recent study comparing the neutrophil response to O3 at a concentration and exposure time

        January 2005                              6-32        DRAFT-DO NOT QUOTE OR CITE

-------
 1      similar to those of the latter three studies, Stenfors and colleagues (2002) were unable to detect a
 2      difference in the increased neutrophil numbers between 15 mild asthmatic and 15 healthy
 3      subjects by bronchial wash at the 6 h post-exposure time point. These results suggest that, at
 4      least with regard to neutrophil influx, differences between healthy and asthmatic individuals
 5      develop gradually following exposure and may not become evident until later in the process.
 6      In another study, mild asthmatics who exhibited a late phase underwent allergen challenge 24 hrs
 7      before a 2 h exposure to 0.27 ppm O3 or filtered air in a cross-over design (Vagaggini et al.,
 8      2002).  At 6 h post-exposure, eosinophil numbers in induced sputum were found to be
 9      significantly greater after O3 than after air. Studies such as these suggest that the time course of
10      eosinophil and neutrophil influx following O3 exposure can occur to levels detectable within the
11      airway lumen by as early as 6 h. They also suggest that the previous or concurrent activation of
12      proinflammatory pathways within the  airway epithelium may enhance the inflammatory effects
13      of O3. For example, in an in vitro study of epithelial cells from the upper and lower respiratory
14      tract, cytokine production induced by rhinovirus infection was enhanced synergistically by
15      concurrent exposure to O3 at 0.2 ppm for 3 h (Spannhake et al, 2002).  The use of bronchial
16      mucosal biopsies has also provided important insight into the modulation by O3 of existing
17      inflammatory processes within asthmatics. In a study of healthy and allergic asthmatic subjects
18      exposed to 0.2 ppm O3 or filtered air for 2  h, biopsies were performed 6 hr following exposure
19      (Bosson et al., 2003). Monoclonal antibodies were used to assess epithelial expression of a
20      variety of cytokines and chemokines.  At baseline (air exposure), asthmatic subjects showed
21      significantly higher expression of interleukins (IL)-4 and -5. Following O3 exposure, the
22      epithelial expression of IL-5, IL-8, granulocyte-macrophage colony-stimulating  factor (GM-
23      CSF) and epithelial cell-derived neutrophil-activating peptide 78 (ENA-78) was significantly
24      greater in asthmatic  subjects, as compared to healthy subjects. In vitro studies of bronchial
25      epithelial cells derived by biopsy from nonatopic, nonasthmatic subjects and asthmatic subjects
26      also demonstrated the preferential release of GM-CSF and also of regulated on activation,
27      normal T cell-expressed and -secreted (RANTES) from asthmatic cells following O3 exposure.
28           The time course of the inflammatory response to O3 in humans has not been explored fully.
29      Nevertheless, studies in which BAL was performed 1-3 h (Devlin et al., 1990; Koren et al.,
30      1991; Seltzer et al., 1986) after exposure to 0.4 ppm O3  demonstrated that the inflammatory
31      response is quickly initiated, and other studies (Koren et al., 1989a,b; Torres et al., 1997;

        January 2005                              6-33        DRAFT-DO NOT QUOTE OR CITE

-------
 1      Scannell et al., 1996; Balmes et al., 1996) indicated that, even 18 h after exposure, inflammatory
 2      mediators such as IL-6 and PMNs were still elevated.  However, different markers show peak
 3      responses at different times. Ozone-induced increases in IL-8, IL-6, and PGE2 are greater
 4      immediately after O3 exposure, whereas BAL levels of fibronectin and plasminogen activator are
 5      greater after 18 h. PMNs and some products (protein, Tissue Factor) are similarly elevated both
 6      1 and 18 h after O3 exposure (Devlin et al., 1996; Torres et al., 1997).  Schelegle et al. (1991)
 7      found increased PMNs in the "proximal airway" lavage at 1, 6, and 24 h after O3 exposure, with
 8      a peak response at 6 h. In a typical BAL sample, PMNs were elevated only at the later time
 9      points.  This is consistent with the greater increase 18 h after exposure seen by Torres et al.
10      (1997). In addition to the influx of PMNs and (in allergic asthmatics) eosinophils, lymphocyte
11      numbers in BAL were also seen to be elevated significantly at 6 h following exposure of healthy
12      subjects to 0.2 ppm O3 for 2 h (Blomberg et al., 1997).  Analysis of these cells by flow cytometry
13      indicated the increased presence of CD3+, CD4+ and CD8+ T cell subsets.  This same laboratory
14      later demonstrated that within 1.5 h following exposure of healthy subjects to the same O3
15      regimen, expression of human leukocyte antigen (HLA)-DR on lavaged macrophages underwent
16      a significant, 2.5-fold increase (Blomberg et al.,  1999). The significance of these alterations in
17      immune system components and those in IL-4 and IL-5 expression described above in the
18      studies of Bosson et al. (2003) has not been fully explored and may suggest a role for O3 in the
19      modulation of immune inflammatory processes.
20
21      6.9.4    Effects of Repeated Exposures and Adaptation of Responses
22          Physiologic and symptomatic responses in humans following repeated exposure to O3 were
23      discussed in Section 6.6. Inflammatory responses upon repeated O3 exposures are discussed in
24      this section. Animal studies suggest that while inflammation may be diminished with repeated
25      exposure, underlying damage to lung epithelial cells continues (Tepper et al.,  1989).  Markers
26      from BALF following both 2-h (Devlin et al., 1997) and 4-h (Christian et al.,  1998; Torres et al.,
27      2000) repeated O3 exposures (up to 5 days) indicate that there is ongoing cellular damage
28      irrespective of the attenuation of some cellular inflammatory responses of the airways,
29      pulmonary function, and symptom responses.
30          Devlin et al. (1997) examined the inflammatory responces of humans repeatedly exposed
31      to 0.4 ppm O3 for 5 consecutive days. Several indicators of inflammation (e.g., PMN influx, IL-

        January 2005                             6-34       DRAFT-DO NOT QUOTE OR CITE

-------
 1      6, PGE2, BAL protein, fibronectin) were attenuated after 5 days of exposure (i.e., values were
 2      not different from FA).  Several markers (LDH, IL-8, total protein, epithelial cells) did not show
 3      attenuation, indicating that tissue damage probably continues to occur during repeated exposure.
 4      The recovery of the inflammatory response occurred for some markers after 10 days, but some
 5      responses were not normalized even after 20 days. The continued presence of markers of
 6      cellular injury indicates a persistent but not necessarily recognized due to attention of
 7      spirometric and symptom responses to O3.
 8           Christian et al. (1998) randomly subjected heathy subjects to a single exposure and to 4
 9      consecutive days of exposure to 0.2 ppm O3 for 4 h. Both "bronchial" and "alveolar" fractions
10      of the BAL showed decreased numbers of PMNs and fibronectin  concentration at day 4 versus
11      the single exposure, and a decrease in IL-6 levels in the alveolar fraction.
12           Following a similar study design and exposure parameters, Torres et al. (2000) found both
13      functional and BAL cellular responses to O3 were abolished at 24 h postexposure following the
14      fourth exposure day. However, levels of total protein, IL-6, IL-8, reduced glutathione and ortho-
15      tyrosine were still increased significantly.  In addition, visual scores for bronchitis, erythema and
16      the numbers of neutrophils in the mucosal biopsies were increased.  Their results indicate that,
17      despite reduction of some  markers of inflammation in BAL and measures of large airway
18      function, inflammation within the airways persists following repeated exposure to O3.
19           Holz, et al. (2002) made a comparison of early and  late responses to allergen challenge
20      following O3 in subjects with allergic rhinitis or allergic asthma. With some variation, both early
21      and late FEVj and cellular responses in the two subject groups were significantly enhanced by 4
22      consecutive days of exposure to 0.125 ppm O3 for 3 h.
23           In another study, Frank and colleagues (2001) exposed healthy subjects to FA and to O3
24      (0.25 ppm, 2 h) on 4 consecutive days each, with pulmonary function measurements being made
25      prior to and following each exposure. BAL was  performed on day 5, 24 h following the last
26      exposure. On day 5, PMN numbers remained significantly higher following O3  compared to FA.
27      Of particular note in this study was the observation that small  airway function, assessed by
28      grouping values for isovolumetric FEF25_75, VmaxSO and Vmax75 into a single value, showed
29      persistent reduction from day 2 through day 5. Following exposure of and asthmatic and healthy
30      subjects for one day to 0.4 ppm O3 for 2 h, Alexis et al. (2000) have also reported that variables
31      representing small airways function (viz., FEF25, FEF50, FEF 60P, FEF75) demonstrated the

        January 2005                              6-3 5        DRAFT-DO NOT QUOTE OR CITE

-------
 1     greatest O3-induced decline in the asthmatic subjects.  These data suggest that techniques
 2     monitoring the function in the small peripheral airway regions, the primary sites of O3 uptake in
 3     the lung, may provide important information regarding both acute and cumulative effects of O3
 4     exposure.
 5
 6     6.9.5   Effect of Anti-Inflammatory and other Mitigating Agents
 7           Pretreatment of healthy subjects with non-steroidal anti-inflammatory drugs (ibuprofen,
 8     etc.) has been found to partially suppress development of airway inflammation and pulmonary
 9     function changes (U.S. Environmental Protection Agency, 1996). Although atropine blocked the
10     increase in Raw in response to O3 exposure, it did not alter the spirometric or symptom
11     responses (Beckett et al., 1985).  Similarly, albuterol and salbutamol, which had no effect on O3-
12     induced changes in spirometry, also had no effect of symptom responses (McKenzie et al., 1987;
13     Gongetal.,  1988). The anti-inflammatory medications indomethacin and ibuprofen, which
14     partially inhibit the spirometric responses to O3 exposure, also cause a reduction in respiratory
15     symptoms (Schelegle et al., 1987; Hazucha et  al., 1994).  Indomethacin attenuates decrements in
16     FEVj and FVC in healthy subjects, but not asthmatics (Alexis et al., 2000). In contrast,
17     inhalation of the corticosteroid budesonide does not prevent or even attenuate O3-induced
18     responses in healthy subjects as assessed by measurements of lung function, bronchial reactivity
19     and airway inflammation (Nightingale et al., 2000). In asthmatic subjects, budesonide  decreases
20     airway neutrophil influx following O3 exposure (Vagaggini et al., 2001). This suggests that
21     corticosteroids may be effective  only when the inflammation is already present, such as in
22     asthmatics.
23
24     6.9.6   Changes in Host Defense  Capability Following Ozone Exposures
25           A number of studies clearly show that a  single acute exposure (1 to 4 h) of humans to
26     moderate concentrations of O3 (0.2 to 0.6 ppm) while exercising at moderate to heavy levels
27     results in a number of cellular and biochemical changes in the lung including an inflammatory
28     response characterized by increased numbers of PMNs, increased permeability of the epithelial
29     cells lining the respiratory tract,  cell damage, and production of proinflammatory cytokines and
30     prostaglandins. This response can be detected as early as 1 h after exposure (Koren et al., 1991;
       January 2005                             6-36        DRAFT-DO NOT QUOTE OR CITE

-------
 1      Schelegle et al., 1991) and persists for at least 18 h (Aris et al., 1993; Koren et al., 1989a). The
 2      response profile of these mediators is not defined adequately, although it is clear that the time
 3      course of response varies for different mediators and cells (Devlin et al.,  1997; Schelegle et al.,
 4      1991).  These changes also occur in humans exposed to 0.08 and 0.10 ppm O3 for 6 to 8 h
 5      (Devlin et al., 1991; Peden et al., 1997).  Decrements in the ability of AMs to phagocytose
 6      microorganisms have also been reported.  Ozone also causes inflammatory changes in the nose,
 7      as indicated by increased levels of PMNs and albumin, a marker for increased epithelial cell
 8      permeability.  Nasal lavage analyses, however, are not necessarily parallel to BAL analyses.
 9           There appears to be no strong correlation between any of the measured cellular and
10      biochemical changes and changes in lung function measurements,  suggesting that different
11      mechanisms may be responsible for these processes (Balmes et al., 1996; Devlin et al., 1991).
12      The idea of different mechanisms  is supported by a study in which ibuprofen, a cyclooxygenase
13      inhibitor, blunted the O3-induced decrements in lung function without altering the O3-induced
14      increase in PMNs or epithelial cell permeability  (Hazucha et al., 1996).
15           In vitro studies suggest that epithelial cells are the primary target of O3 in the lung and that
16      O3 induces them to produce many of the mediators found in the BAL fluid of humans exposed to
17      O3.  Although O3 does not induce AMs to produce these compounds in large quantities, it does
18      directly impair the ability of AMs to phagocytose and kill microorganisms.
19           Only two studies (Foster et al., 1987; Gerrity et al.,  1993) have investigated the effect of
20      O3 exposure on mucociliary particle clearance in humans.  Foster et al. (1987) measured
21      clearance during and after a 2 h exposure to 0.4 ppm O3.  Gerrity et al.  (1993) measured
22      clearance at 2 h postexposure (0.4 ppm O3), by which time, sRaw had returned to baseline and
23      FVC was within 5% of baseline (versus an 11%  decrement immediately postexposure). Foster
24      et al. (1987) found a stimulatory effect of acute O3 exposure on mucociliary clearance. Gerrity
25      et al. (1993), who observed no effect on clearance, suggested that transient clearance increases
26      are coincident to pulmonary function responses.  Investigators in both studies suggested that
27      O3-induced increases in mucociliary clearance could be mediated by cholinergic  receptors.
28
29
        January 2005                              6-37        DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.10  EXTRAPULMONARY EFFECTS OF OZONE
 2           Ozone reacts rapidly on contact with respiratory system tissue and is not absorbed or
 3      transported to extrapulmonary sites to any significant degree as such. Human exposure studies
 4      discussed in the previous criteria documents (U.S. Environmental Protection Agency, 1986,
 5      1996) failed to demonstrate any consistent extrapulmonary effects. More recently, some human
 6      exposure studies have attempted to identify specific markers of exposure to O3 in blood. Foster
 7      et al. (1996) found a reduction in the serum levels of the free radical scavenger a-tocopherol
 8      after O3 exposure.  Liu et al. (1997, 1999) used a salicylate metabolite, 2,3, dehydroxybenzoic
 9      acid (DHBA), to indicate increased levels of hydroxyl radical which hydroxylates salicylate to
10      DHBA.  Increased DHBA levels after exposure to 0.12 and 0.40 ppm suggest that O3 increases
11      production of hydroxyl radical.  The levels of DHBA were correlated with changes in
12      spirometry.
13           Gong et al. (1998) monitored ECG, HR, cardiac output, blood pressure, oxygen saturation,
14      and chemistries, as well as calculating other hemodynamic variables (e.g., stroke volume,
15      vascular resistance, rate-pressure products) in both healthy and hypertensive adult males, 41 to
16      78 years of age. No major acute cardiovascular effects were found in either the normal or
17      hypertensive subjects after exposure to 0.3 ppm O3 for 3 h with intermittent exercise at 30 L/min.
18      Statistically significant O3 effects for both groups combined were a decrease in FEVl3 and
19      increases in HR, rate-pressure product, and the alveolar-to-arterial PO2 gradient, which might be
20      more important in some patients with severe cardiovascular disease.
21
22
23      6.11  EFFECTS OF OZONE MIXED WITH OTHER POLLUTANTS
24           Over the past 10 years only a handful of human controlled studies have examined the
25      effects of pollutant mixtures containing O3.  The results of a controlled study on children (Linn
26      et al., 1997), designed to approximate exposure conditions of an epidemiologic study (Neas
27      et al., 1995) by matching the population and exposure atmosphere (0.1 ppm  O3, 0.1 ppm SO2 and
28      101 |ig/m2 H2SO4), did not support the findings of this epidemiologic study.  The study points
29      out the difficulties in attempting to link the outcomes of epidemiologic and controlled studies.
30      Another vulnerable population, asthmatics, demonstrated enhanced airway reactivity to house
31      dust mite following exposures to O3, NO2, and the combination of the two gases. Spirometric

        January 2005                             6-3 8        DRAFT-DO NOT QUOTE OR CITE

-------
 1      response, however, was impaired only by O3, and O3+NO2 at higher concentrations (Jenkins
 2      et al., 1999). It is plausible that uneven longitudinal absorption of SO2, NO2, and O3 in the
 3      conducting airways may influence a response. Ozone has been found to be scrubbed more
 4      efficiently in proximal airways and to penetrate less into the distal airways than either SO2 and
 5      NO2 (Rigas et al., 1997).  Inhalation of a mixture of PM25 and O3 by healthy subjects increased
 6      brachial artery tone and reactivity (Brook et al., 2002).  Since no other cardiovascular endpoints
 7      were affected by the exposure, the pathophysiological importance of this observation remains
 8      unclear.
 9          All in all, the contention that air pollutant mixtures elicit stronger pathophysiologic effects
10      than individual pollutants of the mix is only weakly supported by human studies of either healthy
11      or at-risk population. The studies summarized in this section complement the studies reviewed
12      in the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996). Regarding the latter, the
13      mobile  laboratory comparative studies of exercising athletes (Avol et al., 1984, 1985) with
14      chamber exposures to oxidant-polluted ambient air (mean O3 concentration of 0.153 ppm) and
15      purified air containing a controlled concentration of generated O3 at 0.16 ppm showed similar
16      pulmonary function responses and symptoms.  These results strongly suggest that acute
17      exposures of coexisting ambient pollutants had minimal contribution to these responses under
18      the typical summer ambient conditions in Southern California.  However, no unifying
19      conclusions can be reached since each study employed different mixtures and examined different
20      aspects of a response [The complexities ofO3 and co-pollutant exposures in animal studies are
21      discussed in Chapter 5.4.4}.
22
23
24      6.12  CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES
25          A large amount of informative O3 exposure-effects data has been obtained in controlled
26      laboratory exposure studies under a variety of different experimental conditions. However,
27      laboratory simulation of the variable pollutant mixtures present in ambient air is not practical.
28      Thus, the exposure effects of one or several artificially generated pollutants (i.e., a simple
29      mixture) on pulmonary function and symptoms may not explain responses to ambient air where
30      complex pollutant mixtures exist.
31

        January 2005                              6-39       DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.12.1  Mobile Laboratory Studies
 2           Quantitatively useful information on the effects of acute exposure to photochemical
 3      oxidants on pulmonary function and symptoms responses from field studies using a mobile
 4      laboratory were presented in prior criteria documents (U.S. Environmental Protection Agency,
 5      1986, 1996).  Relative to controlled exposure studies, mobile laboratory ambient air studies
 6      suffer the additional limation of a dependence on ambient outdoor conditions.  Consistent with
 7      controlled exposure studies, mobile studies in California demonstrated that pulmonary effects
 8      from exposure to ambient air in Los Angeles are related to O3 concentration and level of
 9      exercise. Healthy subjects with a history of allergy also appeared to be more responsive to O3
10      than "nonallergic" subjects (Linn et al., 1980, 1983b), although a standardized evaluation of
11      atopic status was not performed.
12
13      6.12.2  Aircraft Cabin Studies
14           Respiratory symptoms and pulmonary function effects resulting from exposure to O3 in
15      commercial aircraft flying at high altitudes, and in altitude-simulation studies, have been
16      assessed previously (U.S. Environmental Protection  Agency, 1986, 1996).  Commercial aircraft
17      cabin O3 levels were  reported to be very low (average concentration 0.01 to 0.02 ppm) during
18      92 randomly selected smoking  and nonsmoking flights in 1989 (Nagda et al., 1989). None of
19      these flights recorded O3 concentrations exceeding the 3-h time-weighted average (TWA)
20      standard of 0.10 ppm promulgated by the U.S. Federal Aviation Administration (FAA, 1980),
21      probably due to the use of O3-scrubbing catalytic filters (Melton, 1990).
22           Ozone contamination aboard high-altitude aircraft also has been an interest to the U.S. Air
23      Force because of complaints by crew members of frequent symptoms of dryness and irritation of
24      the eyes, nose, and throat and an occasional cough (Hetrick et al., 2000). Despite the lack of
25      ventilation system modifications as used in commercial aircraft, the O3 concentrations never
26      exceeded the FAA ceiling limit of 0.25 ppm and exceeded the 3-h TWA of 0.10 ppm only 7% of
27      the total monitored flight time (43 h).  The authors concluded that extremely low average
28      relative humidity (12%) during flight operations was most likely responsible for the reported
29      symptoms.
30
31

        January 2005                             6-40       DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.13  SUMMARY
 2           Responses in humans exposed to ambient O3 concentrations include decreased inspiratory
 3      capacity; mild bronchoconstriction; rapid, shallow breathing pattern during exercise; and
 4      symptoms of cough and pain on deep inspiration. Reflex inhibition of inspiration results in a
 5      decrease in forced vital capacity (FVC) and, in combination with mild bronchoconstriction,
 6      contributes to a decrease in the forced expiratory volume in 1 s (FEVj).  In addition to
 7      physiological pulmonary responses and respiratory symptoms, O3 exposure also results in airway
 8      hyperresponsiveness, inflammation, immune system activation, and epithelial injury. With
 9      repeated O3 exposures over several days, spirometric and symptom responses become attenuated,
10      but this tolerance is lost after about a week without exposure. Airway responsiveness also
11      appears to be attenuated with repeated O3 exposures, but less than FEVj.
12           Young healthy adults exposed to O3 concentrations > 0.08 ppm develop significant
13      reversible, transient decrements in pulmonary function if minute ventilation ( VE ) or duration of
14      exposure are increased sufficiently.  O3-induced decrements in FEVj do not appear to depend on
15      gender, race, body surface area, height, lung size, or baseline FVC in young healthy adults.
16      Healthy children experience similar spirometric  responses but lesser symptoms from O3
17      exposure relative to young adults. Beyond the age of 18 to 20 yrs, spirometric and symptom
18      responses to O3 exposure begin to decline with increasing age. There is a large degree of
19      intersubject variability in physiologic and symptomatic responses of heathy adults exposed to
20      O3.  However, responses tend to be reproducible within a given individual over a period of
21      several months. With increasing O3 concentration, the distribution FEVj decrements becomes
22      asymmetrical with a few individuals experiencing large decrements.
23           There is a tendency for slightly increased spirometric responses in mild asthmatics and
24      allergic rhinitics relative to healthy subjects.  Spirometric responses in asthmatics appear to be
25      affected by baseline lung function, i.e., responses increase with disease severity. With repeated
26      daily O3 exposures, spirometric responses of asthmatics become attenuated, however, airway
27      responsiveness becomes increased in subjects with preexisting allergic airway disease (with or
28      without  asthma). Possibly due to patient age,  O3 exposure does not appear to cause significant
29      pulmonary function impairment or evidence of cardiovascular strain in patients with
30      cardiovascular  disease or chronic obstructive pulmonary disease relative to healthy subjects.
        January 2005                               6-41        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Available information on recovery from O3 exposure indicates that an initial phase of
 2      recovery in healthy individuals proceeds relatively rapidly, with acute spirometric and symptom
 3      response appears to occur within about 2 to 4 h.  Small residual lung function effects are almost
 4      completely resolved within 24 hours. Effects of O3 on the small airways, assessed by decrement
 5      in FEF25.75, may be due in part to inflammation. Indeed, a prolonged recovery of residual
 6      spirometric decrements following the initial rapid recovery could be due to slowly resolving
 7      airway inflammation. In hyperresponsive individuals, this recovery takes longer, as much as 48
 8      hours, to return to baseline values.  Persistent spirometry changes observed for up to 48 h
 9      postexposure could plausibly be sustained by the inflammatory mediators.  Cellular responses
10      (e.g., release of immuno-modulatory cytokines) appear to still be active as late as 20 h
11      postexposure.  More slowly developing inflammatory and cellular changes may persist for up to
12      48 h, but the time course in humans has not been explored fully.
13           Soluble mediators of inflammation such as the cytokines (IL-6, IL-8) and arachidonic acid
14      metabolites (e.g., PGE2, PGF2a, thromboxane, and leukotrienes [LTs] such as LTB4) have been
15      measured in the BAL fluid of humans exposed to O3.  Many of these compounds have
16      bronchoconstrictive properties and may be involved in increased airway responsiveness
17      following O3 exposure.  Some indicators of inflammation (e.g., PMN influx, IL-6, PGE2, BAL
18      protein, fibronectin) are attenuated with repeated O3 exposures.  Indicating that tissue damage
19      probably continues to occur during repeated O3 exposure, however, other markers (LDH,  IL-8,
20      total protein, epithelial cells) did not show attenuation. There appears to be no strong correlation
21      between any of the measured cellular and biochemical changes and changes in lung function
22      measurements. Whether airway reactivity  or inflammatory responses to O3 are dependent on the
23      age of the exposed individual, such as spirometric responses, has not been determined.
24           Dietary antioxidant supplementation  attenuates O3-induced spirometric responses but not
25      the intensity of subjective symptoms nor inflammatory responses.  Dietary antioxidants also
26      afforded partial protection to asthmatics by attenuating postexposure bronchial
27      hyperresponsiveness.
28
        January 2005                              6-42        DRAFT-DO NOT QUOTE OR CITE

-------
 1     REFERENCES

 2     Adams, W. C. (2000) Ozone dose-response effects of varied equivalent minute ventilation rates.
 3          J. Exposure Anal. Environ. Epidemiol. 10: 217-226.
 4     Adams, W. C. (2002) Comparison of chamber and face-mask 6.6-hour exposures to ozone on
 5          pulmonary function and symptoms responses. Inhalation Toxicol. 14: 745-764.
 6     Adams, W. C. (2003a) Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm
 7          ozone via square-wave and triangular profiles on pulmonary responses. Inhalation Toxicol.
 8          15:265-281.
 9     Adams, W. C. (2003b) Relation of pulmonary responses induced by 6.6-h exposures to 0.08 ppm
10          ozone and 2-h exposures to 0.30 ppm ozone via chamber and face-mask inhalation.
11          Inhalation Toxicol. 15: 745-759.
12     Adams, W. C.; Ollison, W. M. (1997) Effects of prolonged simulated ambient ozone dosing
13          patterns on human pulmonary function and symptomatology. Presented at: 90th annual
14          meeting of the Air & Waste Management Association; June; Toronto, Ontario, Canada.
15          Pittsburgh, PA: Air & Waste Management Association; paper no. 97-MP9.02.
16     Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation effects on pulmonary function
17          and endurance performance. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55:
18          805-812.
19     Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during
20          continuous exercise via the effective dose concept. J. Appl. Physiol.: Respir. Environ.
21          Exercise Physiol.  51: 415-422.
22     Alexis, N.; Urch, B.; Tarlo,  S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000)
23          Cyclooxygenase metabolites play a different role in ozone-induced pulmonary function
24          decline in asthmatics compared to normals. Inhalation Toxicol. 12: 1205-1224.
25     Alpert, S. E.; Walenga, R. W.; Jaspers, I; Qu, Q.; Chen,  L. C. (1997) Ozone inactivates
26          Cyclooxygenase in human tracheal epithelial  cells without altering PGHS-2 mRNA or
27          protein. Am. J. Physiol. 272: L879-L887.
28     Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner, W. E.; Balmes, J. R. (1993)
29          Ozone-induced airway inflammation in human  subjects as determined by airway lavage
30          and biopsy. Am. Rev. Respir. Dis.  148:  1363-1372.
31     Aris, R. M.; Tager, L; Christian, D.; Kelly, T.; Balmes, J. R. (1995) Methacholine
32          responsiveness is  not associated with O3-induced decreases in FEVj. Chest 107: 621-628.
33     Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N.  (2000) Ozone, but not
34          nitrogen dioxide, exposure decreases glutathione peroxidases in epithelial lining fluid of
35          human lung. Am.  J. Respir. Crit. Care Med.  162: 1342-1347.
36     Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative
37          respiratory effects of ozone and ambient oxidant pollution exposure during heavy exercise.
38          J. Air Pollut. Control Assoc. 34: 804-809.
39     Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J. D. (1985)
40          Comparative effects of laboratory generated ozone and ambient oxidant exposure in
41          continuously exercising subjects. In: Lee, S. D., ed. Evaluation of the scientific basis for
42          ozone/oxidants standards: proceedings of an APCA international specialty conference;
43          November 1984; Houston, TX. Pittsburgh, PA:  Air Pollution Control Association;
44          pp. 216-225. (APCA international specialty conference transactions: TR-4).
       January 2005                             6-43        DRAFT-DO NOT QUOTE OR CITE

-------
 1      Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, I.; Christian, D.; Hearne, P. Q.; Kelly, T.; Aris,
 2           R. M.  (1996) Ozone-induced decrements in FEVj and FVC do not correlate with measures
 3           of inflammation. Am. J. Respir. Crit. Care Med. 153: 904-909.
 4      Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.; Christian, D.;
 5           Kelly, T.; Hearne, P. Q.; Ferrando, R.; Welch, B. (1997) Effects of ozone on normal and
 6           potentially sensitive human subjects, part I: airway inflammation and responsiveness to
 7           ozone in normal and asthmatic subjects. Cambridge, MA: Health Effects Institute.
 8           Research report no. 78; pp 1-37, 81-99.
 9      Bascom, R.; Naclerio, R. M.; Fitzgerald, T. K.; Kagey-Sobotka, A.; Proud, D. (1990) Effect of
10           ozone inhalation on the response to nasal  challenge with antigen of allergic subjects. Am.
11           Rev. Respir. Dis. 142: 594-601.
12      Basha, M. A.; Gross, K. B.; Gwizdala, C.  J.; Haidar, A. H.; Popovich, J., Jr. (1994)
13           Bronchoalveolar lavage neutrophilia in asthmatic and healthy volunteers after controlled
14           exposure to ozone and filtered purified air. Chest 106: 1757-1765.
15      Becker, S.; Madden, M. C.; Newman, S. L.; Devlin, R. B.; Koren, H. S. (1991) Modulation of
16           human alveolar macrophage properties by ozone exposure in vitro. Toxicol. Appl.
17           Pharmacol. 110:403-415.
18      Bedi, J. F.; Drechsler-Parks, D. M.; Horvath, S. M. (1985) Duration of increased pulmonary
19           function sensitivity to an initial  ozone exposure. Am. Ind. Hyg. Assoc. J. 46: 731-734.
20      Bergamaschi, E.; De Palma, G.; Mozzoni, P.; Vanni, S.; Vettori, M. V.; Broeckaert, F.; Bernard,
21           A.; Mutti, A. (2001) Polymorphism  of quinone-metabolizing enzymes and susceptibility to
22           ozone-induced acute effects. Am. J.  Respir. Crit. Care Med. 163: 1426-1431.
23      Blomberg, A.; Helleday,  R.; Pourazar, J.;  Stenfors, N.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.;
24           Sandstrom, T. (1997) Early airway and peripheral blood cell responses to 0.20 ppm ozone
25           in healthy human subjects. Eur.  Respir. J. 10(suppl 25): 274S.
26      Blomberg, A.; Mudway,  IS.; Nordenhall, C.; Hedenstrom,  H.; Kelly, F. J.; Frew, A.  J.; Holgate,
27           S. T.;  Sandstrom, T. (1999) Ozone-induced lung function decrements do not correlate with
28           early airway inflammatory or antioxidant responses. Eur. Respir. J. 13: 1418-1428.
29      Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S.  T.; Kelly, f J.;
30           Sandstrom, T.; Wilson, S.; Frew, A.  J.; Blomberg, A. (2003) Ozone-induced bronchial
31           epithelial cytokine expression differs between healthy and asthmatic subjects. Clin. Exp.
32           Allergy 33: 777-782.
33      Brook, R. D.; Brook, J. R.; Urch, B.; Vincent, R.; Rajagopalan, S.; Silverman, F. (2002)
34           Inhalation  of fine particulate air pollution and ozone causes acute arterial vasoconstriction
35           in healthy adults. Circulation 105: 1534-1536.
36      Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes,  J. R.
37           (1998) Ozone-induced inflammation is attenuated with multiday exposure. Am. J. Respir.
38           Crit. Care Med.  158: 532-537.
39      Coleridge, J. C. G.; Coleridge, H. M.; Schelegle, E. S.; Green, J. F.  (1993) Acute inhalation of
40           ozone stimulates bronchial C-fibers  and rapidly adapting receptors in dogs. J. Appl.
41           Physiol. 74: 2345-2352.
42      Corradi, M.; Alinovi,  R.; Goldoni, M.; Vettori,  M.; Folesani, G.; Mozzoni, P.; Cavazzini, S.;
43           Bergamaschi, E.; Rossi, L.; Mutti, A. (2002) Biomarkers of oxidative stress after
44           controlled  human exposure to ozone. Toxicol. Lett. 134: 219-225.
45      Devlin, R. B.; McDonnell, W. F.; Koren, H. S.; Becker, S. (1990) Prolonged exposure of humans
46           to 0.10 and 0.08 ppm ozone results in inflammation in the lung. Presented at: 83rd annual
        January 2005                              6-44        DRAFT-DO NOT QUOTE OR CITE

-------
 1           meeting of the Air & Waste Management Association; June; Pittsburgh, PA. Pittsburgh,
 2           PA: Air & Waste Management Association; paper no. 90-150.2.
 3     Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.;
 4           Koren, H. S. (1991) Exposure of humans to ambient levels of ozone for 6.6 hours causes
 5           cellular and biochemical changes in the lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
 6     Devlin, R. B.; McKinnon, K. P.; Noah, T.; Becker, S.; Koren, H.  S. (1994) Ozone-induced
 7           release of cytokines and fibronectin by alveolar macrophages and airway epithelial cells.
 8           Am. J. Physiol. 266: L612-L619.
 9     Devlin, R. B.; McDonnell, W. F.; Becker, S.; Madden, M. C.; McGee, M. P.; Perez, R.; Hatch,
10           G.; House, D. E.; Koren, H. S. (1996) Time-dependent changes of inflammatory mediators
11           in the lungs of humans exposed to 0.4 ppm ozone for 2 hr: a comparison of mediators
12           found in bronchoalveolar lavage fluid 1 and 18 hr after exposure. Toxicol. Appl.
13           Pharmacol. 138: 176-185.
14     Devlin, R. B.; Folinsbee, L. J.; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.;
15           Koren, H. S. (1997) Inflammation and cell damage induced by repeated exposure of
16           humans to ozone. Inhalation Toxicol. 9: 211-235.
17     Dimeo, M. J.; Glenn, M. G; Holtzman, M. J.; Shelter, J. R.; Nadel, J. A.; Boushey, H. A. (1981)
18           Threshold concentration of ozone causing an increase in bronchial reactivity in humans
19           and adaptation with repeated exposures. Am. Rev. Respir. Dis. 124: 245-248.
20     Fahy, J. V.; Wong, H. H.; Liu, J. T.; Boushey, H. A. (1995) Analysis of induced sputum after air
21           and ozone exposures in healthy subjects. Environ. Res. 70: 77-83.
22     Federal Aviation Administration. (1980) Airplane cabin ozone contamination. F. R. (January 21)
23           45: 3880-3883.
24     Folinsbee, L. J.; Hazucha, M. J. (1989) Persistence of ozone-induced changes in lung function
25           and airway responsiveness. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;  Grant, L. D.,
26           eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
27           US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam,
28           The Netherlands: Elsevier  Science Publishers; pp. 483-492. (Studies in environmental
29           science 35).
30     Folinsbee, L. J.; Hazucha, M. J. (2000) Time course of response to ozone exposure in healthy
31           adult females. Inhalation Toxicol. 12: 151-167.
32     Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure. Aviat.
33           Space Environ. Med. 57: 1136-1143.
34     Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1977) Decrease of maximum work performance
35           following ozone exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 42:
36           531-536.
37     Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978) The influence of exercise
38           on the pulmonary function changes due to exposure to low concentrations of ozone. In:
39           Folinsbee, L. J.; Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F.,
40           eds. Environmental  stress:  individual human adaptations. New York, NY: Academic Press;
41           pp. 125-145.
42     Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly
43           exposed to low concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
44     Folinsbee, L. J.; McDonnell, W. F.; Horstman, D. H. (1988) Pulmonary function and symptom
45           responses after 6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38:
46           28-35.
       January 2005                             6-45        DRAFT-DO NOT QUOTE OR CITE

-------
 1     Folinsbee, L. I; Horstman, D. H.; Kehrl, H. R.; Harder, S.; Abdul-Salaam, S.; Ives, P. J. (1994)
 2           Respiratory responses to repeated prolonged exposure to 0.12 ppm ozone. Am. J. Respir.
 3           Crit. Care Med. 149: 98-105.
 4     Folinsbee, L. J.; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam, S.; Koren, H. S.
 5           (1998) Repeated exposure of humans to ozone: pulmonary function and symptom
 6           responses. Research Triangle Park, NC: U.S. Environmental Protection Agency; National
 7           Center for Environmental Assessment; unpublished data.
 8     Foster, W. M.; Stetkiewicz,  P. T. (1996) Regional clearance of solute from the respiratory
 9           epithelia:  18-20 h postexposure to  ozone. J. Appl. Physiol. 81: 1143-1149.
10     Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters tracheobronchial
11           mucociliary function in humans. J. Appl. Physiol. 63: 996-1002.
12     Foster, W. M.; Wills-Karp, M.; Tankersley, C. G.; Chen, X.; Paquette, N. C. (1996) Bloodborne
13           markers in humans during multiday exposure to ozone. J. Appl. Physiol. 81: 794-800.
14     Foster, W. M.; Brown, R. H.; Macri, K.; Mitchell, C. S. (2000) Bronchial reactivity of healthy
15           subjects: 18-20 h postexposure to ozone. J. Appl. Physiol. 89: 1804-1810.
16     Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993) Enhanced response to ozone
17           exposure during the follicular phase of the menstrual cycle. Environ. Health Perspect.  101:
18           242-244.
19     Foxcroft, W.  J.; Adams, W.  C. (1986) Effects of ozone exposure on four consecutive days on
                                 •
20           work performance and VO2max. J. Appl. Physiol. 61: 960-966.
21     Frampton, M. W.; Morrow,  P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.;
22           Speers, D. M. (1997) Ozone responsiveness in smokers and nonsmokers. Am. J. Respir.
23           Crit. Care Med. 155: 116-121.
24     Frank, R.; Liu, M. C.; Spannhake, E. W.; Mlynarek, S.; Macri, K.; Weinmann, G. G. (2001)
25           Repetitive ozone exposure of young adults: evidence of persistent small airway
26           dysfunction. Am.  J. Respir. Crit. Care Med. 164: 1253-1260.
27     Freed, A. N.; Chou, C. L.; Fuller, S. D.;  Croxton, T. L. (1996) Ozone-induced vagal reflex
28           modulates airways reactivity in rabbits. Respir. Physiol. 105:  95-102.
29     Gabrielson, E. W.; Yu, X.-Y.; Spannhake, E. W. (1994) Comparison of the toxic effects of
30           hydrogen  peroxide and ozone on cultured human bronchial epithelial cells. Environ. Health
31           Perspect.  102: 972-974.
32     Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled
33           particles measured at 2 h after ozone exposure in humans. J. Appl. Physiol. 74: 2984-2989.
34     Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired exercise
35           performance and pulmonary function in elite cyclists during low-level ozone exposure in a
36           hot environment. Am.  Rev. Respir. Dis. 134: 726-733.
37     Gong, H., Jr.; Bedi, J. F.; Horvath, S. M. (1988) Inhaled albuterol does not protect against ozone
38           toxicity in nonasthmatic athletes. Arch. Environ. Health 43: 46-53.
39     Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997a) Responses of older men
40           with and without chronic obstructive pulmonary disease to prolonged ozone exposure.
41           Arch. Environ. Health 52: 18-25.
42     Gong, H., Jr.; McManus, M. S.; Linn, W. S. (1997b) Attenuated response to repeated daily
43           ozone exposures in asthmatic subjects. Arch. Environ. Health 52: 34-41.
44     Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K.
45           R.; Prasad, S. B. (1998) Cardiovascular effects of ozone exposure in human volunteers.
46           Am. J. Respir. Crit. Care Med. 158: 538-546.
        January 2005                              6-46       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Graham, D. E.; Koren, H. S. (1990) Biomarkers of inflammation in ozone-exposed humans:
 2           comparison of the nasal and bronchoalveolar lavage. Am. Rev. Respir. Dis. 142: 152-156.
 3      Graham, D.; Henderson, F.; House, D. (1988) Neutrophil influx measured in nasal lavages of
 4           humans exposed to ozone. Arch. Environ. Health 43: 228-233.
 5      Gwynn, R. C.; Thurston, G. D. (2001) The burden of air pollution: impacts among racial
 6           minorities. Environ. Health Perspect. 109(suppl. 4): 501-506.
 7      Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential
 8           involvement of 4-hydroxynonenal in the response of human lung cells to ozone. Am. J.
 9           Physiol. 274: L8-L16.
10      Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993) Ozone-induced increases
11           in substance P and 8-epi-prostaglandin F2(X in the airways of human subjects. Am. J. Respir.
12           Cell Mol. Biol. 9: 568-572.
13      Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone
14           concentration profiles on pulmonary function. Am. Rev. Respir. Dis. 146: 1487-1493.
15      Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P.  A. (1994) Lung function response of
16           healthy women after sequential exposures to NO2  and O3. Am. J. Respir. Crit. Care Med.
17           150: 642-647.
18      Hazucha, M. J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.;
19           Bromberg, P. A. (1996) Effects of cyclo-oxygenase inhibition on ozone-induced
20           respiratory inflammation and lung function changes. Eur. J. Appl. Physiol. Occup. Med.
21           73: 17-27.
22      Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003) Distribution and reproducibility of
23           spirometric response to ozone by gender and age.  J. Appl. Physiol. 95: 1917-1925.
24      Hetrick,  S. M.;  Gould, W. D.;  Christensen, D. E.  (2000) Inflight cabin ozone aboard long
25           duration C-5 airlift missions: a historical issue revisited. Aviat.  Space Environ.  Med. 71:
26           408-414.
27      Hiltermann, T.  J. N.; Stolk, J.; Hiemstra, P. S.; Fokkens, P. H. B.; Rombout, P. J. A.;  Sont, J. K.;
28           Sterk, P. J.; Dijkman, J. H. (1995) Effect of ozone exposure on maximal airway narrowing
29           in non-asthmatic and asthmatic subjects. Clin. Sci. 89:  619-624.
30      Hiltermann, T.  J. N.; de Bruijne, C. R.; Stolk, J.; Zwinderman, A. H.; Spieksma, F. Th. M.;
31           Roemer, W.; Steerenberg, P. A.; Fischer, P. H.; van Bree, L.; Hiemstra, P. S. (1997)
32           Effects of photochemical air pollution and allergen exposure on upper respiratory tract
33           inflammation in asthmatics. Am. J. Respir. Crit. Care Med. 156: 1765-1772.
34      Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg,  P. A.; Brahim, J. J.;  Sont, J. K.;
35           Sterk, P. J.; Hiemstra, P. S.; Stolk, J. (1999) Ozone-induced inflammation assessed in
36           sputum and bronchial lavage fluid from asthmatics: a new noninvasive tool in
37           epidemiologic studies on air pollution and asthma. Free Radical Biol. Med. 27:  1448-1454.
38      Holz, O.; Torres, R. A.; Timm, P.; Mucke, M.; Richter, K.; Koschyk, S.; Magnussen,  H. (1999)
39           Ozone-induced airway inflammatory changes differ between individuals and are
40           reproducible. Am.  J. Respir. Crit. Care Med. 159:  776-784.
41      Holz, O.; Mucke, M.; Paasch,  K.; Bohme, S.; Timm, P.; Richter,  K.; Magnussen, H.;  Torres, R.
42           A.  (2002) Repeated ozone exposures enhance bronchial allergen responses in subjects with
43           rhinitis or asthma.  Clin. Exp. Allergy. 32: 681-689.
44      Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone
45           concentration and pulmonary response relationships for 6.6-hour exposures with five hours
46           of moderate exercise to 0.08, 0.10, and0.12ppm.  Am. Rev. Respir. Dis. 142: 1158-1163.
        January 2005                              6-47       DRAFT-DO NOT QUOTE OR CITE

-------
 1     Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T.; Folinsbee, L. J. (1995) Comparison of
 2           pulmonary responses of asthmatic and nonasthmatic subjects performing light exercise
 3           while exposed to a low level of ozone. Toxicol. Ind. Health 11: 369-385.
 4     Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979) Pulmonary function and maximum
 5           exercise responses following acute ozone exposure. Aviat. Space Environ. Med. 50:
 6           901-905.
 7     Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of effect.
 8           Am. Rev. Respir. Dis. 123: 496-499.
 9     Housley, D. G.; Eccles, R.; Richards, R. J. (1996) Gender difference in the concentration of the
10           antioxidant uric acid in human nasal lavage. Acta Oto-Laryngol. 116: 751-754.
11     Jenkins, H. S.; Devalia, J. L.; Mister, R. L.; Bevan, A. M.; Rusznak, C.; Davies, R. J.  (1999) The
12           effect of exposure to ozone and nitrogen dioxide on the airway response of atopic
13           asthmatics to inhaled allergen: dose- and time-dependent effects. Am. J. Respir. Crit. Care
14           Med. 160:  33-39.
15     load, J. P.; Kott, K. S.; Brie, J. M. (1996) The local C-fiber contribution to ozone-induced effects
16           on the isolated guinea pig lung. Toxicol. Appl. Pharmacol. 141: 561-567.
17     Torres, R.; Nowak, D.; Magnussen, H.; Speckin, P.; Koschyk, S. (1996) The effect of ozone
18           exposure on allergen responsiveness in subjects with asthma or rhinitis. Am. J. Respir.
19           Crit. Care Med. 153: 56-64.
20     Torres, R. A.; Holz, O.; Zachgo, W.; Timm, P.;  Koschyk, S.; Miiller, B.; Grimminger, F.; Seeger,
21           W.; Kelly, F. J.; Dunster, C.; Frischer, T.; Lubec, G.; Waschewski, M.; Niendorf, A.;
22           Magnussen, H. (2000) The effect of repeated ozone exposures on inflammatory markers in
23           bronchoalveolar lavage fluid and mucosal biopsies. Am. J. Respir. Crit. Care Med. 161:
24           1855-1861.
25
26     Kehrl, H. R.; Hazucha, M. J.; Solic, J. J.; Bromberg, P. A. (1985) Responses of subjects with
27           chronic obstructive pulmonary disease after exposures to 0.3 ppm ozone. Am. Rev. Respir.
28           Dis. 131:719-724.
29     Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.;
30           Bromberg, P. A. (1987) Ozone exposure increases respiratory epithelial permeability in
31           humans. Am. Rev. Respir. Dis. 135: 1124-1128.
32     Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased
33           specific airway reactivity of persons with mild allergic asthma after 7.6 hours of exposure
34           to 0.16 ppm ozone. J. Allergy. Clin. Immunol. 104: 1198-1204.
35     Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in
36           rats exposed to ozone, alone and in mixtures, with acidic aerosols. Inhalation Toxicol.  11:
37           249-264.
38     Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.;
39           Kozumbo, W. J.; Becker, S.; House, D. E.; McDonnell, W. F.; Bromberg, P. A.  (1989a)
40           Ozone-induced inflammation in the lower airways of human subjects. Am. Rev. Respir.
41           Dis. 139: 407-415.
42     Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The
43           inflammatory response in human lung exposed to ambient levels of ozone. In: Schneider,
44           T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research  and its
45           policy implications: proceedings of the 3rd US-Dutch international symposium; May 1988;
46           Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
47           pp. 745-753.  (Studies in environmental science 35).


       January 2005                             6-48       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent
 2           changes of markers associated with inflammation in the lungs of humans exposed to
 3           ambient levels of ozone. Toxicol. Pathol. 19: 406-411.
 4      Kozumbo, W. 1; Hanley, N. M.; Agarwas, S.; Thomas, M.  I; Madden, M. C. (1996) Products of
 5           ozonized arachidonic acid potentiate the formation of DNA single strand breaks in cultured
 6           human lung cells. Environ. Mol. Mutagen. 27: 185-195.
 7      Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. I; D'Arcy, I; Eschenbacher, W. L. (1989)
 8           Ozone-induced changes in pulmonary function and bronchial responsiveness in asthmatics.
 9           J. Appl. Physiol. 66: 217-222.
10      Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G.; Frew, A.;
11           Polak, J.; Holgate, S. (1997a) Effects of ozone on epithelium and sensory nerves in the
12           bronchial mucosa of healthy humans. Am. J. Respir. Crit. Care Med. 156: 943-950.
13      Krishna, M. T.; Blomberg, A.; Biscione, G. L.; Kelly, F.; Sandstrom, T.; Frew,  A.; Holgate, S.
14           (1997b) Short-term ozone exposure upregulates P-selectin in normal human airways. Am.
15           J. Respir. Crit. Care Med. 155: 1798-1803.
16      Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D.  M. (1982)
17           Duration of pulmonary function adaptation to ozone in humans. Am. Ind.  Hyg. Assoc. J.
18           43: 832-837.
19      Kulle, T. J.; Milman, J. H.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Miller, W. R. (1984)
20           Pulmonary function adaptation to ozone in  subjects with chronic bronchitis. Environ. Res.
21           34: 55-63.
22      Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone response relationships in
23           healthy nonsmokers.  Am. Rev. Respir. Dis. 132: 36-41.
24      Lang, D. S.; Torres, R. A.; Miicke, M.; Siegfried, W.; Magnussen, H. (1998) Interactions
25           between human bronchoepithelial cells and lung fibroblasts after ozone exposure in vitro.
26           Toxicol. Lett. 96-97: 13-24.
27      Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U.  (1988) Die Wirkung von Ozon auf die
28           korperliche Leistungsfahigkeit [The effect of ozone on physical activity].  Schweiz Z.
29           Sportmed. 36: 5-10.
30      Linn, W. S.; Jones, M. P.; Bachmayer, E. A.; Spier, C. E.; Mazur, S. F.; Avol, E. L.; Hackney, J.
31           D. (1980) Short-term respiratory effects of polluted ambient air: a laboratory study of
32           volunteers in a high-oxidant community. Am. Rev. Respir. Dis. 121: 243-252.
33      Linn, W. S.; Fischer, D. A.; Medway,  D. A.; Anzar, U. T.; Spier, C. E.; Valencia, L. M.; Venet,
34           T. G.; Hackney, J. D. (1982a) Short-term respiratory effects of 0.12 ppm ozone exposure in
35           volunteers with chronic obstructive pulmonary disease. Am. Rev. Respir.  Dis. 125:
36           658-663.
37      Linn, W. S.; Medway, D. A.; Anzar, U. T.; Valencia, L. M.; Spier, C. E.; Tsao,  F. S.-D.; Fischer,
38           D. A.; Hackney, J. D. (1982b) Persistence of adaptation to ozone in volunteers exposed
39           repeatedly for six weeks. Am. Rev.  Respir. Dis. 125: 491-495.
40      Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Spier, C. E.; Valencia, L. M.; Anzar, U. T.; Hackney,
41           J. D. (1983a) Response to ozone in volunteers with chronic obstructive pulmonary disease.
42           Arch. Environ. Health 38: 278-283.
43
44      Linn, W. S.; Avol, E. L.; Hackney, J. D. (1983b) Effects of ambient oxidant pollutants on
45           humans: a movable environmental chamber study. In: Lee, S. D.; Mustafa, M. G.;
46           Mehlman, M. A., eds. International symposium on the biomedical effects  of ozone and
47           related photochemical oxidants;  March 1982; Pinehurst, NC. Princeton, NJ: Princeton


        January 2005                              6-49        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Scientific Publishers, Inc.; pp. 125-137. (Advances in modern environmental toxicology: v.
 2           5).
 3     Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Fischer, D.
 4           A.; Hackney, J. D. (1986) A dose-response study of healthy, heavily exercising men
 5           exposed to ozone at concentrations near the ambient air quality standard. Toxicol. Ind.
 6           Health 2:  99-112.
 7     Linn, W. S.; Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Avol, E. L. (1997) Chamber
 8           exposures of children to mixed ozone, sulfur dioxide, and sulfuric acid. Arch. Environ.
 9           Health 52: 179-187.
10     Liu, L.; Leech,  J. A.; Urch, R. B.; Silverman, F.  S. (1997) In vivo salicylate hyroxylation: a
11           potential biomarker for assessing acute ozone exposure and effects in humans. Am. J.
12           Respir. Crit. Care Med. 156: 1405-1412.
13     Liu, L.; Leech,  J. A.; Urch, R. B.; Poon, R.; Zimmerman, B.; Kubay, J. M.; Silverman, F. S.
14           (1999) A  comparison of biomarkers of ozone exposure in human plasma, nasal lavage, and
15           sputum. Inhalation Toxicol. 11: 657-674.
16     Marthan, R.; Roux, E.; Savineau, J.-P. (1996) Human bronchial smooth muscle responsiveness
17           after in vitro exposure to oxidizing pollutants. Cell Biol. Toxicol. 12: 245-249.
18     McBride, D. E.; Koenig, J. Q.; Luchtel, D. L.; Williams, P. V.; Henderson, W. R., Jr. (1994)
19           Inflammatory effects of ozone in the upper airways of subjects with asthma. Am. J. Respir.
20           Crit. Care Med.  149: 1192-1197.
21     McDonnell, W. F. (1996) Individual variability in human lung function responses to ozone
22           exposure. Environ. Toxicol. Pharmacol. 2: 171-175.
23     McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.;
24           House, D. E. (1983) Pulmonary effects of ozone exposure during exercise: dose-response
25           characteristics. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol.  54: 1345-1352.
26     McDonnell, W. F., Ill; Horstman, D. H.; Abdul-Salaam, S.; House, D. E. (1985a)
27           Reproducibility of individual responses to  ozone exposure. Am. Rev. Respir. Dis. 131:
28           36-40.
29     McDonnell, W. F., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier, A. M. (1985b)
30           Respiratory responses of vigorously exercising children to 0.12 ppm ozone exposure. Am.
31           Rev. Respir. Dis. 132: 875-879.
32     McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.; Folinsbee, L. J.; Devlin, R. B.;
33           O'Neil, J.  J.; Horstman, D. H. (1991) Respiratory response of humans exposed to low
34           levels of ozone for 6.6 hours. Arch. Environ. Health  46: 145-150.
35     McDonnell, W. F.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual
36           differences in acute response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
37     McDonnell, W. F.; Stewart, P. W.; Andreoni, S.; Seal, E.,  Jr.; Kehrl, H. R.; Horstman, D. H.;
38           Folinsbee, L. J.; Smith, M. V.  (1997) Prediction of ozone-induced FEVj changes: effects of
39           concentration, duration, and ventilation. Am. J. Respir. Crit. Care Med. 156: 715-722.
40     McDonnell, W. F.; Stewart, P. W.; Smith, M. V.; Pan, W.  K.; Pan, J. (1999) Ozone-induced
41           respiratory symptoms: exposure-response models and association with lung function. Eur.
42           Respir. J.  14: 845-853.
43     McGee, M. P.;  Devlin, R.; Saluta, G.; Koren, H. (1990) Tissue factor and factor VII messenger
44           RNAs in human alveolar macrophages: effects of breathing ozone. Blood 75: 122-127.
45     McKenzie, D. C.; Stirling, D. R.; Fadl, S.; Allen, M. (1987) The effects of salbutamol on
46           pulmonary function in cyclists exposed to ozone: a pilot study. Can.  J. Sport Sci. 12: 46-48.
        January 2005                              6-50       DRAFT-DO NOT QUOTE OR CITE

-------
 1      McKinnon, K. P.; Madden, M. C.; Noah, T. L.; Devlin, R. B. (1993) In vitro ozone exposure
 2           increases release of arachidonic acid products from a human bronchial epithelial cell line.
 3           Toxicol. Appl. Pharmacol. 118: 215-223.
 4      Melton, C. E. (1990) Airliner cabin ozone: an updated review. Washington, DC: Federal
 5           Aviation Administration, Office of Aviation Medicine; report no. DOT/FAA/AM-89/13.
 6           Available from: NTIS,  Springfield, VA; AD-A219 264.
 7      Messineo, T. D.; Adams, W. C. (1990) Ozone inhalation effects in females varying widely in
 8           lung size: comparison with males. J. Appl. Physiol. 69: 96-103.
 9      Michelson, P. H.; Dailey, L.; Devlin, R. B.; Peden, D. B. (1999) Ozone effects on the
10           immediate-phase response to allergen in the nasal airways of allergic asthmatic subjects.
11           Otolaryngol. Head Neck Surg. 120: 225-232.
12      Mohammed, S. P.; Higenbottam, T. W.; Adcock, J. J. (1993) Effects of aerosol-applied
13           capsaicin, histamine and prostaglandin-E2 on airway sensory receptors of anaesthetized
14           cats. J. Physiol. Lond. 46: 951-966.
15      Molfmo, N. A.; Wright,  S. C.; Katz, L; Tarlo, S.; Silverman, F.; McClean, P. A.; Szalai, J. P.;
16           Raizenne, M.; Slutsky, A. S.; Zamel, N. (1991) Effect of low concentrations of ozone on
17           inhaled allergen responses in asthmatic subjects.  Lancet 338(8761): 199-203.
18      Mudway, I. S.; Kelly, F. J. (2000)  Ozone and the lung: a sensitive issue. Mol. Aspects. Med. 21:
19           1-48.
20      Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S.  T.;  Sandstrom, T.; Kelly, F. J. (1999)
21           Antioxidant consumption and repletion kinetics in nasal lavage fluid following exposure of
22           healthy human volunteers to ozone. Eur. Respir. J. 13:  1429-1438.
23      Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.;  Marklund, S. L.; Frew, A.
24           J.; Sandstrom, T.; Kelly, F. J. (2001) Differences in basal airway antioxidant
25           concentrations are not predictive of individual responsiveness to ozone: a comparison of
26           healthy and mild asthmatic subjects. Free Radical Biol. Med. 31: 962-974.
27      Nagda, N. L.; Fortmann, R. C.; Koontz, M. D.; Baker,  S. R.; Ginevan, M. E. (1989) Airliner
28           cabin environment: contaminant measurements, health risks, and mitigation options.
29           Washington, DC: U.S. Department of Transportation, Office of the Secretary.  Available
30           from: NTIS, Springfield, VA; PB91-159384.
31      Nayak, A. S. (2003) The asthma and allergic rhinitis link.  Allergy and Asthma Proc.
32           24: 395-403.
33      Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud, D. J.;  Speizer, F. E. (1995) The
34           association of ambient air pollution with twice daily peak expiratory flow rate
35           measurements in children. Am. J. Epidemiol.  141: 111-122.
36      Newson, E. J.; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000)
37           Effects of short-term exposure to 0.2 ppm ozone  on biomarkers of inflammation in sputum,
38           exhaled nitric oxide, and lung function in subjects with mild atopic asthma. J. Occup.
39           Environ. Med. 42:  270-277.
40      Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P. J. (2000) No effect of inhaled
41           budesonide on the  response to inhaled ozone in normal subjects.  Am. J. Respir. Crit. Care
42           Med. 161:479-486.
43      Nikasinovic, L.; Momas, L; Seta, N. (2003) Nasal epithelial and inflammatory response to ozone
44           exposure: a review of laboratory-based studies published since 1985.  J. Toxicol. Environ.
45           HealthB 6: 521-568.
        January 2005                              6-51       DRAFT-DO NOT QUOTE OR CITE

-------
 1     Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998)
 2           Nociceptive mechanisms modulate ozone-induced human lung function decrements. J.
 3           Appl. Physiol. 85:  1863-1870.
 4     Peden, D. B. (2001a) Air pollution in asthma: effect of pollutants on airway inflammation. Ann.
 5           Allergy Asthma Immunol. 87(suppl. 3): 12-17.
 6     Peden, D. B. (200Ib) Effect of pollutants in rhinitis. Curr. Allergy Asthma Rep. 1: 242-246.
 7     Peden, D. B.; Setzer, R. W., Jr.; Devlin, R. B. (1995) Ozone exposure has both a priming effect
 8           on allergen-induced responses and an intrinsic inflammatory action in the nasal airways of
 9           perennially allergic asthmatics. Am. J. Respir. Crit. CareMed. 151:  1336-1345.
10     Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997) Prolonged acute exposure to 0.16
11           ppm ozone induces eosinophilic airway inflammation in asthmatic subjects with allergies.
12           J. Allergy Clin. Immunol. 100: 802-808.
13     Riediker,  M.; Monn, C.; Roller, T.; Stahel, W. A.; Wuthrich, B. (2001) Air pollutants enhance
14           rhinoconjunctivitis symptoms in pollen-allergic individuals. Ann. Allergy Asthma
15           Immunol. 87: 311-318.
16     Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption
17           in the lung: effects of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch.
18           Environ. Health 52: 173-178.
19     Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L;
20           Estela del Rio-Navarro, B.; Hernandez-Avila, M.; London, S. J. (2004) Genetic
21           polymorphism ofGSTMl and antioxidant supplementation influence lung function in
22           relation to ozone exposure in asthmatic children in Mexico City.  Thorax 59: 8-10.
23     Samet, J. M.; Hatch, G. E.; Horstman, D.;  Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine,
24           M.;  McDonnell, W. F.; Devlin, R. B. (2001) Effect of antioxidant supplementation on
25           ozone-induced lung injury in human subjects. Am. J. Respir. Crit. Care Med.  164: 819-825.
26     Scannell,  C.; Chen, L.; Aris, R. M.; Tager, L; Christian, D.; Ferrando, R.;  Welch, B.; Kelly, T.;
27           Balmes, J. R. (1996) Greater ozone-induced inflammatory responses in subjects with
28           asthma. Am. J. Respir. Crit. Care Med. 154: 24-29.
29     Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise time in competitive simulations
30           consequent to low level ozone exposure. Med. Sci.  Sports Exercise 18: 408-414.
31     Schelegle, E. S.; Adams, W. C.;  Siefkin, A. D. (1987) Indomethacin pretreatment reduces
32           ozone-induced pulmonary function decrements in human subjects. Am. Rev. Respir. Dis.
33           136: 1350-1354.
34     Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced
35           neutrophilia in normal humans. Am. Rev. Respir. Dis. 143: 1353-1358.
36     Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G.; Green, J. F. (1993)
37           Contribution of vagal afferents to respiratory reflexes evoked by  acute inhalation of ozone
38           in dogs. J. Appl. Physiol. 74: 2338-2344.
39     Schelegle, E. S.; Eldridge, M. W.; Cross, C. E.; Walby, W. F.; Adams, W. C. (2001) Differential
40           effects  of airway anesthesia on ozone-induced pulmonary responses in human subjects.
41           Am. J. Respir. Crit. CareMed. 163: 1121-1127.
42     Schonfeld, B. R.; Adams, W. C.; Schelegle, E. S. (1989) Duration of enhanced responsiveness
43           upon re-exposure to ozone. Arch. Environ. Health 44: 229-236.
44     Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S.  (1976)
45           Pulmonary responses of rats to ambient levels of ozone: effects of 7-day intermittent or
46           continuous exposure. Lab. Invest. 34: 565-578.
        January 2005                              6-52       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. I; Butler, S. O.; Green,
 2          J.; Raggio, L. (1993) The pulmonary response of white and black adults to six
 3          concentrations of ozone. Am. Rev. Respir. Dis.  147: 804-810.
 4      Seal, E., Jr.; McDonnell, W. F.; House, D. E. (1996) Effects of age, socioeconomic status, and
 5          menstrual cycle on pulmonary response to ozone. Arch. Environ. Health 51: 132-137.
 6      Seltzer, J.; Bigby, B. G.; Stulbarg, M.; Holtzman, M.  J.; Nadel, J. A.; Ueki, I. F.; Leikauf, G D.;
 7          Goetzl, E. J.; Boushey, H. A. (1986) O3-induced change in bronchial reactivity to
 8          methacholine and airway inflammation in humans. J. Appl. Physiol. 60: 1321-1326.
 9      Solic, J. J.; Hazucha, M. J.; Bromberg, P. A. (1982) The acute effects of 0.2 ppm ozone in
10          patients with chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 125: 664-669.
11      Solway, J.; Left, A. R. (1991) Sensory neuropeptides  and airway function. J. Appl. Physiol. 71:
12          2077-2087.
13      Spannhake, E. W.; Reddy, S. P. M.; Jacoby, D. B.;  Yu, X.-Y.; Saatian, B.; Tian, J. (2002)
14          Synergism between rhinovirus infection and oxidant pollutant exposure enhances airway
15          epithelial cell cytokine production. Environ. Health Perspect. 110: 665-670.
16      Stenfors, N.; Pourazar,  J.; Blomberg, A.; Krishna, M.  T.; Mudway, I; Helleday, R.; Kelly, F. J.;
17          Frew, A. J.; Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in
18          asthmatic and healthy subjects. Respir. Med. 96: 352-358.
19      Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. F.; Hatch, G. E. (1989) Unattenuated
20          structural and biochemical alterations in the rat lung during functional adaptation to ozone.
21          Am. Rev. Respir. Dis. 140: 493-501.
22      Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M. G. (1990) Measurements of
23          cardiopulmonary response in awake rats during  acute exposure to near-ambient
24          concentrations of ozone. Fundam. Appl. Toxicol. 10: 7-15.
25      Torres, A.; Utell, M. J.; Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R. J.; Speers,
26          D. M.; Tsai, Y.; Frampton, M. W. (1997) Airway inflammation in smokers and
27          nonsmokers with varying responsiveness to ozone. Am. J. Respir. Crit. Care Med.  156:
28          728-736.
29      Trenga, C. A.; Koenig,  J. Q.; Williams, P.  V. (2001) Dietary antioxidants and ozone-induced
30          bronchial hyperresponsiveness in adults with  asthma. Arch. Environ. Health 56: 242-249.
31      U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other
32          photochemical oxidants. Research Triangle Park, NC: Office of Health and Environmental
33          Assessment, Environmental Criteria  and Assessment Office; report nos.
34          EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA; PB87-142949.
35      U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related
36          photochemical oxidants. Research Triangle Park, NC: Office of Research and
37          Development; report nos. EPA/600/AP-93/004aF-cF. 3v. Available from: NTIS,
38          Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
39          www.epa.gov/ncea/ozone.htm.
40      Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs:
41          intersubject variability in physiologic response.  Boston, MA: Health Effects Institute.
42      Vagaggini, B.; Carnevali, S.; Macchioni, P.; Taccola,  M.; Fornai, E.; Bacci, E.; Bartoli, M. L.;
43          Cianchetti, S.; Dente, F. L.; Di Franco, A.; Giannini, D.; Paggiaro. P. L. (1999) Airway
44          inflammatory response to ozone in subjects with different asthma severity. Eur. Respir. J.
45          13:274-280.
46      Vagaggini, B.; Taccola, M.; Conti, L; Carnevali,  S.; Cianchetti, S.; Bartoli, M. L.; Bacci, E.;
47          Dente, F. L.; Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2001) Budesonide reduces


        January 2005                              6-53        DRAFT-DO NOT QUOTE OR CITE

-------
 1           neutrophilic but not functional airway response to ozone in mild asthmatics. Am. J. Respir.
 2           Crit. Care Med. 164: 2172-2176.
 3      Vagaggini, B.; Taccola, M.; Clanchetti, S.; Carnevali, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.;
 4           Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2002) Ozone exposure increases eosinophilic
 5           airway response induced by previous allergen challenge. Am. J. Respir. Crit. Care Med.
 6           166: 1073-1077.
 7      Voter, K. Z.; Whitin, J. C.; Torres, A.; Morrow, P. E.; Cox, C.; Tsai, Y.; Utell, M. J.; Frampton,
 8           M. W. (2001) Ozone exposure and the production of reactive oxygen species by
 9           bronchoalveolar cells in humans. Inhalation Toxicol. 13: 465-483.
10      Wang, G.; Umstead,  T. M.; Phelps, D. S.;  Al-Mondhiry, H.; Floros, J. (2002) The effect of ozone
11           exposure on the ability of human surfactant protein A variants to stimulate cytokine
12           production. Environ. Health Perspect. 110: 79-84.
13      Weinmann, G. G.; Weidenbach-Gerbase, M.; Foster, W. M.; Zacur, H.; Frank, R. (1995)
14           Evidence for ozone-induced small-airway dysfunction: lack of menstrual-cycle and gender
15           effects. Am. J. Respir. Crit. Care Med. 152: 988-996.
16      Weymer, A. R.; Gong, H., Jr.; Lyness, A.; Linn, W.  S. (1994) Pre-exposure to ozone does not
17           enhance or produce exercise-induced asthma. Am. J. Respir. Crit. Care Med. 149:
18           1413-1419.
19      Yeadon, M.; Wilkinson, D.; Darley-Usmar, V.; O'Leary, V. J.; Payne, A. N. (1992) Mechanisms
20           contributing to  ozone-induced bronchial hyperreactivity in guinea-pigs. Pulm. Pharmacol.
21           5:39-50.
22      Ying, R. L.; Gross, K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990) Indomethacin does not
23           inhibit the ozone-induced increase in bronchial responsiveness in human subjects. Am.
24           Rev. Respir. Dis.  142: 817-821.
25      Yu, M.; Pinkerton, K. E.; Witschi, H. (2002)  Short-term exposure to aged and diluted sidestream
26           cigarette smoke enhances ozone-induced lung injury in B6C3F1 mice. Toxicol. Sci. 65:
27           99-106.
28      Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced
29           airways hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
30
31
        January 2005                              6-54        DRAFT-DO NOT QUOTE OR CITE

-------
 i         ANNEX AX6.  CONTROLLED HUMAN EXPOSURE
 2                   STUDIES OF OZONE AND RELATED
 3                       PHOTOCHEMICAL OXIDANTS
 4
 5
 6     AX6.1    INTRODUCTION
 7          In the previous chapter, results of ozone (O3) studies in laboratory animals and in vitro test
 8     systems were presented. The extrapolation of results from animal studies is one mechanism by
 9     which information on potential adverse human health effects from exposure to O3 is obtained.
10     More direct evidence of human health effects due to O3 exposure can be obtained through
11     controlled human exposure studies of volunteer subjects or through field and epidemiologic
12     studies of populations exposed to ambient O3.  Controlled human exposure studies, discussed in
13     this chapter, typically use fixed concentrations of O3 under carefully regulated environmental
14     conditions and subject activity levels.
15          Most of the scientific information selected for review and evaluation in this chapter comes
16     from the literature published since  1996 which, in addition to further study of physiological
17     pulmonary responses and respiratory  symptoms, has focused on mechanisms of inflammation
18     and cellular responses to injury induced by O3 inhalation.  Older studies are discussed where
19     only limited new data are available and where new and old data are conflicting. The reader is
20     referred to both the  1986 and 1996 Air Quality Criteria documents (U.S. Environmental
21     Protection Agency,  1986, 1996) for a more extensive discussion of older studies. Summary
22     tables of the relevant O3 literature are included for each of the major subsections.
23     In summarizing the  human health effects literature, changes from control are described if
24     statistically significant at a probability (p) value less than 0.05, otherwise trends are noted
25     as such.
26
       January 2005                            AX6-1      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.2  PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE IN
 2              HEALTHY SUBJECTS
 3     AX6.2.1  Introduction
 4           The responses observed in young healthy non-smoking human adults exposed to ambient
 5     O3 concentrations include decreased inspiratory capacity; mild bronchoconstriction; rapid,
 6     shallow breathing pattern during exercise; and symptoms of cough and pain on deep inspiration.
 7     In addition, O3 has been shown to result in airway hyperresponsiveness as demonstrated by an
 8     increased physiological response to a nonspecific bronchoconstrictor, as well as airway injury
 9     and inflammation assessed via bronchoalveolar lavage and biopsy.  Reflex inhibition of
10     inspiration and consequent decrease in inspiratory capacity results in a decrease in forced vital
11     capacity (FVC) and total lung capacity (TLC) and, in combination with mild
12     bronchoconstriction, contributes to a decrease in the forced expiratory volume in 1 s (FEVj).
13     Given that both FEVj and FVC are subject to decrease with O3 exposures, changes in the ratio
14     (FEVj/FVC) become difficult to interpret and so are not discussed.
15           The majority of controlled human studies have investigated the effects of exposure to
16     variable O3 concentrations in healthy subjects performing continuous exercise (CE) or
17     intermittent exercise (IE) for variable periods of time.  These studies have several important
18     limitations: (1) the ability to study only short-term, acute effects; (2) the inability to link short-
19     term effects with long-term consequences; (3) the use of a small number of volunteers that may
20     not be representative of the general population; and (4) the statistical limitations associated with
21     the small sample size.  Nonetheless, studies reviewed in the 1996 EPA criteria document (U.S.
22     Environmental Protection Agency, 1996) provided a large body of data describing the effects
23     and dose-response characteristics of O3 as function of O3 concentration (C), minute ventilation
24     (VE), and duration or time (T) of exposure.  In most of these studies, subjects were exposed to
25     O3 and to filtered air (FA [reported as 0 ppm O3]) as a control. The  most salient observations
26     from these studies were: (1) healthy subjects exposed to O3 concentrations > 0.08 ppm develop
27     significant reversible, transient decrements in pulmonary function if VE or T are increased
28     sufficiently, (2) there is a large degree of intersubject variability in physiologic and symptomatic
29     responses to O3 and these responses tend to be reproducible within a given individual over a
30     several months period, and (3) subjects exposed repeated to O3 over several days develop a
       January 2005                            AX6-2       DRAFT-DO NOT QUOTE OR CITE

-------
 1     tolerance to successive exposures, as demonstrated by an attention of responses, which is lost
 2     after about a week without exposure.
 3          In this section, the effects of single O3 exposures of 1- to 8-h in duration on pulmonary
 4     function in healthy nonsmoking subjects are examined by reviewing studies that investigate:
 5     (1) the O3 exposure-response relationship; (2) intersubject variability, individual sensitivity, and
 6     the association between responses; and (3) mechanisms of pulmonary function responses and the
 7     relationship between tissue-level events and functional responses. Discussion will largely be
 8     limited to studies published subsequent to the 1996 EPA criteria document (U.S. Environmental
 9     Protection Agency, 1996)
10
11     AX6.2.2  Acute Ozone Exposures for Up to 2 Hours
12          At-Rest Exposures. Exposure studies investigating the effects of O3 exposures on sedentary
13     subjects were discussed in the 1986 EPA criteria document (U.S. Environmental Protection
14     Agency, 1986).  The lowest O3 concentration at which significant reductions in FVC and FEVj
15     were reported was 0.5 ppm (Folinsbee et al., 1978; Horvath et al., 1979). Averaging there results
16     of these two studies and correcting for FA responses, exposing resting young adults (n=23,
17     age=22) to 0.5 ppm O3 results in an -4%  reduction in FVC and an -7% reduction FEVj.  At
18     lower O3 concentrations of 0.25 to 0.3 ppm, resting exposures did not significantly affect lung
19     function.
20          Exposures with Exercise. Collectively, the studies reviewed in the 1996 EPA criteria
21     document (U.S. Environmental Protection Agency, 1996) demonstrated that healthy young
22     adults performing moderate to heavy IE or CE of 1 to 2.5 h duration, exposed to 0.12 to
23     0.18 ppm O3 experienced statistically significant decrements in pulmonary function and
24     respiratory symptoms. As an example, 2  hr exposures to  0.12 and 0.18 ppm O3 during heavy IE
25     (exercise VE = 65 L/min) have resulted in FEVj decrements of 2.0 ± 0.8% (mean ± SE; n = 40)
26     and 9.5 ± 1.1% (n = 89), respectively (McDonnell and Smith,  1994).  Significant decrements in
27     pulmonary function have been reported in heavily exercising healthy adults exposed for 1 h with
28     CE at O3 concentrations of 0.12 ppm (Gong et al., 1986),  0.16 ppm (Avol et al., 1984), and
29     0.2 ppm (Adams and Schelegle, 1983; Folinsbee et al., 1984).
30          In an attempt to describe O3 dose-response characteristics, many investigators modeled
31     acute responses as a function of total inhaled O3 dose (C * T x VE), which was found to be a

       January 2005                             AX6-3       DRAFT-DO NOT QUOTE OR CITE

-------
 1      better predictor of response than O3 concentration, VE, or T of exposure, alone. In an analysis of
 2      6 studies with 1 to 2 h exposures to between 0.12 and 0.18 ppm O3 with exercise, Folinsbee et al.
 3      (1988) reported a good correlation (r = 0.81) between total inhaled O3 dose and FEVj
 4      decrements. For a given exposure duration, total inhaled O3 dose can be increased by increases
 5      in C and/or VE . In exposures of fixed duration, results of several studies suggested that O3
 6      concentration was a more important predictor of response or explained more of the variability in
 7      response than VE  (Adams et al., 1981; Folinsbee et al, 1978; Hazucha, 1987).  Based on a review
 8      of previously published studies, Hazucha (1987) noted that relative to the FEVj decrement
 9      occurring at a given C and VE, doubling C (e.g., from 0.1 to 0.2 ppm) would increase the FEVj
10      decrement by 400%, whereas doubling the  VE (e.g., from an exercise VE of 20 to 40 L/min)
11      which would only increase the FEVj decrement by 190%. Thus, C appears to have a greater
12      affect than VE on FEVj responses even when total inhaled O3 doses are equivalent.
13           New studies (i.e., not reviewed in the 1996 EPA criteria document) that provide
14      spirometric responses for up to 2 h exposures are summarized in Table AX6-1. Most of these
15      newer studies have investigated mechanisms affecting responses, inflammation, and/or effects in
16      diseased groups versus healthy adults, accordingly their findings may be summarized differently
17      in several sections of this chapter.  Rather than a FA exposure, some of these studies use O3
18      exposures with placebo as a control. Studies appearing in Table 1, but not discussed in this
19      section, are discussed in other sections of this chapter as indicated within the table.
20           McDonnell  et al.  (1997) pooled the results of eight studies entailing 485 healthy male
21      subjects exposed for 2 h on one occasion to one of six O3 concentrations (0.0, 0.12, 0.18, 0.24,
22      0.30, or 0.40 ppm) at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA).  FEVj
23      was measured preexposure, after 1 h of exposure, and immediately postexposure. Decrements  in
24      FEVj were modeled by sigmoid-shaped curve as a function of subject age, O3 concentration, VE,
25      and T. The modeled decrements reach a plateau with increasing T and dose rate (C * VE). That
26      is, for a given O3 concentration, exercise VE level, and after a certain length of exposure, the
27      FEVj response tends not to  increase further with increasing duration of exposure.  The modeled
28      FEVj responses increased with C x VE  and T, decreased with subject age, but were only
29      minimally affected by body size corrections to VE. Fitted and experimental FEVj decrements

        January 2005                             AX6-4      DRAFT-DO NOT QUOTE OR CITE

-------
                   Table AX6-1. Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours during Exercise3
to
o
o
X
H

6
o


o
H

O
Ozone
Concentration11
ppm
0.0
0.4




0.0
0.2


0.0
0.2




0.0
0.12
0.18
0.24
0.30
0.40

0.4






ug/m3
0
784




0
392


0
392




0
235
353
471
589
784

784






Exposure Duration
and Activity
2hIE
4x15 min
on bicycle,
VE = 30 L/min


2hIE
4x15 min
at VE = 20
L/min/m2 BSA
2hIE
4x15 min
at VE = 20
L/min/m2 BSA


2 h rest or IE
(4x15 min
at VE = 25 or 35
L/min/m2 BSA)



2hIE
20 min mild-mod.
exercise,
1 0 min rest



Number and
Exposure Gender of Subject
Conditions' Subjects Characteristics
NA 5 M, 4 F Healthy adults
25 ± 2 years old

6 M, 7 F Mild atopic asthmatics
22 ± 0.7 years old

20 °C 8 M, 5 F Healthy NS
50% RH median age 23 years


20 °C 10 M, 12 F Healthy NS
50% RH mean age 24 years




22 °C 485 M (each Healthy NS
40% RH subject exposed 1 8 to 36 years old
at one activity mean age 24 years
level to one O3
concentration)


NA 4 M, 5 F Healthy NS
30 ± 3 years old





Observed Effect(s)
O3-induced reductions in FVC (12%, 10%) and FEVj (13%, 11%)
for asthmatic and healthy subjects. Significant reductions in mid-
flows in both asthmatics and healthy subjects. Indomethacin
pretreatment significantly decreased FVC and FEV; responses to
O3 in healthy but not asthmatic subjects. See Section AX6.3. 2 and
Tables AX6-3 andAX6-13.
Median O3-induced decrements of 70 mL, 190 mL, and 400 mL/s
in FVC, FEVj, and FEF25.75, respectively. Spirometric responses
not predicted of inflammatory responses. See Sections AX6.2. 5. 2,
AX6.5.6, andAX6.9.3 andTableAX6-12.
Significant O3-induced decrement in FEVj immediately post-
exposure but not significantly different from baseline 2 h later.
No correlation between Clara cell protein (CC16) and FEV;
decrement. CC16 levels, elevated by O3 exposure, remained
high at 6 h post-exposure, but returned to baseline by 18 h
postexposure. See Table AX6-12
Statistical analysis of 8 experimental chamber studies conducted
between 1980 and 1993 by the U.S. EPA in Chapel Hill, NC.
Decrement in FEVj described by sigmoid-shaped curve as a
function of subject age, O3 concentration, VB and time. Response
decreased with age, was minimally affected by body size
corrections, and was not more sensitive to O3 concentration
than VE. Also see Section AX6.5
Subjects previously in Nightingale et al. (2000) study. Placebo-
control: Immediately postexposure decrements in FVC (9%) and
FEVj (14%) relative to pre-exposure values. FEVj decrement
only 9% at 1 hr postexposure. By 3 h postexposure, recovery in
FVC to 97% and FEV; to 98% of preexposure values. Significant
increases in 8-isoprostane at 4 h postexposure. Budesonide for
2 wk prior to exposure did not affect responses.
Reference
Alexis et al.
(2000)




Blomberg
etal. (1999)


Blomberg
et al. (2003)




McDonnell
etal. (1997)





Montuschi
et al. (2002)





o
HH
H
W

-------
                    Table AX6-1 (con't). Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours during Exercise"
to
o
o
X
Oi
H

6
o


o
H

O
Ozone
Concentration11
ppm ug/m3 and Activity Conditions'
0.0 392 2hIE 20 °C
0.2 4xl5min 50% RH
at VE = 20
L/min/m2 BSA

0.4 784 2hIE NA
20 min mild-mod.
exercise,
10 min rest



0.0 784 2hIE 21 °C
0.4 4x15 min at 40% RH
VE =18 L/min/m2
BSA

2 exposures:
25% subjects
exposed to air-air,
75% to O3-O3
0.0 784 2hIE 20 °C
0.4 4x15 min 40% RH
at VE = 20
L/min/m2 BSA


0.0 490 1 h CE NA
0.25 VE = 30 L/min Face mask
exposure


Number and
Gender of
Subjects
6M, 9F


9M, 6F

6M, 9F






Weak
re spenders
7 M, 13F

Strong
re spenders
21 M, 21 F


Placebo group
15 M, 1 F

Antioxidant
group
13 M, 2F
32 M, 28 F




Subject
Characteristics
Healthy adults
24 years old

Mild asthmatics
29 years old
Healthy NS
mean age ~3 1 years






Healthy NS
20 to 59 years old







Healthy NS
mean age 27 years



Healthy NS
22.6 ± 0.6 years old



Observed Effect(s)
O3-induced FEVl decrement (8%, healthy adults; 3% asthmatics)
and PMN increase (20.6%, healthy adults; 15.2% asthmatics).
Primary goal was to investigate relationship between antioxidant
defenses and O3 responses in asthmatics and healthy adults.
See Tables AX6-3 andAX6-13.
Placebo-control: O3 caused significant decrements in FEV;
(13.5%) and FVC (10%) immediately following exposure,
a small increase in Mch-reactivity, and increased PMNs and
myeloperoxidase in induced sputum at 4 h postexposure. FEVj at
96% and FVC at 97% preexposure values at 3 h postexposure.
Budesonide for 2 wk prior to exposure did not affect spirometric
responses. See Section AX6. 2. 5 and Table AX6-13
Significant O3-induced decrements in spirometric lung function.
Young adults (< 35 years) were significantly more responsive
than older individuals (> 35 years). Sufentanil, a narcotic
analgesic, largely abolished symptom responses and improved
FEVj in strong responders. Naloxone, an opioid antagonist,
did not affect O3 effects in weak responders. See Section
AX6.2.5.1


Placebo and antioxidant groups had O3-induced decrements in
FEVj (20 and 14%) and FVC (13 and 10%), respectively.
Percent neutrophils and IL-6 levels in BAL fluid obtained 1 h
post exposure were not different in the two treatment groups.
See Table AX6-13.

Mean O3-induced FEVj decrements of 15.9% in males and 9.4%
in females (gender differences not significant). FEV; decrements
ranged from -4 to 56%; decrements >15% in 20 subjects and
>40% in 4 subjects. Uptake of O3 greater in males than females,
but uptake not correlated with spirometric responses.
Reference
Mudway etal.
(2001)
Stenfors etal.
(2002)

Nightingale
et al. (2000)





Passannante
etal. (1998)







Samet et al.
(2001)
Steck-Scott
et al. (2004)


Ultman etal.
(2004)



o
HH
H
W
"See Appendix A for abbreviations and acronyms.

bListed from lowest to highest O3 concentration.

'Studies conducted in exposure chamber unless otherwise indicated.

-------
1
2
3
4
5
6
7
following a 2 h exposure at three nominal levels ofVE are illustrated in Figure AX6-1 as a
function of O3 concentration. Their analysis indicated that C was marginally, but not
significantly more important than VE  in predicting FEVj response. Additionally, the McDonnell
et al. (1997) analysis revealed that some prior analyzes of IE protocols may have over estimated
the relative importance of C over VE in predicting FEVj responses by considering only the VE
during exercise and ignoring the VE during periods of rest.
                   24-
                   20-
                •£  16-
                E
                «  12-|
                o
                I   8H
               HI
                    4-
                    0-
                   -4-
                       0.0
                             T = 2h
                            0.1
0.2         0.3
  O3 (ppm)
                                                                       Ve = 40 L/min (n)
                                                                       Ve = 31 L/min («)
                                                                       Ve = 10 L/min (A)
0.4
0.5
      Figure AX6-1. FEVt decrements as a function of O3 concentration following a 2 h
                     exposure with incremental exercise (15 min intervals) or rest. Points are
                     experimental data (mean ± SE) and lines are model predictions for each
                                                       •
                     activity level. Minute ventilation (VE) represent average across intervals
                     of rest and exercise.
      Source: McDonnell et al. (1997).
      January 2005
                                        AX6-7
             DRAFT-DO NOT QUOTE OR CITE

-------
 1          Ultman et al. (2004) measured O3 uptake and pulmonary responses in 60 young heathy
 2     non-smoking adults (32 M, 28 F). A bolus technique was used to quantify the uptake of O3 as a
 3     function of the volume into the lung which the bolus penetrated. From these measurements, the
 4     volumetric depth at which 50% uptake occurred was calculated. This volumetric lung depth was
 5     correlated with conducting airways volume, i.e., a greater fraction of O3 penetrated to deeper into
 6     the lungs of individuals have larger conducting airways volumes. Two weeks after the bolus
 7     measurements, subjects were exposed via a face mask to FA and subsequently two weeks later to
 8     0.25 ppm O3 for 1 h with CE at a target VE of 30 L/min. The breath-by-breath uptake of O3 was
 9     measured.  There was a small but significant reduction in the breath-by-breath uptake of O3 from
10     90.6% on average for the first 15 minutes to 87.3% on average for the last 15 minutes of
11     exposure.  The uptake fraction was significantly greater in males (91.4%) than females (87.1%),
12     which is consistent with the larger fB and smaller VT of the females than males. Uptake was not
13     correlated with spirometric responses. However, there was tendency for males to have greater
14     O3-induced FEVj decrements than females, 15.9% versus 9.4%, respectively.  There was
15     considerable intersubject variability in FEVj decrements which ranged from -4 to 56% with 20
16     subjects having decrements of >15% and 4 subjects with >40% decrements (see Section AX6.4
17     for additional discussion regarding intersubject variability).
18
19     AX6.2.3  Prolonged Ozone Exposures
20          Between 1988 and 1994, a number studies were completed that described the responses of
21     subjects exposed to relatively low (0.08 to 0.16 ppm) O3 concentrations for exposure durations
22     of 4 to 8 h. These studies were discussed in the 1996 criteria document (U.S.  Environmental
23     Protection Agency, 1996) and only a select few are briefly discussed here. Table AX6-2 details
24     newer studies of healthy subjects undergoing prolonged exposures at O3 concentrations ranging
25     from 0.06 to 0.20 ppm. In most  of these studies, statistically significant changes in pulmonary
26     function, symptoms, and airway  responsiveness have been observed during and after exposures
27     to O3 concentrations of 0.08 ppm and higher.  As with studies conducted at higher O3
28     concentrations for shorter periods of time, there is considerable intersubject variability  in
29     response (see Section AX6.4).
30          Folinsbee et al. (1988) first reported the  effects of a 6.6 h exposure to 0.12 ppm O3 in ten
31     young healthy adults (25 ± 4 yr)  with quasi continuous exercise that was intended to simulate a

       January 2005                            AX6-8       DRAFT-DO NOT QUOTE OR CITE

-------
                           Table AX6-2. Pulmonary Function Effects after Prolonged Exposures to Ozone"
to
o
o
X
H

6
o


o
H

O
Ozone Concentration11 Exposure

ppm
Studies with
0.18


0.0
0.20



0.2



0.0
0.24




Studies with
0.0
0.04
0.08
0.12

0.12








ug/m3 and Activity
4 hr Exposures
353 4hIE
(4 x 50 min)
VE = 35L/min
0 4hIE
392 (4 x 50 min cycle
ergometry or
treadmill running
[VE = 40L/min])
392 4 h IE
(4 x 50 min)
VE = 25 L/min/m2
BSA
0 4hIE
470 (4x15 min)
VE =20 L/min



>6 hr Exposures
0 6.6 h
78 IE (6 x 50min)
157 VE = 20 L/min/m2
235 BSA

235 3 day-6.6h/day
IE (6 x 50 min)
VE = 17 L/min/m2,
20 L/min/m2
BSA, and 23
L/min/m2 BSA


Number and
Exposure Gender of Subject
Conditions Subjects Characteristics

23 °C 2 M, 2 F Adults NS, 21
50% RH to 33 years old

20 °C FA: 1 1 M, 3 Adult NS, 19 to
50% RH F 41 years old
03: 9M, 3F


20 °C 42 M, 24 F Adults NS,
50% RH 18 to 50 years
old

24°C 10 M Healthy NS,
40% RH 60 to 69 years
9 M COPD
59 to 71 years



23 °C 15 M, 15 F Healthy NS,
50% RH 22.4 ± 2.4 yrs
old


23 °C 1 5 M, 1 5 F Healthy NS, 1 8
50% RH to 31 years old








Observed Effect(s)

FVC decreased 19% and FEVj decreased 29% in these four
pre-screened sensitive subjects.

Decrease in FVC, FEVj, VT, and SRaw and increase in fB with
O3 exposure compared with FA; total cell count and LDH
increased in isolated left main bronchus lavage and inflammatory
cell influx occurred with O3 exposure compared to FA exposure.

FEVj decreased by 18.6%; Pre-exposure methacholine
responsiveness was weakly correlated with the functional response
to O3 exposure. Symptoms were also weakly correlated with the
FEVj response (r = -0.31 to -0.37)
Healthy: small, 3.3%, decline in FEVj (p=0.03 [not reported in
paper], paired-t on O3 versus FA pre-post FEVj). COPD: 8%
decline in FEV; (p=ns, O3 versus FA). Adjusted for exercise,
ozone effects did not differ significantly between COPD patients
and healthy subjects.
See Section AX6. 5.1.

FEVj and total symptoms at 6.6 h exposure to 0.04 ppm not
significantly different from FA. FEVj (-6.4%) and total symptoms
significant at 6.6 h exposure to 0.08 ppm. FEV; (-15.4%) at 6.6 h
not significantly different between chamber and face mask
exposure to 0.12 ppm.
FEVj at 6.6 h decreased significantly by 9.3%, 11.7%, and 13.9%,
respectively at three different exercise VE rates, but were not
significantly different from each other. Total symptoms at the
highest VE protocol were significantly greater than for the lowest
VE protocol beginning at 4.6 h. Largest subjects (2.2 m2 BSA)
had significantly greater average FEVj decrement for the three
protocols, 18.5% compared to the smallest subjects (1.4 m2
BSA), 6.5%.


Reference

Adams
(2000a)

Aris et al.
(1993)



Aris et al.
(1995)


Gong et al.
(1997a)





Adams (2002)




Adams
(2000b)






o
HH
H
W

-------
                               Table AX6-2 (cont'd). Pulmonary Function Effects after Prolonged Exposures to Ozonea
to
o
o
X
 I
o
Ozone Concentration11
ppm
(a) 0.08
(b) 0.08
(mean) varied
from 0.03 to
0.15

(a) 0.08



(b) 0.30



(a) 0.12

(b)0.12
(mean) varied
from 0.07 to
0.16
ug/m3
235
235
(mean)



157



588



235

235
(mean)


Exposure
Duration
and Activity
6.6 h
IE (6 x 50 min)
VE = 20 L/min/m2
BSA


6.6 h
IE (6 x 50 min)
VE = 20 L/min/m2
BSA
2h
IE (4 x 15 min)
VE = 35 L/min/m2
BSA
6.6 h IE
(6 x 50 min)
(a,b,c) VE = 20
L/min/m2 BSA
(d)VE=12
L/min/m2 BSA
Number and
Exposure Gender of Subject
Conditions Subjects Characteristics
23 °C 15 M Healthy NS,
50% RH 15 F 18 to 25 years
old



23 °C 15 M Healthy NS,
50% RH 15 F 18 to 25 years
old





23 °C 6 M, 6 F Healthy NS,
50% RH 19 to 25 years
old



Observed Effect(s)
(a) FEVj decreased 6.2% after 6.6 h in square-wave exposures.
Total symptoms significantly increased at 5.6 and 6.6 h.
(b) FEV; decreased 5.6 to 6.2% after 4.6 to 6.6 h, respectively,
in varied exposure; total symptoms significantly increased also
after 4.6 to 6.6 h. No significant difference between face mask
and chamber exposures.
Significantly greater FEV; decrement (12.4%) for 2-h, 0.30 ppm
exposure than for 6.6-h, 0.08 ppm exposure (3.6%).






(a) FEVj decreased 11% at 6.6 h in square-wave exposure.
Total symptoms significant from 4.6 to 6.6 h.

(b) FEV; decreased 13% at 6.6 h; not significantly different from
square-wave exposure. Total symptoms significant from 4.6 to
6.6 h.
Reference
Adams
(2003a)






Adams
(2003b)




Adams and
Ollison(1997)




H
6
o
o
H
O
         (c)0.12          235
         (mean) varied    (mean)
         from 0.11 to
         0.13
         (d)0.12
235
(c) FEVj decreased 10.3% at 6.6 h; not significantly different from
square-wave exposure. Total symptoms significant from 4.6 to
6.6 h.

(d) FEV; decreased 3.6% at 6.6 h; significantly less than for
20 L/min/m2 BSA protocols.
         "See Appendix A for abbreviations and acronyms.
         bListed from lowest to highest O3 concentration.
o
HH
H
W

-------
 1      full workday of heavy physical labor.  Except for a 35-min lunch break after 3 h, the subjects
 2      exercised at a moderate level ( VE « 40 L/min) for 50 min of each hour.  Ignoring the lunch
 3      break during which lung function did not change appreciably, approximately linear decreases
 4      were observed in FVC, FEVl3 and FEV25_75 with duration of O3 exposure. Correcting for FA
 5      responses, decrements of 8.2, 14.9, and 26.8% in FVC, FEVl5 and FEV25.75 occurred as a result
 6      of the O3 exposure. Using the same 6.6 h protocol, but a lower O3 concentration of 0.08 ppm,
 7      Horstman et al. (1990) and McDonnell et al. (1991) observed decrements corrected for FA (and
 8      averaged across studies) of 5, 8, and 11% in FVC, FEVb and FEV25_75, respectively, in 60 young
 9      adults (25 ± 5 years old). Horvath et al. (1991) observed a 4% (p = 0.03)1 decrement in FEVj
10      using the forementioned protocol (i.e., 6.6 h and 0.08 ppm O3) in 11 healthy adults (37 ± 4 yr).
11      The smaller decrement observed by Horvath et al.  (1991) versus Horstman et al. (1990) and
12      McDonnell  et al. (1991) is consistent with response decreasing as subject age increases (see
13      Section AX6.5.1}.
14
15      AX6.2.3.1   Effect of Exercise Ventilation Rate on FEVt Response to 6.6 h Ozone Exposure
16          It is well known that response to O3 exposure is a function of VE in studies of 2 h or less in
17      duration (See Section AX6.2.2). It is reasonable to expect that response to a prolonged 6.6-h O3
18      exposure is  also function of VE, although quantitative analyzes are  lacking.
19          In an attempt to quantify this effect, Adams and Ollison (1997) exposed 12 young adults to
20      an average O3 concentration of 0.12 ppm for 6.6 h at varied exercise VE .  They observed a mean
21      FEVj decrements of 10 to 11% in two protocols having a mean exercise VE of 33 L/min and a
22      14% decrement in a protocol with a mean exercise VE  of 36 L/min. These FEVj decrements
23      were significantly greater than the average decrement of 3.6% (not  significantly different from
24      FA response) observed at an exercise VE of only 20 L/min. In a subsequent study of 30 healthy
25      adults (Adams, 2000b), the effect of smaller exercise VE differences on pulmonary function  and
26      symptoms responses to 6.6 h exposure to 0.12 ppm O3 was examined. FEVj decrements of 9.3,
27      11.7, and 13.9% were observed for the exercise VE of 30.2, 35.5, and 40.8 L/min, respectively.
28      Along with the tendency for FEVj responses to increase with VE, total symptoms severity was
              'Based on two-tailed paired t-test of data in Table III of Horvath et al. (1991).

        January 2005                            AX6-11      DRAFT-DO NOT QUOTE OR CITE

-------
 1     found to be significantly greater at the end of the highest VE protocol relative to the lowest VE
 2     protocol. Although the FEVj responses were not significantly different from each other, the
 3     power of the study to detect differences between the three VE was not reported and no analysis
 4     was performed using all of the data (e.g., a mixed effects model). Data from the Adams and
 5     Ollison (1997) and Adams (2000b) studies are illustrated in Figure AX6-2 with data from three
 6     older studies. There are a paucity of data below an exercise VE of 30 L/min. Existing data for
 7     exposure to 0.12 ppm O3 suggests that FEVj responses increase with increasing exercise VE until
 8     at least 35 L/min.
 9
10     AX6.2.3.2  Exercise Ventilation Rate as a Function of Body/Lung Size on FEVt Response
11                to 6.6 h Ozone Exposure
12           Typically, with the assumption that the total inhaled O3 dose should be proportional to the
13     lung size of each individual, exercise VE in 6.6 h exposures has been set as a multiple of body
14     surface area (BSA) (McDonnell et al., 1991) or as a product of eight times FVC (Folinsbee et al.,
15     1988; Frank et al., 2001; Horstman et al., 1990). Utilizing previously published data, McDonnell
16     et al. (1997) developed a statistical model analyzing the effects of O3 concentration, VE, duration
17     of exposure, age, and body and lung size on FEVj response. They concluded that any effect of
18     BSA, height, or baseline FVC on percent decrement in FEVj in this population of 485 young
19     adults was small if it exists at all.  This is consistent with Messineo and Adams (1990),  who
20     examined pulmonary function responses in young adult women having small (n = 14) or large
21     (n = 14) lung sizes (mean FVC  of 3.74 and 5.11 L, respectively). Subject were exposed to
22     0.30 ppm O3 for 1 h with CE (VE = 47 L/min).  There was no significant difference between the
23     group FEVj decrements (22.1 and 25.6% for small and large lung, respectively). In addition,
24     Messineo and Adams (1990) also did a retrospective analysis of 36 young adult males who each
25     had completed similar 1 h exposures to 0.30 ppm O3 with CE (VE « 70 L/min) and found lung
26     size was not real ted with FEVj response.
27           Adams (2000b) studied a group of 30 young adult men and women exposed to 0.12 ppm
28     O3 for 6.6 h on three occasions while exercising 50 min of each hour at one of three different VE
29     levels (viz., 17,  20, and 23 l/min/m2BSA). Their postexposure FEVj responses were regressed
30     as a function of BSA (which was directly related to the absolute amount of VE during exercise

       January 2005                            AX6-12     DRAFT-DO NOT QUOTE OR CITE

-------
                         25
                      & 20
                      0)
                      E
                      S>  15
                      o
                      0)
                      Q
                      £10
                      LU
                      LL
                          5-
• Adams (2000b)
D Adams and Ollison (1997)
A Folinsbee et al. (1988)
A Folinsbee et al. (1994)
• Horstman et al. (1990)
                            15       20      25       30      35
                                          Minute Ventilation (L/min)
                                       40
                    45
      Figure AX6-2.  Average FEVt decrements (±SE) for prolonged 6.6 h exposures to 0.12 ppm
                                               •
                     O3 as a function of exercise VE. Since age affects response to O3 exposure,
                     selected studies had subjects with mean ages between 22 and 25 years.
                     FEVj decrements were calculated as mean O3 responses minus mean air
                     responses.  Unless provided in papers, SE were estimated from variability
                     in post O3 exposure responses. In one case, the SE for VE of 33 L/Min
                     (10.3% decrement) was taken as the SE of data from protocol with VE of
                     33 L/min (11% decrement). All studies used a constant 0.12 ppm O3
                     exposure except two (*) which used 0.115 ppm O3 for hours 1-2 and 5-6
                     and 0.13 ppm O3 for hours  3-4 of exposure.
1     and, thus, primarily responsible for individual differences in total inhaled O3 dose).  The slope
2     was significantly different from zero (p = 0.01), meaning that the smallest subjects, who had the
3     lowest exercise VE (~ 26 L/min), had a lower FEVj decrement (-5%) than the largest subjects
4     (-17%), whose exercise VE was -44 L/min. This relationship was not a gender-based
5     difference, as the mean female's FEVj decrement was -11.2%, which was not significantly
6     different from the male's -12.2% mean value. Similarly, when total symptoms severity
7     response was regressed against BSA, the slope was significantly different than zero (p = 0.0001),
8     with lower values for smaller subjects than for larger subjects.  Results of this study suggest that
      January 2005
               AX6-13
DRAFT-DO NOT QUOTE OR CITE

-------
 1     for the O3 concentration and exposure duration used, responses are more closely related to VE
 2     than VE normalized to BSA.  Further, this observation is in agreement with McDonnell et al.
 3     (1997), who observed no evidence that measurements of lung or body size were significantly
 4     related to FEVj response in 2 h IE exposures.  These authors state that the absence of an
 5     observed relationship between FEVj response and BSA, height, or FVC may be due to the poor
 6     correlation between these variables and airway caliber (Collins et al., 1986; Martin et al., 1987).
 7     Also, the O3 dosimetry study of Bush et al. (1996) indicated that normalization of the O3 dose
 8     would be more appropriately applied as a function of anatomic dead space.
 9
10     AX6.2.3.3 Comparison of 6.6 h Ozone Exposure Pulmonary Responses to Those Observed
11                in 2 h Intermittent Exercise Ozone Exposures
12           It has been shown that greater O3 concentration (Horstman et al.,  1990) and higher VE
13     (Adams, 2000b) each elicit greater FEVj response in prolonged, 6.6-h exposures, but data on the
14     relative effect of O3 concentration, VE, and T in prolonged exposures are very limited and have
15     not been systematically compared to data from shorter (< 2-h) exposures. In a recent study
16     (Adams, 2003b), the group mean FEVj response for a 2-h IE exposure to 0.30 ppm O3 was
17     -12.4%, while that for a 6.6-h exposure to 0.08 ppm O3 was -3.5%. The total inhaled  O3 dose
18     (as the simple product of C XT  x VE ) was 1358 ppm-L for the 2-h exposure and 946 ppm-L for
19     the 6.6-h exposure.  Thus, the FEVj decrement was 3.5 times greater and the total inhaled O3
20     dose was 1.44 times greater for the 2-h exposure compared to the 6.6-h exposure.  This
21     difference illustrates the limitations  of utilizing the concept of total  O3 dose for comparisons
22     between studies of vastly different exposure durations.
23           Adams (2003b) also examined whether prolonged 6.6 h exposure to a relatively low O3
24     concentration (0.08 ppm) and the 2-h IE exposure at a relatively high O3 concentration (0.30
25     ppm) elicited consistent individual subject effects, i.e, were those most or least affected in one
26     exposure also similarly affected in the other? Individual subject O3 exposure reproducibility was
27     first examined via a regression plot of the postexposure FEVj response to the 6.6-h chamber
28     exposure as a function of postexposure FEVj response to  the 2-h chamber exposure. The R2 of
29     0.40, although statistically  significant, was substantially less than that observed in a comparison
30     of individual FEVj response to two 2-h IE exposures by chamber and face mask, respectively
31     (R2 = 0.83).  The Spearman rank order correlation for the  chamber 6.6-h and chamber 2-h

       January 2005                            AX6-14      DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposure comparison was also substantially less (0.49) than that obtained for the two 2-h
 2      exposures (0.85). The primary reason for the greater variability in the chamber 6.6-h exposure
 3      FEVj response as a function of that observed for the two 2-h IE exposures is very likely related
 4      to the increased variability in response upon repeated exposure to O3 concentrations lower than
 5      0.18 ppm (R = 0.57, compared to a mean R of 0.82 at higher concentrations) reported by
 6      McDonnell et al. (1985a).  This rationale is supported by the lower R (0.60) observed by Adams
 7      (2003b) for the FEVj responses found in 6.6 h chamber and face mask exposures to 0.08 ppm
 8      O3, compared to an R of 0.91 observed for responses found at 0.30 ppm O3.
 9
10      AX6.2.4  Triangular Ozone Exposures
11          To further explore the factors that determine responsiveness to O3, Hazucha et al. (1992)
12      designed a protocol to examine the effect of varying, rather than constant,  O3 concentrations.
13      In this study, subjects were exposed to a constant level of 0.12 ppm O3 for 8 h and to an O3 level
14      that increased linearly from 0 to 0.24 ppm  for the first 4 h and then decreased linearly from
15      0.24 to 0 over the second 4 h of the 8 h exposure (triangular concentration profile). Subjects
16      performed moderate exercise (VE -40 L/min) during the first 30 minutes of each hour. The total
17      inhaled O3 dose (i.e., C x T x VE ) for the constant versus the triangular concentration profile was
18      almost identical. FEVj responses are illustrated in Figure AX6-3.  With exposure to the constant
19      0.12 ppm O3, FEVj declined approximately 5% by the fifth hour of exposure and then remained
20      at that level.  This observation clearly indicates a response plateau as suggested in other
21      prolonged exposure studies (Horstman et al., 1990; McDonnell et al., 1991). However, with the
22      triangular O3 concentration profile after a minimal initial response over the first 3 h, Hazucha
23      et al. (1992) observed a substantial decrease in FEVj corresponding to the higher average O3
24      concentration that reached a nadir after 6 h (-10.3%). Despite 2 h of continued exposure to a
25      lower O3 concentration (0.12 to 0.00 ppm,  mean = 0.06 ppm), FEVj improved and was only
26      reduced by 6.3% (relative to the preexposure FEVj) at the end of the 8-h exposure. The authors
27      concluded that total inhaled O3 dose (C x VE x T) was not a sufficient index of O3 exposure and
28      that, as observed by others (Adams et al., 1981; Folinsbee et al., 1978; Hazucha, 1987; Larsen
29      et al., 1991), O3 concentration appears to be more important in determining exposure effects than
30      is either duration or the volume of air breathed during the exposure.  However, it should be noted
31      that the mean O3 concentration for Hazucha et al.'s triangular exposure profile was 0.12 ppm

        January 2005                            AX6-15      DRAFT-DO NOT QUOTE OR CITE

-------
                              4.5

                              4.4
                              4.3
                          HI
                          (0
                              4.2
                              4.1
                              4.0
                              3.9
••••••• 0 ppm
—H— 0.12 ppm
     Constant
—A— Variable
                                     0246
                                           Exposure Duration (h)
       Figure AX6-3. The forced expiratory volume in 1 s (FEVt) is shown in relation to exposure
                      duration (hours) under three exposure conditions. Subjects exercised
                      (minute ventilation « 40 L/min) for 30 min during each hour; FEVt was
                      measured at the end of the intervening rest period. Standard error of the
                      mean for these FEVt averages (not shown) ranged from 120 to 150 mL.
       Source: Hazuchaetal. (1992).
 1     at 4 h, 0.138 ppm at 5 h, 0.14 ppm at 6 h, and 0.133 ppm at 7 h, before falling to 0.12 ppm at 8 h.
 2     The FEVj responses of the last 4 hours (Figure AX6-3) follow a closely similar pattern as the
 3     total mean O3 concentration over the same time period.
 4          It has become increasingly well realized that laboratory simulations of air-pollution risk-
 5     assessment need to employ O3 concentration profiles that more accurately mimic those
 6     encountered during summer daylight ambient air pollution episodes (Adams and Ollison, 1997;
 7     Lefohn and Foley,  1993; Rombout et al., 1986). Neither square-wave O3 exposures or the one
 8     8-h study by Hazucha et al. (1992) that utilized  a triangular shaped varied O3 exposure described
 9     above closely resembles the variable diurnal daylight O3 concentration pattern observed in many
10     urban areas experiencing air-pollution episodes (Lefohn and Foley, 1993).  Recently, 6.6 h less
       January 2005
          AX6-16
DRAFT-DO NOT QUOTE OR CITE

-------
 1     abrupt triangular O3 exposure profiles at lower concentrations more typical of outdoor ambient
 2     conditions have been examined (Adams 2003a; Adams and Ollison, 1997).
 3          Using a face-mask inhalation system, Adams and Ollison (1997) observed no significant
 4     differences in postexposure pulmonary function responses or symptoms between the 6.6-h,
 5     0.12 ppm O3 square-wave exposure; and those observed for a triangular O3 profile in which
 6     concentration was increased steadily from 0.068 ppm to 0.159 ppm at 3.5 h and then decreased
 7     steadily to 0.097 ppm at end exposure.  Further, no attenuation in FEVj response during the last
 8     2 h was observed in either the 6.6 h square-wave or the triangular exposures.  In a subsequent
 9     study (Adams, 2003a), no significant difference was observed in pulmonary function responses
10     or symptoms between face-mask and chamber exposure systems either for a 6.6-h, 0.08 ppm O3
11     square-wave profile or for the triangular O3 exposure beginning at 0.03 ppm, increasing steadily
12     to 0.15 ppm in the fourth hour, and decreasing steadily to 0.05 ppm at 6.6 h (mean = 0.08 ppm).
13     For the chamber-exposure comparison,  postexposure values for FEVj and symptoms were not
14     significantly different from the responses for the square-wave 0.08 ppm O3 exposure. However,
15     analysis showed that FEVj response for the square-wave protocol did not become statistically
16     significant until the 6.6-h postexposure value, while that for the triangular exposure protocol was
17     significant at 4.6 h (when  O3 concentration was 0.15 ppm). Earlier significant FEVj responses
18     for the triangular protocol were accompanied by significant increases in symptoms at 4.6 h,
19     continuing on through the fifth and sixth hours when the mean O3 concentration was 0.065 ppm.
20     Symptoms for the square-wave  0.08 ppm exposure did not become statistically significant until
21     5.6 h.  The rate of FEVj response to the triangular exposure did not decrease as was observed by
22     Hazucha et al. (1992) during the last two hours of their 8-h triangular exposure (Figure AX6-3).
23     Rather, FEVj responses for the triangular exposure showed clear signs of plateauing during the
24     last 2 h; i.e., -5.46% at 4.6 h, -6.27% at 5.6 h, and -5.77% at 6.6 h.  The most probable reason
25     for differences in the triangular O3 profile observations of Hazucha et al. (1992) and those of
26     Adams (2003a) is that the increase and decrease in Hazucha et al.'s study (i.e., 0 to 0.24 ppm and
27     back to 0) encompassed a much greater range of O3 concentrations than those used by Adams
28     (2003a), viz., 0.03 ppm to 0.15 ppm from 3.6 to 4.6 h, then decreasing to 0.05 ppm at 6.6 h.
29     Nonetheless, the greatest FEVj  decrement was observed  at 6 h of Hazucha et al.'s 8 h triangular
30     exposure (Figure AX6-3)  corresponding to the time when total mean O3 concentration was
31     highest (0.14 ppm), with a very similar response at 7 h when total mean O3 concentration was

       January 2005                            AX6-17     DRAFT-DO NOT QUOTE OR CITE

-------
 1      0.138 ppm.  Adams (2003a) observed the greatest FEVj decrement at 5.6 h (-6.27% with total
 2      mean O3 concentration of 0.086 ppm), which was not significantly different than the 4.6-h value
 3      of -5.46% (total mean O3 concentration = 0.0875 ppm).
 4          Whereas FEVj decrements during square-wave O3 exposures between 0.08 to 0.12 ppm
 5      tend to increase with time of exposure (i.e., with steadily increasing total inhaled dose), FEVj
 6      decrements during triangular exposures (Hazucha et al., 1992; Adams, 2003a) occurred 1 to 2 h
 7      after the peak O3 concentration and 1 h to 2 h before the maximal total O3 inhaled dose occurred
 8      at end exposure. This difference, especially because  O3 concentration profiles during summer
 9      daylight air-pollution episodes rarely mimic a square-wave, implies that triangular O3 exposure
10      profiles most frequently observed during summer daylight hours merit further investigation.
11
12      AX6.2.5  Mechanisms of Pulmonary Function Responses
13          Inhalation of O3 for several hours while physically active elicits both subjective respiratory
14      tract symptoms and acute pathophysiologic changes.  The typical symptomatic response
15      consistently reported in studies is that of tracheobronchial airway irritation. This is accompanied
16      by decrements in lung capacities and volumes, bronchoconstriction, airway hyperresponsiveness,
17      airway inflammation, immune system activation, and epithelial injury.  The severity of
18      symptoms and the magnitude of response depend on inhaled dose, O3 sensitivity of an individual
19      and the extent of tolerance resulting from previous exposures. The development of effects is
20      time dependent during both exposure and recovery periods with considerable overlap of evolving
21      and receding effects.
22          Exposure to O3 initiates reflex responses manifested as a decline in spirometric lung
23      function parameters (1FVC, IFEVl3 1FEF25.75), bronchoconstriction (t SRaw) and altered
24      breathing pattern (1VT,  t fB), which becomes more pronounced as exposure progresses and
25      symptoms of throat irritation, cough, substernal soreness and pain on deep inspiration develop.
26      The spirometric lung function decline and the severity of symptoms during a variable (ramp)
27      exposure profile seem to peak a short time (about 1 to 2 h) following the highest concentration of
28      O3 (Hazucha et al., 1992; Adams, 2003a). Exposure to a uniform O3 concentration profile elicits
29      the maximum spirometric response at the end of exposure (Hazucha et al., 1992; Adams, 2003a).
30      Regardless of exposure concentration profile, as the exposure to O3 progresses airway
31      inflammation begins to develop and the immune response at both cellular and subcellular level is

        January 2005                            AX6-18       DRAFT-DO NOT QUOTE OR CITE

-------
 1      activated. Airway hyperreactivity develops slower than pulmonary function effects, while
 2      neutrophilic inflammation of the airways develops even more slowly and reaches the maximum
 3      3 to 6 h postexposure.  The cellular responses (e.g., release of immunoregulatory cytokines)
 4      appear to still be active as late as 20 h postexposure (Torres et al., 2000). Following cessation of
 5      exposure, the recovery in terms of breathing pattern, pulmonary function and airway
 6      hyperreactivity progresses rapidly and is almost complete within 4 to 6 hours in moderately
 7      responsive individuals. Persisting small residual lung function effects are almost completely
 8      resolved within 24 hours.  Following a 2 h exposure to 0.4 ppm O3 with IE, Nightingale et al.
 9      (2000) observed a 13.5% decrement in FEVj. By 3 h postexposure, however, only a 2.7% FEVj
10      decrement persisted. As illustrated in Figure AX6-4, a similar postexposure recovery in FVC
11      was observed. In hyperresponsive individuals,  the recovery takes longer and as much as
12      48 hours to return to baseline values. More slowly developing inflammatory and cellular
13      changes persist for up to 48 hours.  The time sequence, magnitude and the type of responses of
14      this complex series of events, both in terms of development and recovery, indicate that  several
15      mechanisms, activated at different time of exposure must contribute to the overall lung  function
16      response (U.S. Environmental Protection Agency,  1996).
17
18      AX6.2.5.1 Pathophysiologic Mechanisms
19      Breathing pattern changes
20           Human studies consistently report that inhalation of O3 alters the breathing pattern without
21      significantly  affecting minute ventilation. A progressive decrease in tidal volume and a
22      "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
23      exposure suggests a direct modulation of ventilatory control.  These changes parallel a response
24      of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
25      Although alteration of a breathing pattern could be to some degree voluntary, the presence of the
26      response in animals and the absence of perception of the pattern change by subjects, even before
27      appearance of the first subjective symptoms of irritation, suggests an involuntary reflex
28      mechanism.
29           Direct recording from single afferent vagal fibers in animals convincingly demonstrated
30      that bronchial C-fibers and rapidly adapting receptors are the primary vagal afferents responsible
31      for O3-induced changes in ventilatory rate and depth (Coleridge et al., 1993; Hazucha and

        January 2005                            AX6-19      DRAFT-DO NOT QUOTE OR CITE

-------
                         a.
                           LJ
                         b.
                           o
                               4.0-,
                               3.5-
                               2.5-
                               2.0-
                                    I     I     I    I     I     I     I  '  >\
                                   Pre        0    1     2     3     4    24
                                        Time from Exposure (hours)
                               4.5-I
                               4.0-
                               3.5-
                               3.0-
                                   Pre        01234
                                        Time from Exposure (hours)
                        24
      Figure AX6-4a,b.  Recovery of spirometric responses following a 2 h exposure to 0.4 ppm
                        O3 with IE. Immediately postexposure, there were significant
                        decrements (**p < 0.001, ***p < 0.0005) in FVC (10%) and FEVt
                        (13.5%) compared to preexposure values. At 3 h postexposure, FVC
                        and FEVt were at 96 and 97% of preexposure values, respectively.

      Adapted from Nightingale et al. (2000).
1     Sant'Ambrogio, 1993). In spontaneously breathing dogs, an increase in VT/T; (T; decreased more
2     than VT) was attributed to an increased inspiratory drive due to stimulation of rapidly adapting
3     receptors and bronchial C-fibers by O3 (Schelegle et al., 1993). Folinsbee and Hazucha (2000)
4     also observed similar changes in VT/T; and other breath-timing parameters in humans exposed to
      January 2005
AX6-20
DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3 implying activation of the same mechanisms.  They also reported that Pm0A (pressure at
 2      mouth at 0.1 sec of inspiration against a transiently occluded mouthpiece which is considered an
 3      index of inspiratory drive) increased during controlled hypercapnia without a change in the slope
 4      of Pm0 j versus pCO2 relation suggesting that the primary mechanism is an increased inspiratory
 5      drive. Since no significant within-individual differences in ventilatory response to CO2 between
 6      air exposure and O3 exposure were found, the CO2 chemoreceptors did not modulate the
 7      response. Therefore, the principal peripheral mechanism modulating changes in breathing
 8      pattern appears to be direct and indirect stimulation of lung receptors and bronchial C-fibers by
 9      O3 and/or its oxidative products. The activity of these afferents, centrally integrated with input
10      from other sensory pathways, drives the ventilatory controller, which determines the depth and
11      the frequency of breathing.
12           The potential modulation of breathing pattern by activation of sensory afferents located in
13      extrathoracic airways by O3 has not yet been studied in humans.  Laboratory  animal studies have
14      shown that the larynx, pharynx, and nasal mucosa are densely populated by free-ending,
15      unmyelinated sensory afferents resembling nociceptive C-fibers (Spit et al., 1993; Sekizawa and
16      Tsubone, 1994).  They are almost certainly stimulated by O3 and likely  contribute to overall
17      ventilatory and symptomatic responses. Nasal only exposure of rats produced O3-induced
18      changes in breathing pattern that are similar to changes found in humans (Kleinman et al., 1999).
19
20      Symptoms and lung function changes
21           As already discussed, in addition to changes in ventilatory control, O3 inhalation by
22      humans will also induce a variety of symptoms, reduce vital capacity (VC) and related functional
23      measures, and increase airway resistance. Hazucha et al. (1989) postulated that a reduction of
24      VC by O3 is due to a reflex inhibition of inspiration and not due to a voluntary reduction of
25      inspiratory effort. Recently, Schelegle et al. (2001) convincingly demonstrated that a reduction
26      of VC due to O3 is indeed reflex in origin and not a result of subjective discomfort and
27      consequent premature voluntary termination of inspiration. They reported that inhalation of an
28      aerosolized topical anesthetic tetracaine substantially reduced if not abolished O3-induced
29      symptoms that are known to be mediated in part by bronchial C-fibers.  Yet,  such local
30      anesthesia of the upper airway mucosa had a minor and irregular effect  on pulmonary function
31      decrements and tachypnea, strongly supporting neural mediation, i.e., stimulation of both

        January 2005                            AX6-21      DRAFT-DO NOT QUOTE OR CITE

-------
 1      bronchial and pulmonary C-fibers, and not voluntary inhibition of inspiration (due to pain) as the
 2      key mechanism.
 3           The involvement of nociceptive bronchial C-fibers modulated by opioid receptors in
 4      limiting maximal inspiration and eliciting subjective symptoms in humans was studied by
 5      Passannante et al. (1998). The authors hypothesized that highly variable responses among
 6      individuals might reflect the individual's inability or unwillingness to take a full inspiration.
 7      Moreover, development of symptoms of pain on deep inspiration, cough and substernal soreness
 8      suggested that nociceptive mechanism(s) might be involved in O3-induced inhibition of maximal
 9      inspiration.  If this were so, pain suppression or inhibition by opioid receptor agonists should
10      partially or fully reverse symptoms and lung functional impairment. Subjects for this study were
11      pre-screened with exposure to 0.42 ppm O3 and classified either as "weak" (FEVj > 95% of
12      preexposure value), "strong" (FEVj < 85% of preexposure value), or "moderate" responders.
13      Sixty two (28 M, 34 F) healthy volunteers (18 to 59 yrs old), known from the previous screening
14      to be "weak" (n = 20) or "strong" (n = 42) O3-responders, participated in this double-blind
15      crossover study.  Subjects underwent  either two 2 h exposures to air, or two 2 h exposures to
16      0.42 ppm O3, with 15 min IE at 17.5 1/min/m2 BSA. Immediately following postexposure
17      spirometry the "weak" responders were given (in random order) either the potent opioid receptor
18      antagonist naloxone (0.15 mg/kg) or saline, while "strong" responders received (in random
19      order) either the potent, rapid-acting opioid agonist and analgesic sufentanil (0.2 jig/kg), or
20      physiologic saline administered through an indwelling catheter. Administration of saline or
21      naloxone had no significant effect on  the relatively small decrements in FEVj observed in
22      "weak" responders. However, as hypothesized, sufentanil rapidly reversed both the O3-induced
23      symptomatic effects and spirometric decrements (FEV^  p < 0.0001) in "strong" responders
24      (Figure AX6-5).  All the same, the reversal was not complete and the average post-sufentanil
25      FEVj remained significantly below (-7.3%) the preexposure value suggesting involvement of
26      non-opioid receptor modulated mechanisms as well. Uneven suppression of symptoms has
27      implied involvement of both A-6 and  bronchial C-fibers. The plasma  p-endorphin (a potent
28      pain suppressor) levels, though substantially elevated immediately postexposure and post-drug
29      administration, were not related to individuals' O3 responsiveness. These observations have
30      demonstrated that nociceptive mechanisms play a key role in modulating O3-induced inhibition
31      of inspiration. Moreover, these findings are consistent with and further support the concept that

        January 2005                             AX6-22     DRAFT-DO NOT QUOTE OR CITE

-------
             LJJ
                                0.5
 1           1.5

Time (hours)
                     2.5
      Figure AX6-5.  Plot of the mean FE Vt (% baseline) vs. time for ozone exposed cohorts.
                      Solid lines represent data for "strong" males (n = 14; solid squares) and
                      females  (n = 15; solid circles) that received sufentanil and dotted lines
                      represent data for the same cohorts after receiving saline.  Dashed lines
                      represent data for "weak" males (n = 5; open squares) and females (n = 10;
                      open circles) that received naloxone and dot-dash lines represent data for
                      the same cohorts after receiving saline. The arrow denotes the time of drug
                      administration (~2.1 hrs).  Vertical bars associated with the symbols are
                      one-sided SEM.

      Source: Adapted from Passannante et al. (1998).
1     the primary mechanism of O3-induced reduction in inspiratory lung function, is an inhibition of

2     inspiration elicited by stimulation of the C-fibers.  The absence of effect of naloxone in "weak"

3     responders shows that the weak response is not due to excessive endorphin production in those

4     individuals.  However, other neurogenic mechanisms not modulated by opioid receptors may
5     have some though limited role in inspiratory inhibition.
      January 2005
  AX6-23
DRAFT-DO NOT QUOTE OR CITE

-------
 1     Airway hyperreactivity
 2           In addition to limitation of maximal inspiration and its effects on other spirometric
 3     endopoints, activation of airway sensory afferents also plays a role in receptor-mediated
 4     bronchoconstriction and an increase in airway resistance. Despite this common mechanism,
 5     post-O3 pulmonary function changes and either early or late bronchial hyperresponsiveness
 6     (BHR) to inhaled aerosolized methacholine or histamine are poorly correlated either in time or
 7     magnitude.  Fentanyl and indomethacin, the drugs that have been shown to attenuate O3-induced
 8     lung function decrements in humans, did not prevent induction of BHR when administered to
 9     guinea pigs prior to O3 exposure (Yeadon et al., 1992).  Neither does post-O3 BHR seem to be
10     related to airway baseline reactivity. These findings imply that the mechanisms are either not
11     related or are activated independently in time. Animal studies (with limited support from human
12     studies) have suggested that an early post-O3 BHR is, at least in part, vagally mediated (Freed,
13     1996) and that stimulation of C-fibers can lead to increased responsiveness of bronchial smooth
14     muscle independently of systemic and inflammatory changes which may be even absent (load
15     et al.,  1996). In vitro study of isolated human bronchi have reported that O3-induced airway
16     sensitization involves changes in smooth muscle excitation-contraction coupling (Marthan,
17     1996).  Characteristic O3-induced inflammatory airway neutrophilia which at one time was
18     considered a leading BHR mechanism, has been found in a murine model, to be only
19     coincidentally associated with BHR and there was no cause and effect relationship (Zhang et al.,
20     1995).  However, this observation does not rule out involvement of other cells such as
21     eosinophils or T-helper cells in BHR modulation. There is some evidence that release of
22     inflammatory mediators by these cells can sustain BHR and bronchoconstriction.  In vitro and
23     animal studies have also suggested that airway neutral endopeptidase activity can be a strong
24     modulator of BHR (Marthan et al., 1996; Yeadon et al., 1992). Late BHR observed in some
25     studies is plausibly due to a sustained damage of airway epithelium and continual release of
26     inflammatory mediators (Foster et al., 2000).  Thus,  O3-induced BHR appears to be a product of
27     many mechanisms acting at different time periods and levels of the bronchial smooth muscle
28     signaling pathways. [The effects ofO3 on BHR are described in Section AX6.8. ]
29
        January 2005                             AX6-24      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.2.5.2  Mechanisms at a Cellular and Molecular Level
 2           Stimulation of vagal afferents by O3 and reactive products, the primary mechanism of lung
 3      function impairment is enhanced and sustained by what can be considered in this context to be
 4      secondary mechanisms activated at a cellular and molecular level.  The complexity of these
 5      mechanisms is beyond the scope of this section and the reader is directed to Section AX6.9 of
 6      this chapter for greater details.  A comprehensive review by Mudway and Kelly (2000) discusses
 7      the cellular and molecular mechanisms of O3-induced pulmonary response in great detail.
 8
 9      Neurogenic airway inflammation
10           Stimulation of bronchial C-fibers by O3 not only inhibits maximal inspiration but, through
11      local  axon reflexes, induces neurogenic inflammation. This pathophysiologic process is
12      characterized by release of tachykinins and other proinflammatory neuropeptides. Ozone
13      exposure has been shown to elevate C-fiber-associated tachykinin substance P in human
14      bronchial lavage fluid (Hazbun et al. 1993) and to deplete neuropeptides synthesized and
15      released from C-fibers in human airway epithelium rich in substance P-immunoreactive axons.
16      Substance P and other transmitters are known to induce granulocyte adhesion and subsequent
17      transposition into the airways, increase vascular permeability and plasma protein extravasation,
18      cause bronchoconstriction, and promote mucus secretion (Solway and Leff,  1991). Although the
19      initial pathways of neurogenic, antigen-induced, and generally immune-mediated inflammation
20      are not the same, they eventually converge leading to further amplification of airway
21      inflammatory processes by subsequent release of cytokines, eicosanoids, and other mediators.
22      Significantly negative correlations between O3-induced leukotriene (LTC4/D4/E4) production and
23      spirometric decrements (Hazucha et al., 1996), and an increased level of postexposure PGE2, a
24      mediator known to stimulate bronchial C-fibers, show that these mediators play an important
25      role in attenuation of lung function due to O3 exposure (Mohammed et al., 1993; Hazucha et al.,
26      1996). Moreover, because the density of bronchial C-fibers is much lower in the small  than
27      large airways, the reported post O3 dysfunction of small airways assessed by decrement in
28      FEF25.75 (Weinman et al., 1995; Frank et al., 2001) may be due in part to inflammation.  Also,
29      because of the relative slowness of inflammatory responses as compared to reflex effects, O3-
30      triggered inflammatory mechanisms are unlikely to initially contribute to progressive lung
31      function reduction. It is plausible, however, that when fully activated, they  sustain and possibly

        January 2005                             AX6-25      DRAFT-DO NOT QUOTE OR CITE

-------
 1     further aggravate already impaired lung function.  Indeed, a prolonged recovery of residual
 2     spirometric decrements following the initial rapid improvement after exposure termination could
 3     be due to slowly resolving airway inflammation. Bronchial biopsies performed 6 h postexposure
 4     have shown that O3 caused a significant decrease in immunoreactivity to substance P in the
 5     submucosa (Krishna et al., 1997a). A strong negative correlation with FEVj also suggests that
 6     the release of substance P may be a contributing mechanism to persistent post-O3
 7     bronchoconstriction (Krishna et al., 1997a). Persistent spirometry changes observed for up to
 8     48 h postexposure could plausibly be sustained by the inflammatory mediators, many of which
 9     have bronchoconstrictive properties (Blomberg  et al., 1999).
10
11
12     AX6.3    PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
13                IN SUBJECTS WITH PREEXISTING DISEASE
14          This section examines the effects of O3 exposure on pulmonary function in subjects with
15     preexisting disease by reviewing O3 exposure studies that utilized subjects with (1) chronic
16     obstructive pulmonary disease (COPD), (2) asthma, (3) allergic rhinitis, and (4) ischemic heart
17     disease. Studies of subjects with preexisting disease exposed to O3, published subsequent to or
18     not included in the 1996 Air Quality Criteria Document (U.S. Environmental Protection Agency,
19     1996), are summarized in Table AX6-3.  Studies examining increased airway responsiveness
20     after O3 exposure are discussed in Section AX6.8.
21
22     AX6.3.1  Subjects with Chronic Obstructive Pulmonary Disease
23          Five studies of O3-induced responses  in COPD patients were available for inclusion in  the
24     1996 criteria document (U.S. Environmental Protection Agency, 1996). The COPD patients in
25     these studies were exposed during light IE (4  studies) or at rest (1  study) for 1 to 2 hours to O3
26     concentrations between 0.1 and 0.3 ppm. None of theses studies found significant O3-induced
27     changes in pulmonary function. Of the four studies examining arterial oxygen saturation, two
28     reported small but statistically significant O3-induced decreases in the COPD patients.  These
29     limited data suggest COPD patients experience  minimal O3-induced effects for 0.3 ppm O3
30     exposures less than 2 hours in duration. These findings are also consistent decreasing O3 effects
31     with increasing  age (see Section AX6.5.1).

       January 2005                           AX6-26       DRAFT-DO NOT QUOTE OR CITE

-------
o

O

to
O
O
Table AX6-3.  Ozone Exposure in Subjects with Preexisting Disease"
X
ON
to
H
6
o

o
H
O
O
HH
H
W
Ozone
Concentration'
ppm Hg/ni3


Exposure Duration Exposure Number and
and Activity Condition Gender of Subjects


Subject Characteristics


Observed Effect(s)


Reference
Subjects with Chronic Obstructive Pulmonary or Heart Disease
0.0 0
0.24 472




0.3 588



4 h IE 24 °C 9 M
15 min exercise 40% RH 10 M
15 min rest
VE ~ 20 L/min


3 h IE 22 °C 10 M
VE = 30 L/min 50% RH 6 M


COPD patients
Age-matched healthy NS
All subjects 59-71 years old



Hypertension
42-61 years old
Healthy
41-49 years old
No significant changes in FEVj, FVC, or SRaw due to
ozone in COPD patients. Equivocal SaO2 decrement
during 2nd and 3rd hours of ozone exposure in COPD
patients. Adjusted for exercise, ozone effects did not
differ significantly between COPD patients and healthy
subjects.
No major cardiovascular effects in either healthy or
hypertensive subjects.


Gong et al.
(1997a)
Gong and
Tierney (1995)


Gong et al.
(1998)


Subjects with Allergic Rhinitis
0.0 0
0.2 392






0.125 245
0.250 490



0.125 245



0.0 0
0.25 490







1 h CE 20 °C 13 M, 1 F
at VE = 25 L/min/m2 50% RH
BSA





3h IE 27 °C 5 F, 6 M
(10 min rest, 15 min 50 % RH
exercise on bicycle) 6 F, 16 M
VE = 30 L/min

3h IE x 4 days



3hIE, 27 °C 13 M, 11 F
VE = 30 L/min 54% RH
15 min ex/10 min mouthpiece 6 M, 6 F
rest/5 min no O3; exposure
every 30 min.
5M, 5F



Dust mite sensitized asthmatics
mean age 29 ± 5 years






Mild bronchial asthma
20-53 years old
Allergic rhinitis
19-48 years old





Atopic mild asthma

Positive allergen and IgE tests

Healthy NS




FEVj decrement following O3 of 10% not significantly
different from the 4% decrement following FA. Subjects
received dust mite antigen challenge at 0.5 h FA and O3
postexposures and were lavaged 6 h post-challenge.
Amount of allergen producing 1 5% FEVj decrement was
decreased by O3 compared to FA in 9 of 14 subjects.
PMN in proximal airway lavage tended to be greater after
O3 than FA (p=0.06).
Mean early-phase FEVj response and number of > 20%
reductions in FEVj were significantly greater after
0.25 ppm O3 or 4 x 0.125 ppm O3. Most of the > 15%
late-phase FEVj responses occurred after 4 days of
exposure to 0.125 ppm O3, as well as significant
inflammatory effects, as indicated by increased sputum
eosinophils (asthma and allergic rhinitis) and increased
sputum lymphocytes, mast cell tryptase, histamine, and
LDH (asthma only).
O3-induced FEVj decrements of 12.5, 14. 1, and 10.2%
in asthmatics, allergic rhinitics and healthy subjects,
respectively (group differences not significant).
Methacholine responsiveness increased in asthmatics.
Allergen responsiveness: increased significantly after O3
exposure in asthmatics (~ 2 dose shift) and a smaller shift
is rhinitics. No change in healthy. Neither allergen or
methacholine response correlated with lung function and
were not correlated with each.
Chen et al.
(2004)






Holz et al.
(2002)







Jorres et al.
(1996)








-------
o

O

to
O
O
Table AX6-3 (cont'd). Ozone Exposure in Subjects with Preexisting Disease"
X
to
oo
H
6
o

o
H
O
O
HH
H
W
Ozone
Concentration'
ppni fig/m3
Exposure Duration Exposure
and Activity Conditions
Number and
Gender of Subjects Subject Characteristics
Observed Effect(s)
Reference
Subjects with Asthma
0.4 784







0.0 0
0.4 784





0.12 236

0.0 0
0.2 392




0.4 784



0.0 0
0.12 235

0.4 784






2h IE NA
(15 min rest, 15 min exercise
on bicycle)
VE = 30 L/min




2h IE NA
4x15 min on bicycle,
VE = 40 L/min




Rest 22 °C
40% RH
6 h 22 °C
30 min rest/30 min exercise 50% RH
VE ~ 25 L/min



3h 6x15 min cycle ergometer 31 °C
VE » 32L/min 35% RH
5 consecutive days

1 h rest NA
air-antigen
O3-antigen
2 h IE Head mask
15 min exercise exposure
15 min rest »18°C
VE » 20L/min 60% RH



4 F, 5 M Healthy
(25 ± 2 years old)

7 F, 6 M Mild atopic asthma;
beta agonists only
(22 ± 0.7 years old)


15 Healthy adults
18-40 years old

9 Mild atopic asthmatics
18-40 years old


10 M, 5 F atopic asthma

5 M Healthy NS
5 M Asthmatics, physician
diagnosed,
All 18-45 years


8 M , 2 F Asthmatic NS adults
beta-agonist use only
19-48 years old
ATS criteria for asthma
9 M, 6 F Mild allergic asthma; 18 to
49 years of age.

5 M, 1 F Healthy adults
6 M Atopic asthmatics





Significant reductions in FVC (12%, 10%) and
FEV[ (13%, 1 1%) for asthmatic and healthy
subjects, respectively; attenuated by indomethacin
in healthy subjects only. Significant reductions in
mid-flows which tended to be greater in asthmatics
than healthy subjects. Indomethacin treatment
attenuated mid-flow-reductions somewhat more
in asthmatics than healthy subjects.
Sputum collected 24 h before and 4-6 h post O3
exposure. Baseline GDI Ib expression positively
correlated with O3-induced PMN. Increased
expression of mCD14 on macrophages following
O3 compared to FA. Asthmatic PMN response
similar to healthy subjects (also see Table AX6-3).
No spirometric data available.
No effect of O3 on airway response to grass
allergen.
Similar spirometric responses in asthmatic and
healthy. However, preexposure FEV1 and FVC
were both -0.4 L lower on O3-day than FA day.
More PMN's in asthmatics. IL-8 and IL-6 higher
in asthmatics exposed to O3. No relationship of
spirometry and symptoms to inflammation.
FEVj decreased 35% on first exposure day.
Methacholine reactivity increased about ten-fold.
Also see Table AX6- 7 for repeated exposure results.

No effect of O3 on airway response to grass or
ragweed allergen.

FEVj responses of healthy and asthmatic similar
(~ 15% decrease). Maximal FEVj response to
methacholine increased similarly in both groups
(12 h postexposure). Larger increase in PC20 in
healthy subjects. Both groups had increased
PMN's in sputum no correlation of PMN's
and lung function.
Alexis et al.
(2000)






Alexis et al.
(2004)





Ball et al.
(1996)
Basha et al.
(1994)




Gong et al.
(1997b)


Hanania
etal. (1998)

Hiltermann
etal. (1995)






-------
o

O

to
O
O
Table AX6-3 (cont'd).  Ozone Exposure in Subjects with Preexisting Disease"
X
to
VO
H
6
o

o
H
O
Ozone
Concentration'
ppm ug/m
Number and
Exposure Duration and Exposure Gender of
Activity Conditions Subjects
Subject
Characteristics
Observed Effect(s)
Reference
Adult Subjects with Asthma (cont'd)
0.0 0
0.16 314




0.0 0
0.25 490






0.16 314




0.25 490
0.40 784


0.0 0
0.2 392



0.2 396




7.6 h 18° C 13 M
25 min treadmill, 40% RH
25 min cycle/10 min rest 7 M, 10 F
per hour.
VE =27-32 L/min

3hIE, 27 °C 13 M, 11 F
VE = 30 L/min 54% RH
15 min ex/10 min rest/5 mouthpiece 6 M, 6 F
min no O3; every 30 min. exposure

5M, 5F


7.6 h 22 °C 4 M, 5 F
25 min treadmill, 25 min 40 % RH
cycle/10 min rest per
hour.
VE = 25 L/min
VE = 25-45 L/min NA 8 M, 4 F
8 M, 10 F
22 M, 16 F

2 h IE 20 °C 6 M, 9 F
4x15 min 50% RH
at VE = 20
L/min/m2 BSA 9 M, 6 F

2h IE 22 °C 5 F, 4 M
(1 5 min rest, 1 5 min 40 % RH
exercise on bicycle)
VE = 20 L/min/m2 BSA

Healthy NS,
age 19-32 years.
Moderate Asthmatics,
physician diagnosed,
beta agonist users,
age 19-32 years.
Atopic mild asthma

Positive allergen and
IgE tests

Healthy NS


Mild atopic asthma;
no meds 12 hpre-
exposure
20-35 years old

Asthmatics
Allergic rhinitics
Healthy adults
All < 26 years old
Healthy adults
24 years old

Mild asthmatics
29 years old
Mild atopic asthma; no
meds 8 h pre-exposure
21-42 years old


FEVj decreased 19% in asthmatics and only 10% in
non-asthmatics. High responders had worse baseline airway
status. More wheeze in asthmatics after O3.



O3-induced FEVj decrements of 12.5, 14.1, and 10.2% in
asthmatics, allergic rhinitics and healthy subjects, respectively
(group differences not significant). Methacholine
responsiveness increased in asthmatics. Allergen responsiveness
increased after O3 exposure in asthmatics (= 2 dose shift),
a smaller shift occurred in rhinitics, no change occurred in
healthy subjects. Neither allergen nor methacholine responses
were correlated with each other or with lung function.
Significant FEVj decrease of 9.1 % following O3 exposure;
marked individual variability with responses ranging from 2
% to 26 %.


Healthy 12.2% decrease in FEVj, Rhinitics 10.1%,
asthmatics 12.4%


O3-induced FEV; decrement (8%, healthy adults; 3% asthmatics)
and PMN increase (20.6%, healthy adults; 15.2% asthmatics).
Primary goal was to investigate relationship between antioxidant
defenses and O3 responses in asthmatics and healthy adults
(see Tables AX6-3 andAX6 -13).
Significant decrease in FEVj and a trend toward decreases in
mean inspiratory flow, FEF25, and FEF75 after O3 exposure.
No significant differences in FEF50, FVC, TLC, Raw, or sRaw.
No correlation between sputum neutrophils at 6 h postexposure
and FEVj immediately after exposure.
Horstman
etal. (1995)




Jorres et al.
(1996)






Kehrl et al.
(1999)



Magnussen
etal. (1994)


Mudway
etal. (2001)
Stenfors
et al. (2002)

Newson
et al. (2000)



O
HH
H
W

-------
O
o
i
{_} '
CD
>-l
(^
to
O
O
•^












J>

1
O
Table AX6-3 (cont
Ozone
Concentration'
Exposure Duration Exposure
ppm Hg/ni3 and Activity Conditions
Adult Subjects with Asthma (cont'd)

0.4 784 2 h rest 21 "C
40% RH


0.16 314 7.6 h 18° C
25 min treadmill, 40% RH
25 min cycle/ every
hour.
0.0 0 4h 21 °C
0.2 392 50 min exercise, 50% RH
10 min rest each hour.
VE » 45-50 L/min




"See Appendix A for abbreviations and acronyms.
f~li-r*ii«fiH hw rf^ct 3iiH f^vfMvicf*' Axjitliiii rmr>f*ntr3tir*ii
with Preexisting Disease"



Observed Effect(s)


Ozone resulted in nasal inflammation (increased PMN's)
and caused augmented response to nasal allergen challenge.


Increased eosinophils and PMN's after O3 exposure more
in initial (bronchial) fraction. No correlation of eosinophils
and PMN's, FEVj & FVC decreased 14% and 9%
respectively.
FVC, FEVj decreased 17.6% and 25% respectively.
Trend for larger increase in SRaw in asthmatics. Larger
increase in PMN's and protein in asthmatics indicating
more inflammation. No increase in eosinophils.
Spirometry changes in asthmatics similar to healthy
subjects (Aris et al., 1995; Balmes et al., 1997).







Reference


Peden et al.
(1995)


Peden et al.
(1997)


Scannell et al.
(1996)







 H
 6
 o

 o
 H
O
O
HH
H
W

-------
 1           More recently, Gong et al. (1997a) exposed 9 COPD patients (age range, 59 to 71 years;
 2      mean age 66 ± 4 years) and 10 healthy NS (age range, 60 to 69 years; mean age 65 ± 3 years)
 3      to 0.24 ppm for 4 h with interment light exercise («20 L/min).  COPD patients had decreases in
 4      FEVj following both clean air (-11%, p = 0.06) and O3 (-19%, p < 0.01) exposures. These
 5      FEVj decrements, presumably due to exercise, were primarily attributable to four of the patients
 6      who lost greater than 14% of their FEVj following both the air and O3 exposures. Relative to
 7      clean air, O3 caused a statistically insignificant FEVj decrement of - 8% in COPD patients which
 8      was not statistically different from the decrement of -3% in healthy subjects.  Ozone-induced
 9      symptoms, sRaw, SaO2, and postexposure bronchial activity also exhibited little or no difference
10      between the COPD patients and the healthy subjects.
11
12      AX6.3.2  Subjects with Asthma
13           Based on studies reviewed in the 1996 criteria document (U.S. Environmental Protection
14      Agency, 1996) asthmatics appear to be at least as sensitive to acute effects of O3 as healthy
15      nonasthmatic subjects.  At rest, neither adolescent asthmatics nor healthy controls had significant
16      responses as a result of an hour exposure to 0.12 ppm O3. Exposure of adult asthmatics to 0.25
17      ppm O3 for 2 h at rest also caused no significant responses. Preexposure to between 0.10 and
18      0.25 ppm O3 for 1 hr with light IE does not appear to exacerbate exercise-induced asthma
19      (Fernandes et al., 1994; Weymer et al., 1994). At higher exposures (0.4 ppm O3 with heavy IE
20      for 2 h), Kreit et al. (1989) and  Eschenbacher et al. (1989) demonstrated significantly greater
21      FEVj and FEF25.75 decrements in asthmatics than in healthy controls.  With longer duration
22      exposures to lower O3 levels (0.12 ppm with moderate IE for 6.5 h), asthmatics have also shown
23      a tendency for greater FEVj decrements than healthy non-asthmatics (Linn et al., 1994). Newer
24      studies (see Table AX6-3) continue to suggest that asthmatics are at least as sensitive as healthy
25      controls to O3-induced responses.
26           Studies of less than 3 h duration have reported similar or tendencies  for increased O3-
27      induced spirometric responses up to O3 concentrations of 0.4 ppm.  Similar group decrements in
28      FEVj and FVC were reported by Hiltermann et al. (1995), who exposed 6  asthmatics and
29      6 healthy subjects to 0.4 ppm O3 for 2 h with light IE. Alexis et al. (2000) exposed 13  mild
30      atopic asthmatics and 9 healthy subjects for 2 h to 0.4 ppm O3 with IE ( VE = 30 L/min).  Similar
31      O3-induced group decrements in FEVj and FVC were also reported by these investigators.

        January 2005                            AX6-31      DRAFT-DO NOT  QUOTE OR CITE

-------
 1      A tendency, however, for an increased O3-induced reduction in mid-flows (viz., FEF25, FEF50,
 2      FEF60p, FEF75) was reported for the asthmatics relative to the healthy subjects.  In a larger study,
 3      Torres et al. (1996) exposed 24 asthmatics, 12 allergic rhinitis, and 10 healthy subjects to
 4      0.25 ppm O3 for 3 h with IE. Statistically significant O3-induced decreases in FEVj occurred in
 5      all groups, but tended to be lower in healthy controls (allergic rhinitis, -14.1%; asthmatics,
 6      -12.5%; healthy controls, -10.2%). One study reported that asthmatics tended to have less of an
 7      FEVj response to O3 than healthy controls (Mudway et al., 2001). In that study, however, the
 8      asthmatics also tended to be older than the healthy subjects which could partially explain their
 9      lesser response.
10           Studies between 4 and 8 h duration, with O3 concentrations of 0.2 ppm or less, also suggest
11      a tendency for increased O3-induced pulmonary function  responses in asthmatics relative to
12      healthy subjects. Scannell et al. (1996) exposed 18 asthmatics to 0.2 ppm O3 for 4 h with IE
13      (VE -25 L/min/m2 BSA). Baseline and hourly pulmonary function measurements of FEVl3
14      FVC, and sRaw were obtained. Asthmatic responses were compared to 81 healthy  subjects who
15      underwent similar experimental protocols (Aris et al., 1995; Balmes et al., 1996). Asthmatic
16      subjects experienced a significant O3-induced increase in sRaw, FEVj and FVC. The O3-induced
17      increase in sRaw tended to be greater in asthmatics than the healthy subjects, whereas similar
18      group decrements in FEVj and FVC were observed. Basha et al. (1994) also reported similar
19      spirometric responses between 5 asthmatic and 5 healthy  subjects exposed to 0.2 ppm O3 for 6 h
20      with IE.  However, the mean preexposure FEVj in the asthmatics was about 430 mL less (i.e.,
21      ~ 12% decreased) on the O3-day relative to the air-day. In a longer exposure duration (7.6 h)
22      study, Horstman et al. (1995) exposed 17 asthmatics and  13 healthy controls to 0.16 ppm O3 or
23      FA with alternating periods of exercise (50 min, VE -30 L/min) and rest (10 min).  Both groups
24      had significant O3-induced decrements in FEVl3 FVC, and FEV25_75. The asthmatic and healthy
25      subjects had similar O3-induced reductions in FVC. The  FEVj decrement experienced by the
26      asthmatics was significantly greater in the healthy controls (19% versus 10%, respectively).
27      There was also tendency for a greater  O3-induced decrease in FEF25.75 in asthmatics relative to
28      the healthy subjects (24% versus 15%, respectively).
29           With repeated O3 exposures asthmatics, like healthy subjects (see Section AX6.6) develop
30      tolerance. Gong et al. (1997b) exposed  10 asthmatics to 0.4 ppm O3, 3 h per day with IE
31      ( VE -32 L/min), for 5 consecutive days. Symptom and spirometric responses were greatest on

        January 2005                            AX6-32      DRAFT-DO NOT QUOTE OR CITE

-------
 1      the first (-35 % FEVj) and second (-34 % FEVj) exposure days, and progressively diminished
 2      toward baseline levels (-6 % FEVj) by the fifth exposure day. Similar to healthy subjects,
 3      asthmatics lost their tolerance 4 and 7 days later.
 4           Other published studies with similar results (e.g., McBride et al., 1994; Basha et al., 1994;
 5      Peden et al., 1995, 1997; Peden, 2001a; Scannell et al., 1996; Hiltermann et al., 1997, 1999;
 6      Michelson et al., 1999; Vagaggini et al., 1999; Newson et al., 2000; Holz et al., 2002) also
 7      reported that asthmatics have a reproducible and somewhat exaggerated inflammatory response
 8      to acute O3 exposure (see Section AX6.9}. For instance, Scannell et al. (1996) performed lavages
 9      at 18 h post O3 exposure to assess inflammatory responses in asthmatics.  Asthmatic responses
10      were compared to healthy subjects who underwent a similar experimental protocol (Balmes
11      et al., 1996). Ozone-induced increases in BAL neutrophils and total protein were significantly
12      greater in asthmatics than healthy subjects. There was also a trend for an ozone related increased
13      IL-8 in the asthmatics relative to healthy subjects.  Inflammatory responses do not appear to be
14      correlated with lung function responses in either asthmatic or healthy subjects (Balmes et al.,
15      1996, 1997; Holz et al., 1999). This lack of correlations between inflammatory and spirometric
16      responses may be due to differences in the time kinetics of these responses (Stenfors et al.,
17      2002).  In addition, airway responsiveness to inhaled allergens is increased by O3 exposure in
18      subjects with allergic asthma for up  to 24 h (see Section AX6.8).
19           One of the difficulties in comparing O3-induced spirometric responses of healthy subjects
20      versus asthmatics is the variability in responsiveness of asthmatics. Most of the asthma studies
21      were conducted on  subjects with a clinical history of mild disease.  However, classification
22      asthma severity is not only based on functional assessment (e.g., percent predicted FEVj), but
23      also on clinical symptoms, signs, and medication use (Table AX6-4). Although "mild atopic
24      asthmatics" are frequently targeted as an experimental group, the criteria for classification has
25      varied considerably within and across the available published studies.  Although the magnitude
26      of group mean changes in spirometry may not be significantly different between healthy and
27      asthmatic subjects,  many of the studies have reported clinically significant changes in some
28      individuals.
29           Alexis et al. (2000) explored the possibility that the mechanisms of O3-induced spirometric
30      responses may differ between asthmatics and healthy  subjects. Physician-diagnosed mild atopic
31      asthmatics and healthy subjects were pretreated with 75 mg/day of indomethacin (a COX

        January 2005                             AX6-33      DRAFT-DO NOT QUOTE OR CITE

-------
                                                  Table AX6-4. Classification of Asthma Severity1
to
o
o
X

Classification
Severe
persistent
Moderate
persistent

Days with
Step symptoms
4 Continual
3 Daily

Nights with
symptoms
Frequent
> I/week
Lung Function2
FEVlorPEF PEF
% predicted variability
oral (%)
< 60 > 30
between > 30
60 and 80
Medications3

Daily
High-dose inhaled steroids (ICS)
and long-acting inhaled p2-agonist
If needed, add oral steroids
Low-to-medium-dose ICS and
long-acting p2-agonist (preferred)

Quick relief
Short-acting inhaled
p2-agonist, as needed; oral
steroids may be required
Short-acting inhaled
p2-agonist, as needed;
       Mild
       persistent
> 2/week,
but< 1
time/day
> 2/week
       Mild
       intermittent
 < 2/week     < 2/month
                                        Or
                        Medium-dose ICS (another preferred
                        option for children ages < 5 years)
                                        Or
                        Low-to-medium-dose ICS and either
                        leukotriene modifier or theophylline

80          20-30       Low-dose inhaled steroids (preferred)
                                        Or
                        Cromolyn leukotriene modifier,
                        or (except for children aged < 5 years)
                        nedocromil or sustained release
                        theophylline to serum concentration
                        of 5-15 ug/mL

80          < 20        No daily medicine needed
                                                                                                                            oral steroids may be
                                                                                                                            required
Short-acting inhaled
p2-agonist, as needed;
oral steroids may be
required
                                                                                Short-acting inhaled
                                                                                p2-agonist, as needed;
                                                                                oral steroids may be
                                                                                required
       1 Sources: Centers for Disease Control (2003); National Heart, Lung, and Blood Institute (1997, 2003).
       2For adults and children aged > 5 years who can use a spirometer or peak flow meter.
       3The medications listed here are appropriate for treating asthma at different levels of severity. The preferred treatments, dosage, and type of medication
        recommended vary for adults and children and are detailed in the EPR-Update 2002 stepwise approach to therapy. The stepwise approach emphasizes that
        therapy should be stepped up as necessary and stepped down when possible to identify the least amount of medication required to achieve goals of therapy.
        The stepwise approach to care is intended to assist, not replace, the clinical decision-making required to meet individual patient needs.

-------
 1      inhibitor) or placebo and then exposed for 2 h to 0.4 ppm O3 or to FA during mild IE
 2      (VE =30L/m).  The number and severity of O3-induced symptoms were significantly increased
 3      in both asthmatics and healthy subjects. These symptom responses were similar between the
 4      subject groups and unaffected by indomethacin pretreatment. Asthmatics and healthy subjects
 5      also had similar O3-induced reductions in FVC and FEVj.  These restrictive-type responses,
 6      occurring due to the combined effects of bronchoconstriction and reflex inhibition of inspiration
 7      (see Section AX6.2.7), were attenuated by indomethacin in the healthy subj ects but not the
 8      asthmatics. Thus, in healthy subjects but not asthmatics, COX metabolites may contribute to
 9      O3-induced reductions in FVC and FEVj.  As assessed by the magnitude of reductions in
10      mid-flows (viz. FEF25, FEF50, FEF60p, FEF75), the small  airways of the asthmatics tended to be
11      more affected than the healthy subjects. This suggests asthmatics may be more sensitive to
12      small airway effects of O3, which is consistent with the observed increases in inflammation and
13      airway responsiveness.  Indomethacin pretreatment attenuated some of these O3-induced small
14      airways effects (FEF50 in healthy subjects, FEF60p in asthmatics).
15
16      AX6.3.3 Subjects with Allergic Rhinitis
17           Most O3 exposure studies in humans with existing respiratory disease have focused on lung
18      diseases like  COPD and asthma.  However, chronic inflammatory disorders of the nasal airway,
19      especially allergic rhinitis, are very common in the population.  People with allergic rhinitis have
20      genetic risk factors for the development of atopy that predispose them to increased upper airway
21      responsiveness to specific allergens as well as nonspecific air pollutants like O3.  Studies
22      demonstrating the interaction between air pollutants and allergic processes in the human nasal
23      airways and rhinoconjunctival tissue have been reviewed by Peden (2001b) and Riediker et al.
24      (2001), respectively. Ozone exposure of subjects with allergic rhinitis has been shown to induce
25      nasal inflammation  and increase airway responsiveness to nonspecific bronchoconstrictors,
26      although to a lesser degree than experienced by asthmatics.
27           McDonnell et al. (1987) exposed non-asthmatic adults with allergic rhinitis to 0.18 ppm
28      O3.  The allergic rhinitics were no more responsive to O3 than healthy controls, based on
29      symptoms, spirometry, or airway reactivity to histamine although they had a small but
30      significantly  greater increase in SRaw.  The data on subjects with allergic rhinitis and asthmatic
        January 2005                             AX6-3 5      DRAFT-DO NOT QUOTE OR CITE

-------
 1      subjects suggest that both of these groups have a greater rise in Raw to O3 with a relative order
 2      of airway responsiveness to O3 being normal < allergic < asthmatic.
 3           Bascom et al. (1990) studied the upper respiratory response to acute O3 inhalation, nasal
 4      challenge with antigen, and the combination of O3 plus antigen in subjects with allergic rhinitis.
 5      Exposure to O3 caused significant increases in upper and lower airway symptoms, a mixed
 6      inflammatory cell influx with a seven-fold increase in nasal lavage PMNs, a 20-fold increase in
 7      eosinophils, and a 10-fold increase in mononuclear cells, as well as an apparent sloughing of
 8      epithelial cells. McBride et al. (1994) also observed increased nasal PMN's after O3 exposure in
 9      atopic asthmatics. Peden et al. (1995), who studied allergic asthmatics exposed to O3 found that
10      O3 causes an increased response to nasal allergen challenge in addition to nasal inflammatory
11      responses.  Their data suggested that allergic subjects have an increased immediate response to
12      allergen after O3 exposure. In a follow-up study, Michelson et al. (1999) reported that 0.4 ppm
13      O3 did not promote early-phase-response mediator release or enhance the response to allergen
14      challenge in the nasal airways of mild, asymptomatic dust mite-sensitive asthmatic subjects.
15      Ozone did,  however, promote an inflammatory cell influx, which helps induce a more significant
16      late-phase response in this population.
17           Torres et al. (1996) found that O3 causes an increased response to bronchial allergen
18      challenge in subjects with allergic rhinitis. This study also compared responses in subjects with
19      mild allergic asthma (see Sections AX6.3.2 andAX6.8). The subjects were exposed to 0.25 ppm
20      O3 for 3 h with IE. Airway responsiveness to methacholine was determined 1 h before and after
21      exposure; responsiveness to allergen was determined 3 h after exposure.  Statistically significant
22      decreases in FEVj occurred in subjects with allergic rhinitis (13.8%) and allergic asthma
23      (10.6%), and in healthy controls (7.3%). Methacholine responsiveness was statistically
24      increased in asthmatics, but not in subjects with allergic rhinitis. Airway responsiveness to an
25      individual's historical allergen (either grass and birch pollen, house dust mite, or animal dander)
26      was significantly increased after O3 exposure when compared to FA exposure. In subjects with
27      asthma and allergic rhinitis, a maximum percent fall in FEVj of 27.9 % and 7.8%, respectively,
28      occurred 3 days after O3  exposure when they were challenged with of the highest common dose
29      of allergen. The authors concluded that subjects with allergic rhinitis, but without asthma, could
30      be at risk if a high O3 exposure is  followed by a high dose of allergen.
        January 2005                            AX6-36      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Holz et al. (2002) extended the results of Torres et al. (1996) by demonstrating that
 2      repeated daily exposure to lower concentrations of O3 (0.125 ppm for 4 days) causes an
 3      increased response to bronchial allergen challenge in subjects with preexisting allergic airway
 4      disease, with or without asthma. There was no major difference in the  pattern of bronchial
 5      allergen response between subjects with asthma or rhinitis, except for a 10-fold increase in the
 6      dose of allergen required to elicit a similar response (> 20% decrease in FEVj) in the asthmatic
 7      subjects.  Early phase responses were more consistent in subjects with  rhinitis and late-phase
 8      responses were more pronounced in subjects with asthma.  There also was a tendency towards a
 9      greater effect of O3 in subjects with greater baseline response to specific allergens chosen on the
10      basis of skin prick test and history (viz., grass, rye, birch, or alder pollen, house dust mite, or
11      animal dander).  These data suggest that the presence of allergic bronchial sensitization, but not a
12      history of asthma, is a key determinant of increased airway allergen responsiveness with O3.
13      [A more complete discussion of airway responsiveness is found in Section AX6.8]
14
15      AX6.3.4  Subjects with Cardiovascular Disease
16           Superko et al. (1984) exposed six middle-aged males with angina-symptom-limited
17      exercise tolerance for 40  min to FA and to 0.2 and 0.3 O3 while they were exercising
18      continuously according to a protocol simulating their angina-symptom-limited exercise training
19      prescription (meanVE= 35 L/min). No significant pulmonary function impairment or evidence
20      of cardiovascular strain induced by O3 inhalation was observed. Gong  et al. (1998) exposed
21      hypertensive and healthy adult males, 41 to 78 years of age, to 0.3 ppm O3 for 3 h with IE at
22      30 L/min.  The ECG was monitored by telemetry, blood pressure by cuff measurement, and a
23      venous catheter was inserted for measurement of routing blood chemistries and cardiac ezymes.
24      Pulmonary artery and radial artery catheters were placed percutaneously for additional blood
25      sampling and for measurement of hemodynamic pressures, cardiac output, and SaO2. Other
26      hemodynamic variables were calculated, including cardiac index, stroke volume, pulmonary and
27      systemic vascular resistance, left and right ventricular stroke-work indices,  and rate-pressure
28      product. Spirometric volumes (FVC, FEVj) and respiratory symptoms were measured before
29      and after the O3 exposures.  The overall results did not indicate any major acute cardiovascular
30      effects of O3 in either the hypertensive or normal subjects.  Statistically significant O3 effects for
31      both groups combined were a decrease in FEVj and increases in FIR, rate-pressure product, and

        January 2005                             AX6-37      DRAFT-DO NOT QUOTE OR CITE

-------
 1     the alveolar-to-arterial PO2 gradient, suggesting that impaired gas exchange was being
 2     compensated for by increased myocardial work.  These effects might be more important in some
 3     patients with severe cardiovascular disease.
 4
 5
 6     AX6.4     INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
 7                 OF RESPONSE
 8          Analysis of the factors that contribute to intersubject variability is important for the
 9     understanding of individual responses, mechanisms of response, and health risks associated with
10     acute O3 exposures. Bates et al. (1972) noted that variation between individuals in sensitivity
11     and response was evident in respiratory symptoms and pulmonary function following O3
12     exposure.  A large degree of intersubject variability in response to O3 has been consistently
13     reported in the literature (Adams et al., 1981; Aris et al.,  1995; Folinsbee et al.,  1978; Kulle
14     et al., 1985; McDonnell et al., 1983). Kulle et al. (1985) noted that the magnitude of variability
15     between individuals in FEVj responses increases with O3 concentration.  Similarly, McDonnell
16     et al. (1983) observed FEVj decrements ranging from 3 to 48% (mean 18%) in 29 young adult
17     males exposed to 0.40 ppm O3 for 2 h during heavy IE. At a lower O3 concentration of
18     0.18 ppm, 20 similarly exposed subjects had FEVj decrements ranging from 0 o 23%
19     (mean = 6%), while those exposed to FA (n = 20) had decrements ranging from -2% to 6%
20     (mean = 1%) (McDonnell et al., 1983).  All  of the subjects in these studies were young adult
21     males.  (Intersubject variability related to age and gender is discussed in Sections AX6.5.1 and
22     AX6.5.2, respectively.)
23          More recently, McDonnell (1996) examined the FEVj response data from three 6.6 h
24     exposure studies of young adult males conducted at the EPA Health Effects Research Laboratory
25     in Chapel Hill, NC (Folinsbee et al., 1988; Horstman et al., 1990; McDonnell et al., 1991).
26     The response distributions for subjects at each of four O3 concentrations (0.0, 0.08, 0.10, and
27     0.12 ppm) are illustrated in Figure AX6-6.  It is apparent that the FEVj responses in FA are
28     small with most tightly grouped around zero. With increasing O3 concentration, the mean
29     response increases as does the variability about the mean. At higher O3 concentrations, the
30     distribution of response becomes asymmetric with a few  individuals experiencing large FEVj
31     decrements. The response distribution in Figure AX6-6  allows estimates of the number or

       January 2005                            AX6-3 8     DRAFT-DO NOT QUOTE OR CITE

-------
                            tn
                            +j
                            o
O
+J
c


o
                                 40-
                                 30-
                                                    0.12 ppm
                                  -15-10 -5 0 5 10 15 20 25 30 35 40 45 50
                                 50-
                                 30-
                                                    0.10 ppm
                                                  i  .  .  .  n
                                  -15-10 -5 0 5 10 15 20 25 30 35 40 45 50
                                 40-
                                 30-
                                 20-
                                 10-
                                                    0.08 ppm
                                               II  I  ni—in
                                  -15-10 -5 0 5 10 15 20 25 30 35 40 45 50

40-
30-
20-
10-
n-




PI
~ll







I — I




0 ppm



n
I
                                  -15-10 -5 0 5 10 15 20 25 30 35 40 45 50



                                      FEV-, (% Decrement)
Figure AX6-6.  Frequency distributions of percent decrements in FEVt for 6.6-h exposure


                to four concentrations of ozone.




Source: McDonnell (1996).
January 2005
             AX6-39       DRAFT-DO NOT QUOTE OR CITE

-------
 1      percentage of subjects responding in excess of a certain level. With FA exposure, none of 87
 2      subjects had a FEVj decrement in excess of 10%; however, 26%, 31%, and 46% exceeded a 10%
 3      decrement at 0.08, 0.10, and 0.12 ppm, respectively. FEVj decrements as large as 30 to 50%
 4      were even observed in some individuals.  In 6.6-h face mask exposures of young adults (half
 5      women) to 0.08 ppm O3, Adams (2002) found that 6 of 30 subjects (20%) had > 10% decrements
 6      in FEVj.  The response distributions in Figure AX6-6 underlines the wide range of response to
 7      O3 under prolonged exposure conditions and reinforces the observations by others consequent to
 8      2 h IE exposures at higher O3 concentrations (Horvath et al.,1981; McDonnell et al., 1983).
 9           Some of the intersubject variability in response to O3 inhalation may be due to intrasubject
10      variability, i.e., how reproducible the measured responses are in an individual between several
11      O3 exposures. The more reproducible the subject's response, the more precisely it indicates
12      his/her intrinsic responsiveness.  McDonnell et al. (1985a) examined the reproducibility of
13      individual responses to O3 in healthy human subjects (n = 32) who underwent repeated
14      exposures within a period of 21 to 385  days (mean = 88 days; no median reported) at one of five
15      O3 concentrations ranging from 0.12 to 0.40 ppm.  Reproducibility was assessed using the
16      intraclass correlation coefficient (R). The most reproducible responses studied were FVC
17      (R = 0.92) and FEVj (R = 0.91). However, at the lowest concentration, 0.12 ppm, relatively
18      poor FEVj reproducibility was observed (R = 0.58) due, in part, to a lack of specific O3 response
19      or a uniformly small response in the majority of subjects. McDonnell et al. (1985a) concluded
20      that for 2  h IE O3 exposures equal to or greater than 0.18 ppm, the intersubject differences in
21      magnitude of change in FVC and FEVj are quite reproducible over time and likely due to
22      differences in intrinsic responsiveness of individual subjects. Hazucha et al. (2003) exposed
23      47 subjects on three occasions for 1.5 h, with moderate intensity IE, to 0.40 to 0.42 ppm O3.
24      Reproducibility of FEVj responses was related to the length of time between re-exposures,
25      with a Spearman correlation R of 0.54 obtained between responses for exposures 1 and
26      2 (median = 105 days), and an R of 0.85 between responses for exposures 2 and 3
27      (median = 7 days).
28          Identification of mechanisms of response and health risks associated with acute O3
29      exposures are complicated by a poor association between various O3-induced responses.
30      For example, McDonnell et al. (1983) observed a very low correlation between changes in  sRaw
31      and FVC  (r = -0.16) for 135 subjects exposed to O3 concentrations ranging from 0.12 to

        January 2005                            AX6-40      DRAFT-DO NOT QUOTE OR CITE

-------
 1      0.40 ppm for 2.5 h with IE. In a retrospective study of 485 male subjects (ages 18 to 36 yrs)
 2      exposed for 2 h to one of six O3 concentrations at one of three activity levels, McDonnell et al.
 3      (1999) observed significant, but low, Spearman rank order correlations between FEVj response
 4      and symptoms  of cough (R = 0.39), shortness of breath (R = 0.41), and pain on deep inspiration
 5      (R = 0.30).  The authors concluded from their data that the O3-induced responses are related
 6      mechanistically to some degree, but that there is not a single factor which is responsible for the
 7      observed individual differences in O3 responsiveness across the spectrum of symptom and lung
 8      function responses. This conclusion is supported by differences in reproducibility observed by
 9      McDonnell et al., (1985a).  Compared to the intraclass correlation coefficient for FEVj (R =
10      0.91), relatively low but statistically significant R values for symptoms ranged from 0.37 to 0.77,
11      with that for sRaw being 0.54. The reproducibility correlations for fB (R = -0.20) and VT
12      (R = -0.03) were not  statistically significant.
13           The effect of this large intersubject variability on the ability to predict individual
14      responsiveness to O3 was demonstrated by McDonnell et al.  (1993).  These investigators
15      analyzed the data of 290 male subjects (18 to 32 years of age) who underwent repeat 2 h IE
16      exposures to one or more O3 concentrations ranging from 0.12 to 0.40 ppm in order to identify
17      personal characteristics (i.e., age, height, baseline pulmonary functions, presence  of allergies,
18      and past smoking history) that might predict individual differences in FEVj response.  Only age
19      contributed significantly to intersubject responsiveness (younger subjects were more
20      responsive), accounting for just 4% of the observed variance. Interestingly, O3 concentration
21      accounted for only 31% of the variance, strongly suggesting the importance of as  yet undefined
22      individual characteristics that determine FEVj responsiveness to O3. A more general form of
23      this model was developed to investigate the O3 exposure FEVj response relationship (McDonnell
24      et al.,  1997). These authors used data from 485 male subjects (age = 18 to  36 years) exposed
25      once for 2 h to  one of six O3 concentrations (ranging from 0.0 to 0.40 ppm) at one of 3 activity
26      levels (rest, n= 78; moderate IE, n = 92; or heavy IE, n = 314).  In addition to investigating the
27      influence of subject's age, the model focused on determining whether FEVj response was more
28      sensitive to changes in C than to changes in VE, and whether the magnitude of responses is
29      independent of differences in lung size.  It was found that the unweighted proportion of the
30      variability in individual responses explained by C, VE, T,  and age was 41%, with no evidence
31      that the sensitivity of  FEVj response to VE was different than changes in C, and no evidence that

        January 2005                             AX6-41      DRAFT-DO NOT QUOTE OR CITE

-------
 1      magnitude of response was related to measures of body or lung size. The authors concluded that
 2      much inter-individual variability in FEVj response to O3 remains unexplained.
 3
 4
 5      AX6.5     INFLUENCE OF AGE, GENDER, ETHNIC, ENVIRONMENTAL
 6                 AND OTHER FACTORS
 7      AX6.5.1   Influence of Age
 8          On the basis of results reported from epidemiologic studies, children and adolescents are
 9      considered to be at increased risk, but not necessarily more responsive, to ambient oxidants than
10      adults. However, findings of controlled laboratory studies that have examined the acute effects
11      of O3 on children and adolescents do not completely support this assertion (Table AX6-5).
12      Children experience about the same decrements in spirometric endpoints as young adults
13      exposed to comparable O3 doses (McDonnell et al., 1985b; Avol et al., 1987).  In contrast to
14      young adults, however, they had no symptomatic response, which may put them at an increased
15      risk for continued exposure.  Similarly, young adults (Linn et al., 1986; Avol et al., 1984) have
16      shown comparable spirometric function response when exposed to low O3 dose under similar
17      conditions.  Among adults, however, it has been repeatedly demonstrated that older individuals
18      respond to O3 inhalation with less intense lung function changes than younger adults. Thus,
19      children, adolescents, and young adults appear to be about equally responsive to O3, but more
20      responsive than middle-aged and older adults when exposed to a comparable dose of O3 (U.S.
21      Environmental Protection Agency, 1996).
22          Gong et al. (1997a)  studied ten healthy men (60 to 69 years old) and nine  COPD patients
23      (59 to 71 years old) from  the Los Angeles area who were exposed to 0.24 ppm O3 while
24      intermittently exercising every 15 min at a light load (-20 L/min) for 4 h. 2Healthy subjects
25      showed a small but significant O3-induced FEVj decrement of 3.3% (p = 0.03 [not reported in
26      paper] paired-t on O3 versus FA pre-post FEVj). Small but statistically nonsignificant changes
27      were also observed for respiratory symptoms, airway resistance and arterial O2 saturation.  In the
28      COPD patients, there was an 8% FEVj decrement due to O3 exposure which was not
              2Personal communication from authors, correction to Table 2 in Gong et al. (1997a), the %FEV{ change at
       the end of the ozone exposure for subject ID 2195 should read 4.9 and not the published value of -4.3, the mean and
       standard deviation reported in the table are correct.

       January 2005                            AX6-42      DRAFT-DO NOT QUOTE OR CITE

-------
                             Table AX6-5. Age Differences in Pulmonary Function Responses to Ozonea
to
o
o
X
H

6
o


o
H

O
o
HH
H
W
Ozone
Concentration1"
ppm
0.40


0.42


0.0
0.40



0.0
0.24



0.0
0.12
0.18
0.24
0.30
0.40

0.0
0.12
0.18
0.24
0.30
0.40



ug/m3
784


823


0
784



0
470



0
235
353
471
589
784

0
235
353
471
589
784



Exposure Duration
and Activity
2 h IE (15' ex/1 5' rest)
VE - 33-45 L/min
(47 subjects only)
1. 5 h IE (20' ex/10' rest)
VE * 33-45 L/min
(All subjects)
2 h, IE (15' ex/1 5' rest)
VE- 18 L/min/m2 BSA
2 exposures: 25% of subj .
exposed to air-air,
75% exposed to O3-O3
4 h, IE (15' ex/ 15' rest)
VE =20 L/min



2 h rest or IE
(4x15 min
at VE = 25 or 35
L/min/m2 BSA)



2.33h IE
(4x15 min
atVE = 25
L/min/m2 BSA)





Exposure
Conditions
-22 °C
40% RH
treadmill



21 °C
40% RH
treadmill


24°C
40% RH



22 °C
40% RH





22 °C
40% RH







Number and
Gender of
Subjects
146 M
94 F




28 M
34 F



10 M

9M


485 WM
(each subject
exposed at one
activity level
to one O3
concentration)

371 (WM,
BM, WF, BF;
-25% per
group) each
subject
exposed to one
03
concentration

Subject
Characteristics
Healthy NS
18 to 60 years old




Healthy NS
18 to 57 years old
Healthy NS
18 to 59 years old

Healthy NS
60 to 69 years old
COPD
59 to 71 years old

Healthy NS
18 to 36 years old
mean age 24 years




Healthy NS
18 to 35 years old
mean age 24 years






Observed Effect(s)
Young individuals of both gender (< 35 years)
significantly more responsive than older
subjects. Strong responses are less common
over the age of 35 years, especially in women.
The variability of an individual's responsiveness
to repeated exposures to O3 decreases with age.
Significant decrements in spirometric lung
function in all groups. Young males and females
(< 35 years) were significantly more responsive
than older individuals (> 35 years).

Healthy: small, 3.3%, decline in FEVj (p=0.03 [not
reported in paper], paired-t on O3 versus FA pre-post
FEVj). COPD: 8% decline in FEVj (p=ns, O3 versus
FA). Adjusted for exercise, ozone effects did not differ
significantly between COPD patients and healthy
subjects.
Statistical analysis of 8 experimental chamber
studies conducted between 1980 and 1993 by the
U.S. EPA in Chapel Hill, NC. O3-induced
decrement in FEV[ predicted to decrease with
age. FEV; response of a 30 year old predicted to
be 50% the response of a 20 year old. Also see
Table 6-1
Statistical analysis of experimental data
collected between 1983 and 1990 by the U.S.
EPA in Chapel Hill, NC. Cyinduced decrement
in FEV; predicted to decrease with age. FEV;
response of a 30 year old predicted to be 65%
the response of a 20 year old. No effect of
menstrual cycle phase on FEV; response.
Inconsistent effect of social economic status on
FEV[ response.
Reference
Hazucha et al.
(2003)




Passannante
etal. (1998)



Gong et al.
(1997a)



McDonnell
etal. (1997)





Seal et al.
(1996)








-------
3
to
o
o













X
rv
j\




O
>
H
6
o
2?
O
H
O
O
m
rn
O


Ozone
Concentration1"
ppm
0.18
0.24
0.30
0.40
0.45



0.45




0.45





0.12


0.20
0.30




0.113C +
other
ambient
pollutants
ug/m3
353
470
588
784
882



882




882





235


392
588




221



xauii; r^^w-«j ^'
Exposure Duration
and Activity
2.33h
IE
VE = 20 L/min/m2 BSA

Ih, CE
VE * 26 L/min
2h,IE
VE * 26 L/min
2 h, IE (20' ex/20' rest)
Male: VE = 28.5L/min
Female: VE = 26.1 L/min


2 h, IE (20' ex/20' rest)
VE - 26 L/min




1 h IE (mouthpiece)
VE = 4 to 5 x resting

1 h (mouthpiece)
50' rest/10' ex for first 7
males, 20' rest/10' ex for
remaining subjects
Male: VE - 29 L/min
Female: VE - 23 L/min
1 h CE (bicycle)
VE - 22 L/min


k^uui uy* -f»gc i^inci (;iiv^9
Number and
Exposure Gender of
Conditions Subjects
NA 48 WF, 55 BF



-23 °C 7M
58% RH 5 F
cycle/treadmill

23 °C 10 M,
46% RH 6 F
cycle/treadmill


-24 °C 8 M
63% RH
cycle
8F


22 °C 5 M, 7 F
75% RH
treadmill
-22 °C 9M, 10 F
> 75% RH
treadmill



32.7 °C 33 M, 33 F
-43% RH
cycle

111 i uiiiiirii
-------
                              Table AX6-5 (cont'd). Age Differences in Pulmonary Function Responses to Ozone"
to
o
o
Ozone
Concentration1"
ppm
0.45
0.12
0.18

exposure Liuranon
u,g/m3 and Activity
882 2 h, IE (20' ex/20' rest)
VE « 26 L/min
235 40 min (mouthpiece)
IE, 10 min exercise at
VE = 32.6 L/min
353 40 min (mouthpiece)
IE, 10 min exercise at
VE = 41.3L/min
Number and
Exposure Gender of
Conditions Subjects
-23 °C 8 M,
53% RH 8 F
cycle
NA 3 M, 7 F
treadmill
4M, 6F

Subject
Characteristics Observed Effect(s)
Healthy NS, Mean decrement in FEV; = 5 .6 ± 1 3%; range of
51 to 76 years old decrements = 0 to 12%.
Healthy NS, No significant change in FEV^ increased RT
14 to 19 years old with exposure to 0.18 ppm O3. Some subjects
responded to 5 to 10 mg/mL methacholine after
0.18-ppm O3 exposure, whereas none responded
to 25 mg/mL methacholine at baseline
bronchochallenge.
Reference
Drechsler-
Parks et al.
(1987a,b)
Koenig et al.
(1987)

X
        "See Appendix A for abbreviations and acronyms.

        bListed from lowest to highest O3 concentration.

        GOzone concentration is the mean of a range of ambient concentrations.
H

6
o


o
H

O
o
HH
H
W

-------
 1      significantly different from the response in the healthy subjects. The authors have concluded
 2      that typical ambient concentrations of O3 are unlikely to induce "a clinically significant acute
 3      lung dysfunction" in exposed older men. However, they also acknowledged that the "worst
 4      case" scenario of O3 exposure used in their study causes acute spirometric responses.
 5           Although Gong et al. (1997a) and others (see Table 6-5) have examined responses to O3
 6      exposure in subjects of various ages, the exposure conditions differ between most studies
 7      so that age effects remain uncertain.  Three recent studies, which analyzed large data sets
 8      (> 240 subjects) of similarly exposed subjects, show clearly discernable changes in FEVj
 9      responses to O3 as a function of age.
10           Seal et al. (1996) analyzed O3-induced spirometric responses in 371 young nonsmokers
11      (18 to 35 years of age).  The subject population was approximately 25% white males, 25% white
12      females, 25% black males, and 25% black females.  Each subject was exposed once to  0.0,  0.12,
13      0.18, 0.24, 0.30, or 0.40 ppm ozone for 2.3 h during IE at a VE  of 25 L/min/m2 BSA. A logistic
14      function was used to model and test the significance of age, socioeconomic status (SES), and
15      menstrual cycle phase as predictors of FEVj response to O3 exposure. Menstrual cycle phase
16      was not a significant. SES was inconsistent with the greatest response observed in the  medium
17      SES and the lowest response in high  SES. FEVj responses decreased with subject age. On
18      average, regardless of the O3 concentration, the response of 25, 30, and 35 year old individuals
19      are predicted to be 83, 65, and 48% (respectively) of the response in 20 year olds.  For  example,
20      in 20 year old exposed to 0.12 ppm ozone (2.3 h IE, VE = 25 L/min/m2 BSA) a 5.4% decrement
21      in FEVj is predicted, whereas, a  similarly exposed 35 yr old is only predicted to have a 2.6%
22      decrement. The Seal et al. (1996) model is limited to predicting FEVj responses immediately
23      postexposure in individuals exposed for 2.3 h during IE at a VE of 25 L/min/m2 BSA.
24           McDonnell et al. (1997)  examined FEVj responses in 485 healthy white males (18 to
25      36 years of age) exposed once for 2 h to an O3 concentration of 0.0, 0.12, 0.18, 0.24, 0.30, or
26      0.40 ppm at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA).  FEVj was
27      measured preexposure, after 1  h  of exposure, and immediately postexposure.  Decrements in
28      FEVj were modeled by sigmoid-shaped curve as a function of subject age, O3 concentration, VE,
29      and duration of exposure. Regardless of the O3 concentration or duration of exposure,  the
30      average responses of 25, 30, and 35 year old individuals are predicted to be 69, 48,  and 33%


        January 2005                            AX6-46      DRAFT-DO NOT QUOTE OR CITE

-------
 1     (respectively) of the response in 20 year olds. The McDonnell et al. (1997) model is best suited
 2     to predicting FEVj responses in while males exposed to O3 for 2 h or less under IE conditions.
 3          Hazucha et al. (2003) analyzed the distribution of O3 responsiveness in subjects (146 M,
 4     94 F) between 18 and 60 years of age.  Subjects were exposed to 0.42 ppm O3 for 1.5 h with IE
 5     at VE = 20 L/min/m2 BSA. Figure AX6-7 illustrates FEVj responses to O3 exposure as a
 6     function of subject age. Consistent with the discussion in Section 6.4, a large degree of
 7     intersubject variability is evident in Figure AX6-7. Across all ages, 18% of subjects were weak
 8     responders (<5% FEVj decrement), 39% were moderate responders, and 43% were strong
 9     responders (> 15% FEVj decrement). Younger subjects (<35 years of age) were predominately
10     strong responders, whereas,  older subjects (>35 years of age) were mainly weak responders.
11     In males, the FEVj responses of 25, 35, and 50 year olds are predicted to be 94, 83, and 50%
12     (respectively) of the average response in 20 year olds. In females, the FEVj responses of 25, 35,
13     and 50 year olds are predicted to be 82, 46, and 18% (respectively) of the average response in
14     20 year olds. The Hazucha et al. (1996) model is limited to predicting FEVj responses
15     immediately postexposure in individuals exposed to 0.42 ppm O3 for 1.5 h during IE at a VE of
16     20 L/min/m2 BSA.
           110-
           100-
         -5T 9°H
         
-------
 1           The pathophysiologic mechanisms behind the pronounced age-dependent, gender-
 2      differential rate of loss of O3 responsiveness are unclear. Passannante et al. (1998) have
 3      previously demonstrated that O3-induced spirometric decrements (FEVj) in healthy young and
 4      middle-aged adults are principally neural in origin, involving opioid-modulated sensory
 5      bronchial C-fibers. (The methodological details of this study are presented in Section AX6.2.3 of
 6      this chapter}  The peripheral afferents are most likely the primary site of action, which would be
 7      compatible with a reflex action as well as a cortical mechanism. The pattern of progressive
 8      decline, as well as the subsequent rate of recovery of spirometric lung function, suggest
 9      involvement of both direct and indirect (possibly by PGE2a) stimulation and/or sensitization of
10      vagal sensory fibers.  (For details, see Section AX6.2.3.1 of this chapter }
11           The additional pulmonary function  data published since the release of last O3 criteria
12      document (U.S. Environmental Protection Agency, 1996) and reviewed in this section reinforce
13      the conclusions reached in that document. Children and adolescents are not more responsive to
14      O3 than young adults when exposed under controlled laboratory conditions. However, they are
15      more responsive than middle-aged and older individuals.  Young individuals between the age of
16      18 and 25 years appear to be the most sensitive to O3. With progressing age, the sensitivity to O3
17      declines and at an older age (> 60 yrs) appears to be minimal except for some very responsive
18      individuals. Endpoints other than FEVj may show a different age-related pattern of
19      responsiveness.
20
21      AX6.5.2   Gender and Hormonal  Influences
22           The few late 1970 and early 1980 studies specifically designed to determine symptomatic
23      and lung function responses of females to O3 were inconsistent. Some studies have concluded
24      that females might be more sensitive to O3 than  males, while others found no gender differences
25      (U.S. Environmental Protection Agency,  1996). During the subsequent decade, seven studies
26      designed to systematically explore gender-based differences in lung function following O3
27      exposure were completed (Table  AX6-6). Protocols included mouthpiece and chamber
28      exposures, young and old individuals, normalization of ventilation to BSA or FVC, continuous
29      and intermittent exercise, control for menstrual cycle phase, and the use of equivalent effective
30      dose of O3 during exposures. These studies have generally reported no statistically significant
31      differences in pulmonary function between males and females (Adams et al., 1987; Drechsler-

        January 2005                             AX6-48     DRAFT-DO NOT QUOTE OR CITE

-------
to
o
o
                    Table AX6-6.  Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
Ozone
Concentration11
ppm ug/m3 and Activity Conditions'
0.0 490 1 h CE NA
0.25 VE = 30 L/min Face mask
exposure







Number and Gender Subject
of Subjects Characteristics
32 M, 28 F Healthy NS
22.6 ±0.6 years old








Observed Effect(s)
Mean O3-induced FEVj decrements of 15.9%
in males and 9.4% in females (gender
differences not significant). FEV; decrements
ranged from -4 to 56%; decrements >15% in
20 subjects and >40% in 4 subjects. Uptake
of O3 greater in males than females, but
uptake not correlated with spirometric
responses.


Reference
Ultman et al.
(2004)






X
VO
H

6
o


o
H

O
o
HH
H
W
0.40

0.42


0.0
0.35





0.0
0.4



0
0.12
0.24
0.30
0.40

0.0
0.35



784

823


0
686





0
784



0
235
470
588
784

0
686



2 h, IE (15' ex/1 5' rest)
VE = 33-45 L/min
1. 5 h IE (20' ex/10' rest)
VE = 33-45 L/min

1.25h, IE
(30' ex/1 5' rest/30' ex)
VE = 40 L/min




2 h, IE (15' ex/1 5' rest)
VE» 18L/min/m2BSA
2 exposures: 25% of
subj. exposed to air-air,
75% exposed to O3-O3
2.33 h IE (15' ex/15' rest)
VE = 20 L/min/m2 BSA
one exposure per subject



2.15 h, IE
(30' ex/30' rest)



22 °C 146 M
40% RH 94 F
treadmill


22 °C 19 F
40% RH
treadmill




21 °C 28 M
40% RH 34 F
treadmill


22 °C 48 WF, 55 BF
40% RH
treadmill



19-24 °C 12 M
48-55% RH 12 F
treadmill


Healthy NS,
1 8 to 60 years old



O3 responders
22.1 ±2. 7 years old





Healthy NS, 20-59 years old




Healthy NS,
18 to 35 years old




Healthy NS,
5 F follicular and 7 luteal
phase exposure,
regular menstrual cycles, 18
to 35 years old
No significant gender differences in FEVj
among young (< 35 years) and older
individuals. Strong responses are less
common over the age of 35 years, especially
in women.
FVC and FIVC changes about the same,
-13%, FEVj -20%. Increased airway
responsiveness to methacholine. Persistence
of small effects on both inspired and expired
spirometry past 18 h. Chemoreceptors not
activated but ventilatory drive was
accelerated.
Significant decrements in spirometric lung
function. No significant differences in FEV;
between young females and males and older
females and males either in responders or
non-responders subgroups.
Significant menstrual cycle phase * race
interaction for FEVj. No significant
menstrual cycle phase effect when blacks and
whites were analyzed separately.
No significant menstrual phase effects for
SRaw or cough score.
Changes in FVC, FEV,, FEF25.75, Vmaj[50%, and
Vmax25% were similar during both the follicular
and luteal phases. No significant difference
between males and females.

Hazucha
et al. (2003)



Folinsbee
and Hazucha
(2000)




Passannante
etal. (1998)



Seal et al.
(1996)




Weinmann
etal. (1995)




-------
to
o
o
                       Table AX6-6 (cont'd). Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
Ozone
Concentration11
ppm ug/m3
0.3 588

and Activity
1 hCE
VE = 50 L/min
Exposure Number and Gender Subject
Conditions' of Subjects Characteristics
NA 9 F Healthy NS, regular
menstrual cycles,
20 to 34 years old
Observed Effect(s)
FEVj decreased 13.1% during the mid-luteal
phase and 18.1% during the follicular phase.
Decrement in FEF25.75 was significantly larger
Reference
Fox et al.
(1993)
                                                                                                                   during the follicular phase than the mid-luteal
                                                                                                                   phase. Changes in FVC were similar in
                                                                                                                   both phases.
X
ON
(^
O
 H
 6
 o
 o
 H
O
         o
         0.12
         0.18
         0.24
         0.30
         0.40
         0.0
         0.45
         0.3
0
235
353
470
588
784
         0      0
         0.18   353
         0.30   588
0
882
         0      0
         0.20   392
         0.30   588
                588
2.33 h (15'ex/15'rest)
VE = 25 L/min/m2 BSA
(one exposure/subject)
          1 h (mouthpiece), CE
          VE = 47 L/min
          exposures > 4 days apart
2 h, IE (20' ex/20' rest)
VE = 28.5 L/min for M
VE = 26.1 L/min for F
repeated O3 exposures
22 °C
40% RH
treadmill
                        21 to 25 °C
                        45 to 60% RH
                        cycle
 23.1 °C
46.1%RH
cycle/treadmill
30 to 33 F and 30 to    Healthy NS, 18 to 35 years    Decrements in FEV1; increases in SRaw and    Seal et al.
          1 h (mouthpiece) IE        -22 °C
          (50'rest/10'ex first 7 M)    > 75% RH
          (20' rest/10' ex all others)    treadmill
33 M in each
concentration group;
total of
372 individuals
participated

14 F
                                                   14 F
                                                                   10M
                                                                   6F
                                          9M, 10 F
          1 h (mouthpiece), CE
                                                 21 to 25 °C
                                                 45 to 60% RH
                                                 cycle
                                         20 M

                                         20 F
old, blacks and whites
                                       FVC = 5.11 ± 0.53 L,
                                       NS, 20 to 24 years old

                                       FVC = 3.74 ± 0.30 L,
                                       NS, 19 to 23 years old
                     Healthy NS,
                     60 to 89 years old

                     Healthy NS,
                     64 to 71 years old

                     Healthy NS,
                     55 to 74 years old
                                       NS, 18 to 30 years old

                                       NS, 19 to 25 years old
                                                                                                                   cough, correlated with O3 concentration.        (1993)
                                                                                                                   There were no significant differences between
                                                                                                                   the responses of males and females.
                           Small lung group, FVC = 3.74 ± 0.30 L.        Messineo
                           Large lung group, FVC = 5.11 ± 0.53 L.        and Adams
                           Significant concentration-response effect on     (1990)
                           FVC and FEVjj lung size had no effect on
                           percentage decrements in FVC or FEVj.
                           Mean decrement in FEVj = 5.7%.             Bedi et al.
                           Decrements in FVC and FEV; were the only    (1989)
                           pulmonary functions significantly altered by
                           O3 exposure. No significant differences
                           between responses of men and women.

                           No change in any spirometic measure for       Reisenauer
                           either group. Females had 13% increase in RT   et al. (1988)
                           after 0.30-ppm exposure. Gender differences
                           not evaluated.
                                                Significant decrements in FVC, FEV1; and
                                                FEF25.75 following O3 exposure.
                                                No significant differences between men and
                                                women for spirometry or SRaw.
                                                                    Adams et al.
                                                                    (1987)
o
HH
H
W

-------
                   Table AX6-6 (cont'd).  Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
c
!-J
-<
to
o
o










Ozone
Concentration11


ppm ug/m3 and Activity
0.0 0 2 h, IE (20' rest/ 20' ex)
0.45 882 VE = 27.9 L/min for M
VE»25.4L/minforF
repeated O3 exposures

0.48 941 2h, IE
VE = 25 L/min





Exposure
Conditions'
24 °C
58% RH
cycle


21 °C
(WBGT)
cycle




Number and Gender Subject
of Subjects Characteristics
8M Healthy NS,
5 1 to 69 years old

8F Healthy NS,
56 to 76 years old
10 F Healthy NS,
19 to 36 years old






Observed Effect(s)
Range of responses in FEVj:
0 to - 12% (mean = -5.6%).
No significant difference in responses of men
and women.
Tendency for women to have greater effects.
Mean decrement in FEV; = 22.4%.
Significant decrements in all spirometric
measurements. Results not significantly
different from a similar study on males
(Drechsler-Parks et al., 1984).



Reference
Drechsler-
Parks et al.
(1987a,b)


Horvath
etal. (1986)



X
H
6
o

o
H
O
        " See Appendix A for abbreviations and acronyms.

        b Listed from lowest to highest O3 concentration.

        'WBGT = 0.7 Twetbulb + 0.3 T.
o
HH
H
W

-------
 1     Parks, et al., 1987a; Messineo and Adams, 1990; Seal et al., 1993; Weinmann et al., 1995)
 2     although in some studies females appeared to experience a slightly greater decline then males
 3     (Drechsler-Parks et al., 1987a; Messineo and Adams, 1990). The comparative evaluations were
 4     based on responses that included spirometry, airway resistance, non-specific bronchial
 5     responsiveness (NSBR) determinations, and changes in frequency and severity of respiratory
 6     symptoms. However, depending on how the O3 dose was calculated and normalized, the
 7     findings of at least three studies may be interpreted as showing that females are more sensitive to
 8     O3 than males. The findings of the seven studies are presented in detail in Section 7.2.1.3 of the
 9     previous O3 criteria document (U.S. Environmental Protection Agency, 1996).
10           Some support for a possible increased sensitivity of females to O3 comes from a study of
11     uric acid concentration in nasal lavage fluid (NLF).  Housley et al. (1996) found that the NLF of
12     females contains smaller amounts of uric acid than the NLF of males. The primary source of
13     uric acid is plasma; therefore,  lower nasal concentrations would reflect lower plasma
14     concentrations of this antioxidant. The authors have speculated that in females, both lower
15     plasma and NLF levels (of uric acid) can plausibly make them more susceptible to oxidant
16     injury, since local antioxidant protection may not be as effective as with higher levels of uric
17     acid, and consequently more free O3  can penetrate deeper into the lung.
18           Several studies also have suggested that anatomical differences in the lung size and the
19     airways between males and  females,  and subsequent differences in O3 distribution and
20     absorption, may influence O3 sensitivity and potentially differential O3 response. The study of
21     Messinio and Adams (1990) have, however, convincingly demonstrated that the effective dose to
22     the lung, and not the lung size, determines the magnitude of (FEVj) response. Furthermore, the
23     O3 dosimetry experiments of Bush et al. (1996) have shown that despite gender differences in
24     longitudinal distribution of O3, the absorption distribution in conducting airways was the same
25     for both sexes when expressed as a ratio of penetration to anatomic dead space volume. This
26     implies that gender differences, if any, are not due to differences in (normal) lung anatomy.
27     The data also have shown that routine adjustment of O3 dose for body size and gender
28     differences would be more important if normalized to anatomic dead space rather than the usual
29     FVCorBSA.
30           One of the  secondary objectives of a study designed to examine the role of neural
31     mechanisms involved in limiting maximal inspiration following O3 exposure has been to

       January 2005                            AX6-52     DRAFT-DO NOT QUOTE OR CITE

-------
 1      determine if gender differences occur.  A group of healthy males (n = 28) and females (n = 34)
 2      were exposed to 0.42 ppm O3 for 2 h with IE. The methodological details of the study are
 3      presented in Section AX6.2.5.1 of this document. As Figure AX6-4 shows, the differences
 4      between males and females were, at any condition, measurement point, and O3 sensitivity status
 5      only minimal and not significant (Passannante et al., 1998).
 6           In another investigation, Folinsbee and Hazucha (2000) exposed a group of
 7      19 O3-responsive young females (average age of 22 years, prescreened for O3  responsiveness by
 8      earlier exposure) to air and 0.35 ppm O3. The randomized 75-min exposures included two
 9      30-min exercise periods at a VE of 40 L/min. In addition to  standard pulmonary function tests,
10      they employed several techniques used for the first time in human air pollution studies
11      assessment of O3 effects. The average lung function decline from a pre-exposure value was  13%
12      for FVC, 19.9 % for FEVj, and 30% for FEF25.75. The infrequently measured forced inspiratory
13      vital capacity (FIVC) was the same as FVC suggesting that the lung volume limiting
14      mechanisms are the same. The reduction in peak inspiratory flow (PIF) most likely reflects an
15      overall reduction in inspiratory effort associated with neurally mediated inhibition of inspiration.
16      Persistence  of small inspiratory and expiratory spirometric effects, airway resistance, and airway
17      responsiveness to methacholine for up to 18 h postexposure suggests that recovery of pulmonary
18      function after O3 exposure involves more than the simple  removal of an irritant.  Incomplete
19      repair of damaged epithelium and still unresolved airway  inflammation are the likely causes of
20      the residual  effects that in some individuals persisted beyond 24 h postexposure.  However, by
21      42 hours no residual effects were detected. No significant changes were found in ventilatory
22      response to  CO2 between air and O3 exposures, suggesting that chemoreceptors were not affected
23      by O3. However, O3 inhalation did result in accelerated timing of breathing and a modest
24      increase in inspiratory drive. These observations are consistent with, and further supportive of,
25      the primary  mechanisms of O3-induced reduction in inspiratory lung function, namely an
26      inhibition of inspiration elicited by stimulation of the C-fibers and other pulmonary receptors.
27      Because the measures of inspiratory and chemical drive to assess  O3 effects were not reported in
28      any previous human study, no comparisons are possible.  Because no male subjects were
29      recruited for the study, it is not possible to compare gender effects. Despite being O3-responsive,
30      however, the average post-O3 decline in expiratory lung function from preexposure (13% for
31      FVC; 19.9% for FEV^ 30% for FEF25.75) was similar to that seen in female cohorts studied by

        January 2005                            AX6-53      DRAFT-DO NOT QUOTE OR CITE

-------
 1      other investigators under similar conditions of exposure. These were the same studies that found
 2      no gender differences in O3 sensitivity (Adams et al., 1987; Messineo and Adams, 1990).
 3           The study by Hazucha et al. (2003), discussed in the previous section, has in addition to
 4      aging also examined gender differences in O3 responsiveness. The male (n = 146) and female
 5      (n = 94) cohorts were classified into young (19 to 35 year-old) and middle-aged (35 to 60 year-
 6      old) groups.  This classification was selected in order to facilitate comparison with data reported
 7      previously by other laboratories. Using a linear regression spline model (with a break point at
 8      35 years), the authors reported that the rate of loss of sensitivity is about three times as high in
 9      young females as in young males (p< 0.003). In young females, the average estimated decline in
10      FEVj response is 0.71% per year, while in young males it is 0.19% per year. Middle-aged
11      groups of both genders show about the same rate of decline (0.36 to 0.39%, respectively).
12      At 60 years of age, the model estimates  about a 5% post-O3 exposure decline in FEVj for males,
13      but only a 1.3% decline for females. These observations suggest that young females lose O3
14      sensitivity faster than young males, but by middle age, the rate is about the same for both
15      genders. Descriptive statistics show that there were practically no differences in the mean value,
16      standard error of the mean,  and coefficient of variation for % FEVj decrement between the group
17      of young males (n = 125; 83.7 ± 1.1%; CV = 13.5%) and young females (n = 73;  83.4 ± 1.25%;
18      CV = 12.8%). A straight linear regression model of these data was illustrated in Figure AX6-7.
19      The slopes, significant in both males (r = 0.242; p = 0.003) and females (r = 0.488; p = 0.001),
20      represent the decline in responsiveness of 0.29% and 0.55% per year respectively, as assessed by
21      FEVi.
22           Two earlier studies of the effects of the menstrual cycle phase on O3 responsiveness have
23      reported conflicting results  (U.S. Environmental  Protection Agency, 1996). Weinmann et al.
24      (1995) found no significant lung function effects related to menstrual cycle, although during the
25      luteal phase the effects were slightly more pronounced than during the follicular phase; while
26      Fox et al., (1993) reported that follicular phase enhanced O3 responsiveness. In a more recent
27      investigation of possible modulatory effects of hormonal changes during menstrual cycle on O3
28      response, young women (n = 150) 18 to 35 years old were exposed once to one of multiple O3
29      concentrations (0.0, 0.12, 0.18, 0.24, 0.30, 0.40 ppm) for 140 min with IE at 35 L/min/m2 BSA.
30      The women's menstrual cycle phase was determined immediately prior to O3 exposure.  Post-O3,
31      no significant differences in % predicted FEVj changes that could be related to the menstrual

        January 2005                            AX6-54     DRAFT-DO NOT QUOTE OR CITE

-------
 1      cycle phase were found. Admittedly, a less precise method of determining menstrual cycle
 2      phase used in this study could have weakened the statistical power.  Unfortunately, the direction
 3      and magnitude of O3 response as related to the menstrual cycle phases were not reported (Seal
 4      et al., 1996).  Considering the inconclusiveness of findings of this study and the inconsistency of
 5      results between the two earlier studies, it is not possible to make any firm conclusions about the
 6      influence of the menstrual cycle on responses to O3 exposure.
 7           Additional studies presented in this section clarify an open-ended conclusion reached in the
 8      previous O3 criteria document (U.S Environmental Protection Agency, 1996) regarding the
 9      influence of age on O3 responsiveness. Healthy young males and females are about equally
10      responsive to O3, although the rate of loss of sensitivity is higher in females than in males.
11      Middle-aged men and women are generally much less responsive to O3 than younger individuals.
12      Within this range, males appear to be slightly more responsive than females, but the rate of age-
13      related loss in FEVj is about the same. The O3 sensitivity may vary during the menstrual cycle;
14      however, this variability appears to be minimal.
15
16      AX6.5.3 Racial, Ethnic and Socioeconomic Status Factors
17           In the only laboratory study designed to compare spirometric responses of whites and
18      blacks exposed to a range of O3 concentrations (0 to 0.4 ppm), Seal et al. (1993) reported
19      inconsistent and statistically insignificant FEVj differences between white and black males and
20      females within various exposure levels. Perhaps, with larger cohorts the tendency for greater
21      responses of black than white males may become significant. Thus, based on this  study it is still
22      unclear if race is a modifier of O3 sensitivity,  although the findings of epidemiologic studies
23      reported in the previous criteria document "can be considered suggestive of an ethnic difference"
24      (U.S. Environmental Protection Agency, 1996). However, as Gwynn and Thurston (2001)
25      pointed out, it appears that it is more the socioeconomic status (SES) and overall quality of
26      healthcare that drives PM10- and O3-related hospital admissions than an innate or acquired
27      sensitivity to pollutants.
28           This assertion is somewhat supported by the study of Seal et al. (1996) who employed a
29      family history questionnaire to examine the influence of SES on the O3 responsiveness of
30      352 healthy,  18- to 35-year-old black and white subjects.  Each subject was exposed  once under
31      controlled laboratory conditions to either air or 0.12, 0.18, 0.24, 0.30,  0.40 ppm O3 for 140 min

        January 2005                           AX6-55       DRAFT-DO NOT QUOTE OR CITE

-------
 1      with 15 min IE at 35 L/min/m2 BSA. An answer to the "Education of the father" question was
 2      selected as a surrogate variable for SES status.  No other qualifying indices of SES were used or
 3      potential bial examined.  Of the three SES categories, individuals in the middle SES category
 4      showed greater concentration-dependent decline in % predicted FEVj (4-5% @ 0.4 ppm O3) than
 5      low and high SES groups. The authors did not have an "immediately clear" explanation for this
 6      finding.  The SES to %predicted FEVj relationship by gender-race group was apparently
 7      examined as well; however, these results were not presented.  Perhaps a more comprehensive
 8      and quantitative evaluation of SES status would have identified the key factors and clarified the
 9      interpretation of these findings. With such a paucity of data it is not possible to discern the
10      influence of racial or other related factors on O3 sensitivity.
11
12      AX6.5.4   Influence of Physical Activity
13           Apart from the importance of increased minute ventilation on the inhaled dose of O3 during
14      increased physical activity, including work, recreational exercise, and more structured exercise
15      like sports, no systematic effort has been made to study other potential physical factors that may
16      modulate O3 response. The typical physiologic response of the body to exercise is to increase
17      both the rate and depth of breathing, as well as increase other responses such as heart rate, blood
18      pressure, oxygen uptake, and lung diffusion capacity.
19           Physical activity increases minute ventilation in proportion to work load.  At rest, and
20      during light exercise, the dominant route of breathing is through the nose. The nose not only
21      humidifies air, among other physiologic functions, but also absorbs O3 thus decreasing the
22      overall dose. As the intensity of exercise increases, the minute ventilation increases and the
23      breathing switches from nasal to oronasal mode. There is considerable individual variation in
24      the onset of oronasal breathing, which ranges from 24 to 46 L/min (Niinimaa et al., 1980).
25      During heavy exercise, ventilation is dominated by oral breathing.  Consequently, the residence
26      time of inhaled air in the nose and the airways is shorter, reducing the uptake of O3 (Kabel et al.,
27      1994). Moreover, increasing inspiratory flow and tidal volume  shifts the longitudinal
28      distribution of O3 to the peripheral airways, which are more sensitive  to injury than the larger,
29      proximal airways.  Ozone uptake studies of human lung showed that at simulated quiet
30      breathing, 50% of O3 was absorbed in the upper airways, 50% in the conducting airways, and
        January 2005                             AX6-56      DRAFT-DO NOT QUOTE OR CITE

-------
 1      none reached the small airways (Hu et al. 1994). With ventilation simulating heavy exercise
 2      (60 L/min), the respective O3 uptakes were 10% (upper airways), 65% (conducting airways), and
 3      25% (small airways).  These observations imply that equal O3 dose (C x T x VE) will have a
 4      greater effect on pulmonary function and inflammatory responses when inhaled during heavy
 5      physical activity than when inhaled during lighter activity. Although, Ultman et al. (2004)
 6      recently reported that spirometric response are not correlated with O3 uptake.  (See Chapter 4 of
 1      this document for more information on the dosimetry ofO3)
 8          Other physiologic factors activated in response to physical activity are unlikely to have as
 9      much impact on O3 responsiveness as does minute ventilation; however, their potential influence
10      has not been investigated.
11
12      AX6.5.5  Environmental Factors
13          Since the 1996 O3 criteria document not a single human laboratory study has examined the
14      potential influence of environmental factors such as rural versus urban environment, passive
15      cigarette smoke exposure, and bioactive admixtures such as endotoxin on healthy individual's
16      pulmonary function changes due to O3 (U.S. Environmental Protection Agency, 1996).
17          Some of the unresolved issues, e.g., health effects of ETS and O3 interaction, which need to
18      be examined in human studies were explored very recently in laboratory animal studies
19      (presented in a greater detail in Chapter 5). In one study on mice, preexposure of animals to
20      sidestream cigarette smoke (ETS surrogate), which elicited no immediate effects, resulted in a
21      potentiation of subsequent O3-induced inflammatory response.  This finding suggests that typical
22      adverse effects of ETS do not necessarily have to  elicit an immediate response to ETS, but may
23      in fact potentiate the effects of a subsequent exposure to another pollutant like O3 (Yu et al.,
24      2002). The key mechanism by which smoke inhalation may potentiate subsequent oxidant injury
25      appears to be damage to cell membranes and the resulting increase in epithelial permeability.
26      Disruption of this protective layer may facilitate as well as accelerate injury to subepithelial
27      structures when subsequently exposed to other pollutants (Bhalla, 2002).  Although this may be a
28      plausible mechanism in nonsmokers and acute smokers exposed to ETS and other pollutants,
29      studies involving chronic smokers who most likely already have chronic airway inflammation do
30      not seems to show exaggerated response with exposure to O3.
31

        January 2005                            AX6-57      DRAFT-DO NOT QUOTE OR CITE

-------
 1          More than 25 years ago, Hazucha et al. (1973) reported that the spirometric lung function
 2     of smokers declined significantly less than that of nonsmokers when exposed to 0.37 ppm O3.
 3     The findings of this study have been confirmed and expanded (Table AX6-7). Frampton et al.
 4     (1997a) found that exposure of current smokers (n=34) and never smokers (n=56) to 0.22 ppm
 5     O3 for 4 h with IE for 20 min of each 30 min period at 40 to 46 L/min, induced a substantially
 6     smaller decline in FVC, FEVj and SGaw of smokers than never smokers.  Smokers also
 7     demonstrated a much narrower distribution of spirometric endpoints than never smokers.
 8     Similarly, nonspecific airway responsiveness to methacholine was decreased in smokers.
 9     However, both groups showed the consistency of response from exposure to exposure. It should
10     be noted, that despite seemingly lesser response, the smokers were more symptomatic post air
11     exposure than never smokers but the opposite was true for O3 exposure. This would suggest that
12     underlying chronic airway inflammation present in smokers has blunted stimulation of bronchial
13     C-fibers and other pulmonary receptors, the receptors substantially responsible for post O3 lung
14     function decrements. In addition to desensitization, the other "protective" mechanisms active in
15     smokers may be an increase in the mucus layer conferring not only a mechanical protection, but
16     also acting as an O3 scavenger. Another plausible explanation of a diminished responsiveness of
17     smokers may be related to elevated levels of reduced glutathione (GSH), tissue antioxidant,
18     found in epithelial lining fluid of chronic but not acute smokers (MacNee et al., 1996).
19          Despite some differences in a release of proinflammatory cytokines and subsequent
20     recruitment of inflammatory cells, both smokers and nonsmokers developed airway
21     inflammation following O3 exposure. This was demonstrated by the Torres et al. (1997) study
22     that involved exposures of about equal size cohorts of otherwise healthy young smokers,
23     nonsmoker O3 nonresponders (< 5% FEVj post O3 decrement) and nonsmoker O3 responders
24     (> 15% FEVj post O3 decrement) to air and two 0.22 ppm O3 atmospheres for 4 hours,
25     alternating 20 min of moderate exercise (25 L/min/m2 BSA) with 10 min of rest. Both O3
26     exposures were followed by nasal lavage (NL) and bronchoalveolar lavage (BAL) performed
27     immediately post one of exposures and 18 hr later following the other exposure. Neither O3
28     responsiveness nor smoking alters the magnitude or the time course of O3-induced airway
29     inflammation. The overall cell recovery was lower immediately post exposure but higher,
30     particularly in nonsmokers, 18 h post O3 exposure when compared to control (air) in all groups.
31     Recovery of lymphocytes, PMNs and AMs in both alveolar and bronchial lavage fluid showed

       January 2005                           AX6-58     DRAFT-DO NOT QUOTE OR CITE

-------
                                      Table AX6-7. Influence of Ethnic, Environmental, and Other Factors
to
o
o
X
ON

(^
VO
H

6
o


o
H

O
o
HH
H
W
Ozone
Concentration

ppm Hg/m3
0.0 0
0.4 780







0.0 0
0.12 235


Exposure
Duration Exposure
and Activity Conditions
2 h 20°C
IE, 15' ex/15' rest 40% RH
VE = 20 L/min/m2
BSA





0.75 h 60% RH
IE, 15' ex/15' rest
VE = 40-46 L/min

Number
and Gender
of Subjects
15 M, 1 F



13 M, 2 F




5 M, 12 F




Subject
Characteristics
Placebo group:
healthy NS
avg. age 27 yrs.

Antiox. Suppl.
Gr: Healthy NS
avg. age 27 yrs.


Asthmatics
sensitive to SO2
19 to 38 yrs old



Observed Effect(s)
PF decrements in the supplementation
group were signif. smaller for FVC
(p < 0.046) and near significant for FEV{
(p < 0.055). The inflammatory response
(BAL) showed no significant differences
between the two groups either in the
recovery of cellular components or the
concentrations and types of inflammatory
cytokines.
No significant differences due to O3
between placebo and antioxidant
supplement cohort in either PF or



Reference
Samet et al.
(2001)







Trenga et al.
(2001)

bronchial hyperresponsiveness to 0.1 ppm

SO2. The overall results interpreted as a

demonstration of protective effect of

antioxidants from O3, particularly in

"severe" asthmatics.
0.0
0.22
0.22



0.0
0.12-
0.24a





0 4h 21°C 25(M/F)
43 1 IE, 20' ex/10' rest 37% RH
431 VE = 40-46 L/min



0 2.17 h 22°Cor30°C 5 M, 4F
235-470a IE, 10' ex/ 10' rest 45-55% RH
VE = 36-39 L/min





Healthy NS
O3 responders and
nonresponders
18 to 40 yrs old


Healthy NS
24 to 32 yrs old






Glutathione peroxidase (GPx) activity and
eGPx protein level were significantly (p =
0.0001) depleted in ELF for at least 18 h
post-exp. In BAL both endpoints were
elevated (ns). No association between
cell injury, PF, or GPx activity.
FEVj decreased signif. (p < 0.5) by -8%
at22°Cand~6.5%at30°C. 19 h post-
exp decline of 2.3% still signif.
(p<0.05). SGaw signif. (p<0.05)
declined at 30°C but not at 22°C. The
BHR assessed 19 h post-exp. as PC50
sGaw methacholine signif. (p < 0.05)
higher at both temperatures.
Avissar et al.
(2000)b




Foster et al.
(2000)







-------
                                Table AX6-7 (cont'd). Influence of Ethnic, Environmental, and Other Factors
to
o
o
X
Oi
H
6
o
o
H
O
Ozone
Concentration
ppm Hg/ni3
0.0 0
0.40 780
Exposure Number
Duration Exposure and Gender Subject
and Activity Conditions of Subjects Characteristics
2 h 6 M, 9 F Healthy NS
IE, 20' ex/ 10' rest avg. age 3 1 yrs.
Observed Effect(s)
Corticosteroid pretreatment had no effects
on post O3 decline in PF, PMN response,
Reference
Nightingale
et al. (2000)
                           VE = mild to mod.
                                                                                         and sputum cell count under both the
                                                                                         placebo and treatment conditions.
                                                                                         Methacholine PC20 FEV{ was equally
                                                                                         decreased in both cond. 4 h after
                                                                                         exposure. No changes in exhaled NO and
                                                                                         CO
0.0 0
0.22 431



0.0 0
0.22 431
0.22 431





4h 21°C
IE 37% RH
20' ex/10' rest
VE = 40-46 L/min

4h 21°C
IE, 20' ex/10' rest 37% RH
VE = 25 L/min/m2
BSA




90 M 56 never smokers
34 current
smokers
18 to 40 yrs. old

10 M, 2 F NS, O3 nonresp.,
avg. age 25 yrs.;
10 M, 3 F NS, O3 resp.,
avg. age 25 yrs;
1 1 M, 2 F smokers avg.
age 28 yrs


Smokers are less responsive to O3 as
assessed by spirometric and
plethysmographic variables. Neither age,
gender, nor methacholine responsiveness
were predictive of O3 response.
Neither O3 responsiveness nor smoking
has altered the magnitude and the time
course of O3-induced airway
inflammation. Inflammation involved all
types of cells accessible by B AL. The
recovery profile of these cells overtime
was very similar for all groups showing
highest values 18 hpostexposure.
Frampton et al.
(1997a,c)b



Torres et al.
(1997)b
Frampton et al.
(1997a,c)b




        aRamp exposure from 0.12 ppm to 0.24 ppm and back to 0.12 ppm at the end of exposure.
        bRelated studies, sharing of some subjects .
o
HH
H
W

-------
 1      the largest increase in response to O3 in all groups, with nonsmokers showing greater relative
 2      increases than smokers.  Of the two cytokines, IL-6 and IL-8, IL-6 was substantially and
 3      significantly (p < 0.0002) elevated immediately post exposure but returned back to control 18 h
 4      later in all groups; but only nonsmokers' effects were significantly higher (p < 0.024). IL-8
 5      showed a similar pattern of response but the increase in all groups, though still significant
 6      (p < 0.0001), was not as high as for IL-6. Between group differences were not significant. This
 7      inflammatory response involved all types of cells present in BAL fluid and the recovery profile
 8      of these cells over time was very similar for all groups.  In contrast to BAL, NL did not prove to
 9      be a reliable marker of airway inflammation.  The lack of association between lung function
10      changes (spirometry) and airway inflammation for all three groups confirms similar observations
11      reported from other laboratories. This divergence of mechanisms is further enhanced by an
12      observation that a substantially different spirometric response between O3 responders and
13      nonresponders, the airway inflammatory response of the two groups was very similar, both in
14      terms of magnitude and pattern (Torres et al., 1997).
15          The influence of ambient temperature on pulmonary effects induced by O3 exposure in
16      humans has been studied infrequently under controlled laboratory conditions. Several
17      experimental human studies published more than 20 years ago reported additive effects of heat
18      and O3 exposure (see U.S. Environmental Protection Agency, 1986, 1996). In the study of
19      Foster et al. (2000) 9 young (mean age 27 years) healthy subjects (4F/5M) were  exposed for
20      130 min (IE 10 min @ 36 to 39 1/min) to filtered air and to ramp profile O3 at 22° and 30 °C,
21      45-55% RH. The order of exposures was randomized.  The O3 exposure started  at 0.12 ppm,
22      reached the peak of 0.24 ppm mid-way through and subsequently declined to 0.12 ppm at the
23      end of exposure.  Ozone inhalation decreased VT and increased fB as compared to baseline at
24      both temperatures. At the end of exposure FEVj decreased significantly (p < 0.5) by -8% at
25      22°C and -6.5% at 30 °C. One  day (19 h) later, the decline of 2.3% from baseline was still
26      significant (p < 0.05). FVC decrements were smaller and significant only at 22 °C immediately
27      postexposure. SGaw significantly (p < 0.05) declined at 30°C but not at 22 °C.  A day later,
28      sGaw was elevated above the baseline for all conditions. The nonspecific bronchial
29      responsiveness (NSBR) to methacloline assessed as PC50 sGaw was significantly (p <  0.05)
30      higher one day following O3 exposure at both temperatures but more so at 30 °C. Thus, these
31      findings indicate that elevated temperature has partially attenuated spirometric response but

        January 2005                            AX6-61      DRAFT-DO NOT QUOTE OR CITE

-------
 1      enhanced airway reactivity. Numerous studies have reported an increase in NSBR immediately
 2      after exposure to O3. Whether the late NSBR reported in this study is a persistent residual effect
 3      of an earlier increase in airway responsiveness, or is a true one day lag effect cannot be
 4      determined from this study. Whatever the origin, however, a delayed increase in airway
 5      responsiveness raises a question of potentially increased susceptibility of an individual to
 6      respiratory impairment, particularly if the suggested mechanism of disrupted epithelial
 7      membrane holds true.
 8
 9      AX6.5.6 Oxidant-Antioxidant Balance
10           Oxidant-antioxidant balance has been considered as one of the determinants of O3
11      responsiveness. Amateur cyclists who took antioxidant supplements (vitamins C, E,and
12      p-carotene) for three months showed no decrements in spirometric lung function when cycling
13      on days with high O3 levels.  In contrast, matched control  group of cyclists not pretreated with
14      vitamin supplements experienced an almost 2% decline in FVC and FEVj and > 5% reduction in
15      PEF during the same activity period.  Adjustment of data for confounders such as PM10 and NO2
16      did not change the findings. Apparently, substantially elevated levels of plasma antioxidants
17      may afford some protection against lung function impairment (Grievink et al.,1998, 1999).
18           Both laboratory animal and human studies have repeatedly demonstrated that antioxidant
19      compounds present the first line of defense against the oxidative stress. Thus, upregulation of
20      both enzymatic and nonenzymatic antioxidant systems is critical to airway epithelial protection
21      from exposure to oxidants such as O3  and NO2 (see Table AX6-7). As an extension  of an earlier
22      study focused on pulmonary function  changes (Frampton et al., 1997a), Avissar  et al. (2000)
23      hypothesized that concentration of glutathione peroxidase (GPx), one of the antioxidants in
24      epithelial lining fluid (ELF), is related to O3 and NO2 responsiveness. They exposed healthy
25      young nonsmokers (n=25), O3-responders, and non-responders to filtered air and twice to
26      0.22 ppm O3 for 4 h (IE, 20' ex 710' rest, @ VE 40 to 46 L/min). In the NO2 part  of the study,
27      subjects were exposed to air and twice to NO2 (0.6 and 1.5 ppm) for 3 h, with IE of 10 min of
28      each 30 min @ VE of 40 L/min. Ozone exposure elicited a typical pulmonary function response
29      with neutrophilic airway inflammation in both responders and non-responders.  The GPx activity
30      was significantly reduced (p = 0.0001) and eGPx protein significantly depleted (p = 0.0001) in
31      epithelial lining fluid (ELF) for at least 18 h postexposure. In contrast, both GPx and eGPx were

        January 2005                            AX6-62     DRAFT-DO NOT QUOTE OR CITE

-------
 1      slightly elevated in bronchoalveolar lavage fluid (BALF). However, neither of the two NO2
 2      exposures had a significant effect on pulmonary function, airway neutrophilia, epithelial
 3      permeability, GPx activity, or eGPx protein level in either ELF or BALF. The lack of a
 4      significant response to NO2 has been attributed to the weak oxidative properties of this gas.
 5      No association has been observed between cell injury, assessed by ELF albumin, or pulmonary
 6      function and GPx activity for O3 exposure. Thus, it is unclear what role antioxidants may have
 7      in modulation of O3-induced lung function and inflammatory responses. The authors found a
 8      negative association between lower baseline eGPx protein concentration in ELF and post-O3
 9      neutrophilia to be an important predictor of O3-induced inflammation; however, the causal
10      relationship has not been established.
11           The effects of dietary antioxidant supplementation on O3-induced pulmonary and
12      inflammatory response of young healthy individuals has been investigated by Samet et al.
13      (2001). Under controlled conditions, subjects received ascorbate restricted diet for three weeks.
14      After the first week of prescribed diet, subjects were randomly assigned into two groups,  and
15      exposed to air (2 h, IE every  15 min at 20 L/min/m2 BSA). Thereafter, one group received daily
16      placebo pills and the other a  daily supplement of ascorbate, a-tocopherol and a vegetable juice
17      for the next two weeks.  At the end of a two week period subjects were exposed to 0.4 ppm O3
18      under otherwise similar conditions as in sham exposures.  Serum concentration of antioxidants
19      determined prior to O3 exposure showed that subjects receiving supplements had substantially
20      higher concentrations of ascorbate, tocopherol and carotenoid in blood than the control group.
21      Plasma levels of glutathione  and uric acid (cellular antioxidants) remained essentially the same.
22      Ozone exposure reduced spirometric lung function in both groups; however, the average
23      decrements in the supplementation group were smaller for FVC (p = 0.046) and FEVj
24      (p = 0.055) when compared to the placebo group.  There was no significant correlation between
25      individual lung function changes and respective plasma levels of antioxidants. Individuals in
26      both groups experienced typical post O3 subjective symptoms of equal severity.  Similarly, the
27      inflammatory response as assessed by BALF showed no significant differences between the two
28      groups either in the recovery of cellular components or the types and concentrations of
29      inflammatory cytokines. Because of the complexity of protocol, the study was not designed as a
30      cross-over type.  However, it is unlikely that the fixed air-O3 sequence of exposures influenced
31      the findings in any substantial way. Although the study did not elucidate the protective

        January 2005                            AX6-63      DRAFT-DO NOT QUOTE OR CITE

-------
 1      mechanisms, it has demonstrated the value of dietary antioxidants in attenuating lung function
 2      effects of O3.  This observation may appear to contradict the findings of Avissar's and colleagues
 3      study (2000) discussed above; however, neither study found association between lung function
 4      changes and glutathione levels.  The lack of such association suggests that activation of
 5      antioxidant protective mechanisms is seemingly independent of mechanisms eliciting lung-
 6      function changes and that dietary antioxidants afford protection via a different pathway than
 7      tissue-dependent antioxidant enzymes. Moreover, the findings of this study have provided
 8      additional evidence that symptomatic, functional, inflammatory, and antioxidant responses are
 9      operating through substantially independent mechanisms.
10           Further evidence that the levels and activity of antioxidant enzymes in ELF may not be
11      predictive or indicative of O3-induced lung function or inflammatory effects has been provided
12      by a study of Blomberg et al.  (1999).  No association was found between the respiratory tract
13      lining fluid  redox potential level, an indicator of antioxidants balance, and either spirometric or
14      inflammatory changes induced by a moderate exposure of young individuals to O3 (0.2 ppm/2 h,
15      intermittent exercise at 20 L/min/m2 BSA). However, O3 exposure  caused a partial depletion of
16      antioxidants (uric acid, GSH,  EC-SOD) in nasal ELF and a compensatory increase in plasma uric
17      acid, affording at least some local protection (Mudway et al. 1999). More recently, Mudway
18      et al. (2001) investigated the effect of baseline antioxidant levels on response to a 2-h exposure
19      to 0.2 ppm O3 in 15 asthmatic and 15 healthy subjects. In the BALF of 15 healthy subjects,
20      significant O3-induced reductions in ascorbate and increases in glutathione disulphide and
21      EC-SOD were observed, whereas, levels were unaffected by O3 exposure in the asthmatics.
22      In both groups, BALF levels of uric acid and a-Tocopherol were unaffected by O3.
23           Trenga et al. (2001) studied the potential protective effects of dietary antioxidants (500 mg
24      vitamin C and 400 IU of vitamin E) on bronchial responsiveness of young to middle-aged
25      asthmatics who were prescreened for their hyperreactivity to SO2. Prior to the 1st exposure,
26      subjects took either two supplements  or two placebo pills at breakfast time for 4 weeks. They
27      continued taking respective pills for another week when the 2nd exposure took place.  The 45-min
28      exposures to air and 0.12 ppm O3 (15 min IE @ 3 x resting VE ) via mouthpiece were
29      randomized. Each exposure was followed by two 10-min challenges to 0.10 and 0.25 ppm SO2
30      with exercise to determine bronchial hyperresponsiveness.  Due to variability of baseline lung
31      function at different test sessions, and the way the data have been presented, it is difficult to

        January 2005                             AX6-64      DRAFT-DO NOT QUOTE OR CITE

-------
 1      interpret the results. The authors have reported no significant differences due to O3 between the
 2      placebo and supplemented cohort in either lung function or bronchial hyperresponsiveness to
 3      0.1 ppm  SO2. However, post hoc classification of subjects into "mild" and "severe" asthmatics
 4      (based on responses to SO2 during screening session) produced an unusual finding. In "severe"
 5      asthmatics, the challenge with 0.25 ppm SO2 completely reversed O3-induced whereas 0.1 ppm
 6      SO2-enhanced decrements  for PEF and FEF25.75. The overall results of the study were interpreted
 7      by the investigators as a demonstration of the protective effect of antioxidants from O3 exposure,
 8      particularly in "severe" asthmatics. It has been repeatedly demonstrated that pulmonary function
 9      response to O3 is reflex in origin, involving stimulation of bronchial C-fibers (see Section
10      AX6.2.5.1 for more information).  With  a relatively low O3 dose used in this study the reflex
11      response may be a dominant mechanism. Numerous animal and a few human studies used high
12      concentration of SO2 to "knock off the lung receptors.  It is plausible, therefore, that the
13      reversal of O3-induced spirometric decrements are due to a suppression of bronchial C-fibers
14      activation and a subsequent reversal of a reflex response. Thus, the observed recovery of
15      "severe" asthmatics following the second SO2 challenge reported by Trenga and colleagues
16      (2001) may not be related to antioxidant protection.
17
18      AX6.5.7  Genetic and Other Factors
19           It has been repeatedly postulated that genetic factors may play an important role in
20      individual responsiveness to ozone. Recent studies (Bergamaschi et al., 2001; Corradi et al,
21      2002; Romieu et al, 2004)  have indeed found that genetic polymorphisms of various enzymes,
22      namely NAD(P)H:quinone oxidoreductase (NQO1) and glutathione-S-trasferase Ml (GSTM1),
23      may play an important role in attenuating oxidative stress of airway epithelium. Bergamaschi
24      and colleagues (2001) studied young nonsmokers (15 F, 9 M; mean age 28.5 years) who cycled
25      for two hours on a cycling  circuit in a city park on days with the average ozone  concentration
26      ranging from 32 to 103 ppb.  There was  no control study group nor the intensity of bicycling has
27      been reported .  Since spirometry was done within 30 min post-ride, it is difficult to gage how
28      much of the statistically significant (p =  0.026) mean decrement of 160 ml in FEVj of 8/24
29      individuals with NQO1 wild type and GSTM1  null genotypes was due to ozone. Individuals
30      with other genotype combinations including GSTM1 null had a mean post-ride  decrement of
31      FEVj of only 40 mL. The  post-ride serum level of Clara cell protein (CC16), a  biomarker of

        January 2005                            AX6-65      DRAFT-DO NOT QUOTE OR CITE

-------
 1      airway permeability, has been elevated in both subgroups. Only a "susceptible" subgroup
 2      carrying NQO1 wild type in combination with GSTM1 null genotype, serum concentration of
 3      CC16 showed positive correlation with ambient concentration of ozone and negative correlation
 4      with FEVj changes. Despite some interesting observations, the study results should be
 5      interpreted cautiously.
 6           Pretreatment of healthy young subjects with inhaled corticosteroids (2 x 800 |ig/day
 7      budesonide, a maximal clinical dose) for 2 weeks prior to O3 exposure (0.4 ppm/2 h, alternating
 8      20 min exercise at SOW with 10 min rest) had no apparent effect on a typical lung function
 9      decline or inflammatory response to exposure. Because of the complexity of the protocol, the
10      study was not a cross-over design and no control air exposures were conducted. Both the
11      placebo and treatment conditions caused the same magnitude of changes.  Similarly, nonspecific
12      bronchial reactivity to methacholine (PC20 FEVj) was increased about the same 4 h after
13      exposure. Neither absolute nor relative sputum cell counts were affected by budesonide
14      treatment and O3 induced a typical neutrophilic response in both groups. Upregulation of pro-
15      inflammatory mediators measured in sputum was not different between the groups either. The
16      markers of inflammation and oxidative stress, exhaled NO and CO, as well as the reactive
17      product nitrite measured in exhaled breath condensate, respectively, were not significantly
18      influenced by budesonide. However, considering all these findings as a whole, budesonide
19      seemed to have a moderating, although not statistically significant, effect on O3-induced
20      response (Nightingale et al., 2000).  Budesonide is an antiinflammatory drug that in laboratory
21      animal studies partially suppressed neutrophilic inflammation caused by O3 (Stevens et al.,
22      1994). Because the dose of budesonide was at therapeutic maximal levels, the pharmacologic
23      action of this drug and the site of action of O3 do not apparently coincide.
24
25
26      AX6.6     REPEATED EXPOSURES TO OZONE
27           Repeated daily exposure to O3 in the laboratory for 4 or 5  days leads to attenuated changes
28      in pulmonary function responses and symptoms (Hackney et al., 1977a; U.S. Environmental
29      Protection Agency, 1986, 1996). A summary of studies investigating FEVj responses to
30      repeated daily exposure for up to 5 days is given in Table AX6-8. The FEVj responses to
31      repeated O3 exposure typically have shown an increased response on the second exposure day

        January 2005                           AX6-66      DRAFT-DO NOT QUOTE OR CITE

-------
fa
to
O
o










X
ON
ON



O
l>
H
6
o
0
H
O
Table AX6-8. Changes in Forced Expiratory Volume in One Second ,
Ozone
Concentration1"
ppm
0.12
0.20
0.20
0.20
0.20
0.25

0.35
0.35
0.35

0.35
0.40
0.40
0.4
0.4
0.42
0.45
0.45
0.47
0.5
0.5
ug/m3
235
392
392
392
392
490

686
686
686

686
784
784
784
784
823
882
882
921
980
980
Exposure Duration
and Activity0
6.6 h, IE (40)
2 h, IE (30)
2 h, IE (18 and 30)
2 h, IE (18 and 30)
1 h, CE (60)
1 h, CE (63)

2 h, IE (30)
1 h, CE (60)
1 h, CE (60)

1 h, CE (60)
3 h, IE (4-5 x resting)
3 h, IE (4-5 x resting)
2 h, IE (65)
3 h, IE (32)
2 h, IE (30)
2 h, IE (27)
2 h, IE (27)
2 h, IE (3 x resting)
2 h, IE (30)
2. 5 h, IE (2 x resting)
After Repeated Daily Exposure to Ozone"
Number and Gender Percent Change in FEV; on Consecutive
of Subjects Exposure Days References'1
17M
10M
8M, 13 F
9
15M
4M,2F
5M,2F
10M
8M
10M
10M
15M
13M5
HFf
8M
8M, 2Fh
24 M
1M, 5F
10M,6F
8M, 2F8
8M
6
First
-12.79
+1.4
-3.0
-8.7
-5.02
-20.2
-18.8
-5.3
-31.0
-16.1
-14.4
-15.9
-9.2
-8.8
-18.0
-34.7
-21.1
-13.3
-5.8
-11.4
-8.7
-2.7
Second
-8.73
+2.7
-4.5
-10.1
-7.8
-34.8
—
-5.0
-41.0
-30.4
—
-24.6
-10.8
-12.9
-29.9
-31.1
-26.4
—
-5.6
-22.9
-16.5
-4.9
Third
-2.54
-1.6
-1.1
-3.2
—
—
-22.3
-2.2
-33.0
—
-20.6
-5.3
-4.1
-21.1
-18.5
-18.0
-22.8
-1.9
-11.9
-3.5
-2.4
Fourth Fifth
-0.6 0.2 Folinsbeeetal. (1994)
— — Folinsbeeetal. (1980)
— — Glineretal. (1983)
— — Glineretal. (1983)
— — Brookes etal. (1989)
— — Folinsbee and Horvath (1986)
— —
— — Folinsbeeetal. (1980)
-25.0 — Foxcroft and Adams (1986)
— — Schonfeld etal. (1989)
— —
— — Brookes etal. (1989)
-0.7 -1.0 Kulle etal. (1982)
-3.0 -1.6 Kulle etal. (1982)
-7.0 -4.4 Folinsbeeetal. (1998)
-12.0 -6.2 Gong etal. (1997b)
-6.3 -2.3 Horvath etal. (1981)
— — Bedi etal. (1985)
— — Bedi etal. (1989)
-4.3 — Linn etal. (1982)
— — Folinsbeeetal. (1980)
-0.7 — Hackney etal. (1977a)
"See Appendix A for abbreviations and acronyms.
bT,isteH from lowest to highest O, concentration
O
HH
H
W
'Exposure duration and intensity of IE or CE were variable; VE (number in parentheses) given in liters per minute or as a multiple of resting ventilation.
*For a more complete discussion of these studies, see Table AX6-9 and U.S. Environmental Protection Agency (1986).
"Subjects were especially sensitive on prior exposure to 0.42 ppm O3 as evidenced by a decrease in FEV[ of more than 20%. These nine subjects are a subset of the total group of
21 individuals used in this study.
'Bronchial reactivity to  a methacholine challenge also was studied.
8Seven subjects completed entire experiment.
hSubjects had mild asthma.

-------
 1      (Day 2) compared to the initial (Day 1) exposure response. This is readily apparent in repeated
 2      exposures to a range of concentrations from 0.4 to 0.5 ppm O3 accompanied by moderate
 3      exercise (Folinsbee et al., 1980; Horvath et al., 1981; Linn et al., 1982), and at lower
 4      concentrations, 0.20 to 0.35 ppm, when accompanied by heavy exercise (Brookes et al., 1989;
 5      Folinsbee and Horvath, 1986; Foxcroft and Adams, 1986; Schonfeld et al., 1989).  Mechanisms
 6      for enhanced pulmonary function responses on Day 2 have not been established, although
 7      persistence of acute O3-induced damage for greater than 24 h may be important (Folinsbee et al.,
 8      1993). An enhanced Day 2 FEVj response was less obvious or absent in exposures at lower
 9      concentrations or those that caused relatively small group mean O3-induced decrements.
10      For example, Bedi et al. (1988) found no enhancement of the relatively small pulmonary
11      function responses in older subjects (median age, 65 years) exposed repeatedly to O3.  Three
12      reports (Bedi et al., 1985; Folinsbee and Horvath, 1986; Schonfeld et al., 1989) demonstrated
13      that enhanced pulmonary function responsiveness was present within 12 h, lasted for at least
14      24 h and possibly 48 h, but was absent after 72 h.
15           After 3 to 5 days of consecutive daily exposures to O3, FEVj responses are markedly
16      diminished or absent. One study (Horvath et al., 1981) suggested that the rapidity of this decline
17      in FEVj response was related to the magnitude of the subjects' initial responses to O3  or their
18      "sensitivity." A summary of studies examining the effects of repeated exposures to O3 on FEVj
19      and other pulmonary function, symptoms, and airway inflammation is given in Table AX6-9.
20      Studies examining persistence of the attenuation of pulmonary function responses following
21      4 days of repeated exposure (Horvath et al., 1981; Kulle et al., 1982; Linn et al., 1982) indicate
22      that attenuation is relatively short-lived, being partially reversed within 3 to 7 days and typically
23      abolished within 1 to 2 weeks. Repeated exposures separated by 1  week (for up to 6 weeks)
24      apparently do not induce attenuation of the pulmonary function response (Linn et al.,  1982).
25      Gong et al. (1997b) studied the effects of repeated exposure to 0.4 ppm O3 in a group  of mild
26      asthmatics and observed a similar pattern of responses as those seen previously in healthy
27      subjects. The attenuation of pulmonary responses reached after 5 days of consecutive O3
28      exposure was partially lost at 4 and 7 days post exposure.
29           In addition to the significant attenuation or absence of pulmonary function responses after
30      several consecutive daily O3 exposures, symptoms of cough and chest discomfort usually
31      associated with O3 exposure generally  are substantially reduced or absent (Folinsbee et al.,  1980,

        January 2005                            AX6-68      DRAFT-DO NOT QUOTE OR CITE

-------
                            Table AX6-9. Pulmonary Function Effects with Repeated Exposures to Ozone"
to
o
o
X
Oi

ON
VO
H

6
o


o
H

O
Ozone
Concentration11 Exposure
ppm ug/m3 Activity
0.25 490 2 h IE, (30 min
rest, 30 min
exercise), VE =
39 L/min


0.2 392 4 h IE
(4 x 30 min
exercise), VE =
14.8 L/min/m2
BSA
0.2 392 4 h IE
(4 x 30 min
exercise), VE =
25 L/min/m2
BSA

0.4 784 3h/day for 5
days
IE ( 1 5 min rest,
15 min exercise)
VE = 32 L/min
0.12 235 6.6 h
50 min
exercise/10 min
rest, 30 min
lunch
VE = 38.8L/min


0.4 784 2 h IE (15 min
rest, 15 min
exercise
VE ~ 60 L/min
Number and
Gender of Subject
Exposure Conditions Subjects Characteristics
21.4°C 5M, 3F Healthy, NS
43.9%RH
4 days consecutive FA
exposure; 4 days
consecutive O3
exposure
1 day FA, 1 day, O3; 15 M, 8 F Healthy, NS
4 days consecutive 21 to 35 years old
exposure to O3


20 °C 50% RH (1 day, 9 M, 6 F Healthy, NS
O3; 4 days consecutive 23 to 37 years old
exposure to O3



31 °C 8M, 2F Mild asthma
35% RH adult
5 consecutive days
plus follow up @ 4 or
7 days
18°C 17 M Healthy NS
40% RH
five consecutive
daily exposures




5 days consecutive O3 16 M Healthy NS
exposure


Observed Effect(s)
FVC and FEVj decrements were significantly
attenuated on Day 4 of O3 exposure compared to day 1
of O3 exposure. Significant small airway function
depression accompanied by significant reutrophilia in
BALE one day following the end of O3 exposure.

FEVj decrement and symptoms significantly reduced
on Day 4 of O3 exposure compared to Day 1 of O3
exposure. Airway inflammation of mucosa persisted on
Day 4 although some inflammatory markers in BALE
attenuated significantly.
Significant decrease in FVC, FEVj, SRaw, and
symptoms on Day 4 of O3 exposure compared to a
single day of O3 exposure. Number of PMNs,
fibronectin, and IL6 in BALE were significantly
decreased on Day 4 compared to a single day of O3
exposure.
FEVj decreased 35% on day 1 and only 6% on day 5.
Bronchial reactivity increased after day 1 and remained
elevated. Adaptation of asthmatics is similar to healthy
subjects but may be slower and less complete.

FEVj responses were maximal on first day of exposure
(- 13%), less on second day (-9%), absent thereafter.
Symptoms only the first 2 days. Methacholine airway
responsiveness was at least doubled on all exposure
days, but was highest on the second day of O3. Airway
responsiveness was still higher than air control after 5
days of O3 exposure. Trend to lessened response, but it
was not achieved after 5 days.
O3-exposure FEVj decrement was greater on day 2,
29.9%, than day 1, 18.0%, then decreased on day 3,
21 .1%, day 4, 7% and day 5, 4.4%

Reference
Frank et al.
(2001)




Jorres et al.
(2000)



Christian et al.
(1998)




Gong et al.
(1997b)



Folinsbee et al.
(1994)






Folinsbee et al.
(1998)
Devlin et al.
(1997)
o
HH
H
W

-------

to
o
o
                                  Table AX6-9 (cont'd).  Pulmonary Function Effects with Repeated Exposures to Ozonea
X
Ozone
Concentration11
ppm ug/m3 and Activity
0.45 882 2 h
IE
Exposure Conditions
23.3 °C
63% RH
Number
and Gender
of Subjects
10 M, 6F
Subject
Characteristics
Healthy NS
60 to 89 years old
Observed Effect(s)
Overall increase in symptoms, but no single
symptom increased significantly. FVC
Reference
Bedi et al.
(1989)
                              (3 x 20 min exercise)
          0.20/0.20   392/392   1 h
          0.35/0.20   686/392   CEat60L/min
          0.35/0.35   686/686
            0.35
                      686
            0.45
                      882
60 min CE
VB = 60 L/min
2hIE
(3 x 20 min exercise)
                     Exposed for 3 consecutive
                     days, not exposed for 2 days,
                     then exposed to 0.45 ppm
                     again for 1 day
                     21 to 25 °C
                     40 to 60% RH
                     (three 2-day sets
                     of exposures)
                                                   21 to 25 °C
                                                   40 to 60% RH
                                                   (two exposures for each
                                                   subject separated by 24, 48,
                                                   72, or 120 h)
                                                   23.3 °C
                                                   62.5% RH
                                                   (three exposures with
                                                   a minimum 1-week interval)
                                                  15M
40 M
(4 groups
of 10)
8M, 8F
              median 65 years old;
              mean FVC = 3.99 L;
              meanFEV^S.Ol L;
              FEV/FVC range =
              61 to 85%
              Healthy aerobically
              trained NS,
              FVC = 4.24 to 6.98 L
NS; nonallergic,
non-Los Angeles
residents for > 6 mo;
=25 years old
Healthy NS, 61 years
old for M and
65 years old for F
(FVC = 4.97 L for
M and 3.11 LforF)
decreased 111 mL and 104 mL on Days 1 and 2,
respectively. FEVl fell by 171 and 164 mL, and
FEV3 fell by 185 and 172 mL. No significant
changes on Days 3 and 4 or with FA.  FEVl
changes were -5.8, -5.6, -1.9, and -1.7% on
the four O3 days.

Consecutive days of exposure to 0.20 ppm        Brookes et al.
produced similar FEVj responses on each day     (1989)
(-5.02, -7.80); 0.35/0.20 ppm pair caused
increased response to 0.20 ppm on second day
(-8.74); 0.35/0.35 ppm caused much increased
response on Day 2 (-15.9, -24.6). Symptoms
were worse  on the second exposure to 0.35 ppm,
but not with second exposure to 0.20 ppm.

No differences between responses to exposures    Schonfeld et al.
separated by 72 or 120 h. Enhanced YEVl        (1989)
response at24 h (-16.1% vs. -30.4%). Possible
enhanced response at 48 h (-14.4% vs.
-20.6%). Similar trends observed for breathing
pattern and SRaw.

Spirometric changes were not reproducible from   Bedi et al.
time to time after O3 exposure R < 0.50).          (1988)
Repeat exposures to air yielded consistent
responses.
            0.18       353     2h
                              IE (heavy)
                              VE = 60 to "
                              (35 L/min/m2 BSA)
                     31°C                        59 adult
                     35% RH                     Los Angeles
                     (screen exposures             residents
                     in spring 1986;                12
                     second exposures in            responsive
                     summer/fall 1986 and winter    13
                     1987 and spring 1987 for       nonresponsi
                     responders and                ve
                     nonresponders only)
              Responders:
               19 to 40 years old
               6 atopic,
               2 asthmatic,
               4 normal

              Nonresponders:
              18 to 39 years old,
               13 normal
                     Responders had AFEVj = -12.4% after initial
                     screening; nonresponders had no change.
                     Responders had nonsignificant response in late
                     summer or early winter, but were responsive
                     again in early spring (spring 1986, -385 mL;
                     Autumn 1986, -17 mL; winter 1987, +16 mL;
                     spring 1987, -347 mL). Nonresponders did not
                     change with season. Suggests that responders
                     responses may vary with ambient exposure, but
                     nonresponders generally remain nonresponsive.
                                            Linn et al.
                                            (1988)
                                            (also see
                                            Hackney et al..
                                            1989)

-------
to
o
o
                                 Table AX6-9 (cont'd).  Pulmonary Function Effects with Repeated Exposures to Ozonea
Ozone
Concentration11
ppm ug/m3 and Activity
Number
and Gender Subject
Exposure Conditions of Subjects Characteristics
Observed Effect(s) Reference
            0.45
         (+ 0.30
         PAN)
                      882
2h
IE (20 min rest,
20 min exercise)
VB = 27 L/min
22 °C
60% RH
5 days consecutive
exposure to PAN + O3
3 M, 5 F      Healthy NS,
             Mean age = 24 years
FEVj decreased ~ 19% with O3 alone, ~ 15% on
Day 1 of O3 + PAN, =5% on Day 5 of O3 +
PAN, =7% 3 days after 5 days of O3 + PAN,
= 15% after 5 days of O3 + PAN. Similar to
other repeated O3 exposure studies, O3 responses
peaked after 2 days, were depressed 3 days later,
and responses returned 7 days later.  PAN
probably had no effect on repeated to O3
exposure responses.
Drechsler-Parks
etal. (1987b)
(also see
Table AX6-14)
X
H
6
o
o
H
O
0.35 686 =lhCE 22 to 25 °C 8M
(see paper for details) 35 to 50% RH
(1 day FA; 1 day O3; 4 days
consecutive exposure to O3)
Aerobically trained
healthy NS (some
were known O3
sensitive),
22.4 ±2.2 years old
Largest FEV; decrease on second of 4 days O3
exposure (-40% mean decrease). Trend for
attenuation of pulmonary function response not
complete in 4 days. VOj^ decreased with
single acute O3 exposure (-6%) but was not
significant after 4 days of O3 exposure (-4%).
Performance time was less after acute O3 (21 1 s)
exposure than after FA (253 s).
Foxcroft and
Adams (1986)
         "See Appendix A for abbreviations and acronyms.
         bListed from lowest to highest O, concentration.
o
HH
H
W

-------
 1      1994; Foxcroft and Adams, 1986; Linn et al., 1982). Airway responsiveness to methacholine is
 2      increased with an initial O3 exposure (Holtzman et al., 1979; Folinsbee et al.,  1988), may be
 3      further increased with subsequent exposures (Folinsbee et al.,1994), and shows a tendency for
 4      the increased response to diminish with repeated exposure (Dimeo et al., 1981; Kulle et al.,
 5      1982). The initially enhanced and then lessened response may be related to changes that occur
 6      during the repair of pulmonary epithelia damaged as a consequence of O3 exposure.
 7      Inflammatory responses (Koren et al., 1989a), epithelial damage, and changes in permeability
 8      (Kehrl et al.,  1987) might explain a portion of these responses. By blocking pulmonary function
 9      responses and symptoms with indomethacin pretreatment, Schonfeld et al. (1989) demonstrated
10      that in the absence of an initial response, pulmonary function and symptoms effects were not
11      enhanced on Day 2 by repeated exposure to 0.35 ppm O3. These results suggest that airway
12      inflammation and the release of cyclooxygenase products of arachidonic acid play a role in the
13      enhanced pulmonary function responses and symptoms observed upon reexposure to O3 within
14      48 h.
15          Response to laboratory O3 exposure as a function of the season of the year in the South
16      Coast Air Basin of Los Angeles, CA, has been examined in several studies (Avol et al., 1988;
17      Hackney et al., 1989; Linn et al., 1988).  Their primary purpose was to determine whether O3
18      responsive subjects would remain responsive after regular ambient exposure during the "smog
19      season". The subjects were exposed to 0.18 ppm O3 for 2 h with heavy IE on four occasions,
20      spring, fall, winter, and the following spring. The marked difference in FEVj response between
21      responsive and nonresponsive subjects seen initially (-12.4% versus +1%) no longer was present
22      after the summer smog season (fall test) or 3 to 5 months later (winter test). However, when the
23      subjects were exposed to O3 during the following spring, the responsive subjects again had
24      significantly larger changes in FEVl5 suggesting a seasonal variation in response.
25          Brookes et al. (1989) and Gliner et al. (1983) tested whether initial exposure to one O3
26      concentration could alter response to subsequent exposure to a different O3 concentration.
27      Gliner et al. (1983) showed that FEVj response to 0.40 ppm O3 was not influenced by previously
28      being exposed to 0.20 ppm O3 for 2 h on 3 consecutive days. Brookes et al. (1989) found
29      enhanced FEVj and symptoms upon exposure to 0.20 ppm after previous exposure to 0.35 ppm
30      O3. These  observations suggest that, although preexposure to low concentrations of O3 may not
        January 2005                            AX6-72      DRAFT-DO NOT QUOTE OR CITE

-------
 1      influence responses to higher concentrations, preexposure to a high concentration of O3 can
 2      significantly increase responses to a lower concentration on the following day.
 3           Foxcroft and Adams (1986) demonstrated that decrements in exercise performance seen
 4      after 1 h of exposure to 0.35 ppm O3 with heavy CE were significantly less after 4 consecutive
 5      days exposure than they were after a single acute exposure. Further, exercise performance,
 6      VO2max, VEmax and HR^ were not significantly different after 4 days of O3 exposure compared to
 7      those observed in a FA exposure.  Despite the change in exercise performance, Foxcroft and
 8      Adams (1986) did not observe a significant attenuation of FEVj response, although symptoms
 9      were significantly reduced. However, these investigators selected known O3-sensitive subjects
10      whose FEVj decrements exceeded 30% on the first 3 days of exposure. The large magnitude of
11      these responses, the trend for the responses to decrease on the third and fourth day, the decreased
12      symptoms, and the observations by Horvath et al. (1981) that O3-sensitive subjects adapt slowly,
13      suggest that attenuation of response would have occurred if the exposure series had been
14      continued for another  1 or 2 days.  These observations support the contention advanced by
15      Horvath et al. (1981) that the progression of attenuation of response is a function of initial "O3
16      sensitivity."
17           Drechsler-Parks  et al. (1987b) examined the response to repeated exposures to 0.45 ppm O3
18      plus 0.30 ppm peroxyacetyl nitrate (PAN) in 8 healthy subjects and found similar FEVj
19      responses to exposures to O3 (-19%) and to O3 plus PAN (-15%).  Thus, PAN did not increase
20      responses to O3.  Further, repeated exposure to the PAN plus O3 mixture resulted in  similar
21      changes to those seen with repeated O3 exposure alone.  The FEVj responses fell to less than
22      - 5% after the fifth day, with the attenuation of response persisting 3 days after the repeated
23      exposures, but being absent after 7 days.  These observations suggest that PAN does not
24      influence the attenuation of response to repeated O3 exposure.  If the PAN responses are
25      considered negligible, this study confirms  the observation that the attenuation of O3  responses
26      with chamber exposures lasts no longer than 1 week.  [More discussion on the interaction ofO3
27      with other pollutants can be found in Section AX6.11.]
28           Folinsbee et al. (1993) exposed a group of 16 healthy males to 0.4 ppm O3 for 2 h/day on
29      5 consecutive days. Subjects performed heavy IE (VE = 60 to 70 L/min). Decrements in FEVj
30      averaged 18.0, 29.9, 21.1, 7.0, and 4.4% on the 5 exposure days. However, baseline preexposure
31      FEVj decreased from the first day's preexposure measurement and was depressed by an average

        January 2005                            AX6-73      DRAFT-DO NOT QUOTE OR CITE

-------
 1      of about 5% by the third day. This study illustrates that, with high-concentration and heavy-
 2      exercise exposures, pulmonary function responses may not be completely recovered within 24 h.
 3      During this study, BALF also was obtained immediately after the Day 5 exposure, with results
 4      reported by Devlin et al. (1997).  These authors found that some inflammation and cellular
 5      responses associated with acute O3 exposure were also attenuated after 5 consecutive days of O3
 6      exposure (compared to historical data for responses after a single-day exposure), although
 7      indicators of epithelial cell damage—not seen immediately after acute exposure—were present
 8      in BALF after the fifth day of exposure. When reexposed again 2 weeks later, changes in BALF
 9      indicated that epithelial cells appeared fully repaired (Devlin et al., 1997).
10           Frank et al. (2001) exposed 8 healthy young adults to 0.25 ppm O3 for 2 h with moderate
11      IE (exercise VE = 40 L/min) on 4 consecutive days. In addition to standard pulmonary function
12      measures, isovolumetric FEF25.75, Vmax50 and Vmax75 were grouped into a single value representing
13      small airway function (SAWgrp).  Exercise ventilatory pattern was also monitored each day,
14      while peripheral airway resistance was measured by bronchoscopy followed by  lavage on Day 5.
15      The authors observed two patterns of functional response in their subjects— attenuation and
16      persistent.  Values of FVC and FEVj showed significant attenuation by Day 4 compared to Day
17      1 values. However,  SAWgrp and rapid shallow breathing during exercise persisted on Day 4
18      compared to Day 1, and were accompanied by significant neutrophilia in BALF 1 day following
19      the end of O3 exposure. Frank et al. (2001) suggested that both types of functional response (i.e.,
20      attenuation and persistence) are linked causally to inflammation. They contend that the
21      attenuation component is attributable at least  in part to a reduction in local tissue dose during
22      repetitive exposure that is likely to result from the biochemical, mechanical, and morphological
23      changes set in motion by  inflammation. They speculated that the persistent component
24      represents the inefficiencies incurred through inflammation.  Whether the persistent small airway
25      dysfunction is a forerunner of more permanent change in the event that oxidant  stress is extended
26      over lengthy periods of time is unknown.
27           Early repeated multihour (6 to 8 h) exposures focused on exposures to low concentrations
28      of O3 between 0.08 and 0.12 ppm (Folinsbee  et al., 1994; Horvath et al., 1991; Linn  et al., 1994).
29      Horvath et al. (1991) exposed subjects for 2 consecutive days to 0.08 ppm using the 6.6-h
30      prolonged exposure protocol (see Table AX6-2). They observed small pre- to postexposure
31      changes in FEVj (-2.5%) on the first day, but no change on the second day. Linn et al. (1994)

        January 2005                            AX6-74      DRAFT-DO NOT QUOTE OR CITE

-------
 1      observed a 1.7% decrease in FEVj in healthy subjects after 6.6 h exposure to 0.12 ppm O3.
 2      A second consecutive day exposure to O3 yielded even smaller (< 1%) responses. In a group of
 3      asthmatics exposed under similar conditions (Linn et al., 1994), the FEVj response on the first
 4      day was -8.6% which was reduced to -6.7% on day 2, both significantly greater than those
 5      observed for the nonasthmatics group. The observations of Horvath et al. (1991) and Linn et al.
 6      (1994) elicited a somewhat different pattern of response (no enhancement of response after the
 7      first exposure) than that seen at higher concentrations in 2 h exposures with heavy exercise
 8      (Tables AX6-8 and AX6-9). However, the subjects studied by Horvath et al. (1991) were
 9      exposed only to 0.08 ppm O3 and were somewhat older (30 to 43 yrs) than the subjects studied
10      by Folinsbee et al. (1994), mean  age of 25 yrs, while the nonasthmatic subjects studied by Linn
11      et al. (1994) were also older (mean = 32 yrs), had lower exercise VE (-20%) and were residents
12      of Los Angeles who often encountered ambient levels of O3 at or above 0.12 ppm.
13          Folinsbee et al. (1994) exposed 17 subjects to 0.12 ppm O3 for 6.6 h, with 50 min of
14      moderately heavy exercise ( VE = 39 L/min) each hour, on 5 consecutive days.  Compared  with
15      FA, the percentage changes in FEVi over the five days were -12.8%, -8.7%, -2.5%, -0.06%,
16      and +0.18%. A parallel attenuation of symptoms was observed, but the effect of O3 in enhancing
17      airway responsiveness (measured by increase in SRaw upon methacholine challenge) over
18      5 days was not attenuated (3.67,  4.55, 3.99, 3.24, and 3.74, compared to 2.22 in FA control).
19      Nasal lavage revealed no increases in neutrophils except on the first O3 exposure day.
20          Christian et al. (1998) exposed 15 adults (6 females and 9 males; mean age = 29.1 yrs) to
21      4 consecutive days at 0.20 ppm O3 for 4 h, with 30 mm of IE (exercise VE = 25 L/min/m2)  each
22      hour. Measures of FEVb FVC, and symptoms were all significantly reduced on Day 1, further
23      decreased on Day 2, and then attenuated to near FA control values on Day 4. The pattern of
24      SRaw response was  similar, being greatest on Day 2 and no different from FA control on Day 4.
25      BAL was done on Day 5 and showed that neutrophil recruitment to the respiratory tract was
26      attenuated with repeated short-term exposures, compared to Day 1 control O3 exposure, while
27      airway epithelial injury appeared to continue as reflected by no attenuation of IL-6,  IL-8, total
28      protein, and LDH. The authors concluded that such injury might lead to airway remodeling,
29      which has been observed in several animal studies (Brummer et al., 1977; Schwartz et al.,  1976;
30      Tepper et al., 1989; Van Bree et  al., 1989). In a similar study to that of Christian et al. (1998),
31      Torres et al. (2000) exposed 23 adults (8 females and  15 males; mean age = 27.9 yrs) on

        January 2005                            AX6-75      DRAFT-DO NOT QUOTE OR CITE

-------
 1      4 consecutive days to 0.20 ppm O3 for 4 h, with 30 min of IE (exercise VE = 26 L/min) each
 2      hour. The authors observed that FEVj was significantly reduced and symptoms were
 3      significantly increased on Day 1.  On Day 2, FEVj was further decreased, while symptoms
 4      remained unchanged. By Day 4, both FEVj and symptoms were attenuated to near FA, control
 5      values.  Twenty hours after the Day 4 exposure, BAL and bronchial mucosal biopsies were
 6      performed.  These authors found via bronchial mucosal biopsies that inflammation of the
 7      bronchial mucosa persisted after repeated O3 exposure, despite attenuation of some inflammatory
 8      markers in BALF and attenuation of lung function responses and symptoms. Further, Torres
 9      et al. (2000) observed persistent although small decrease in baseline FEVj measured before
10      exposure, thereby suggesting that there are different time scales of the functional responses to
11      O3, which may  reflect different mechanisms.  The levels of protein remaining elevated after
12      repeated exposures confirms the findings of others (Christian et al., 1998; Devlin et al., 1997),
13      and suggests that there is ongoing cellular damage irrespective of the attenuation of cellular
14      inflammatory responses with the airways.  [Further discussion on the inflammatory responses to
15      O3 can be found in Section AX6.9. ]
16          Based on  studies cited here and in the previous O3 criteria documents (U.S. Environmental
17      Protection Agency, 1986, 1996), several conclusions can be drawn about repeated 1- to 2-h O3
18      exposures. Repeated exposures to O3 can cause an enhanced (i.e., greater) lung function
19      response on the second day of exposure. This enhancement appears to be dependent on the
20      interval between the  exposures (24 h is associated with the greatest increase) and is absent with
21      intervals > 3 days. As shown in Figure AX6-8, an enhanced response also appears to depend on
22      O3 concentration and to some extent on the magnitude of the initial response.  Small responses to
23      the first O3 exposure  are less likely to result in an enhanced response on the second day of O3
24      exposure. Repeated  daily exposure also results in attenuation of pulmonary function responses,
25      typically after 3 to 5  days of exposure. This attenuated response persists for less than 1 week or
26      as long as 2 weeks. In temporal conjunction with the pulmonary function changes, symptoms
27      induced by O3,  such as cough and chest discomfort, also are attenuated with repeated exposure.
28      Ozone-induced changes in airway responsiveness attenuate more slowly than pulmonary
29      function responses and symptoms. Attenuation of the changes in airway responsiveness appear
30      to persist longer than changes in pulmonary function, although this has been studied only on a
31      limited basis. In longer-duration (6.6 h), lower-concentration studies that do not cause an

        January 2005                            AX6-76      DRAFT-DO NOT QUOTE OR CITE

-------
           c
           o
           o
           •s
          HI
100-

 95-

 90-

 85-

 80-

 75-

 70-

 65-
                   T        I        I         I        I        I
                  0.0      0.5      1.0       1.5      2.0      2.5
                                               Time (days)
                                                       i
                                                      3.0
                    I
                   3.5
4.0
      Figure AX6-8.  Regression curves were fitted to day-by-day postexposure FEVt values
                     obtained after repeated daily acute exposures to O3 for 2 to 3 h with
                     intermittent exercise at a VE of 24 to 43 L/min (adaptation studies).
                     Symbols represent the results from individual studies conducted at 0.2 ppm
                     for 2 h (+), 0.35 ppm for 2 h (•), 0.4 ppm for 2 h (+), 0.5 ppm for 2 h (#),
                     and 0.54 ppm for 3 h (A). Also shown for comparison are the FEVt values
                     obtained after exposure to 0.12 ppm O3 for 10 h (•).

      Source: Modified from Hazucha (1993).
1     enhanced second-day response, the attenuation of response to O3 appears to proceed more
2     rapidly. Inflammatory markers from BALF on the day following both 2 h (Devlin et al., 1997)

3     and 4 h (Christian et al., 1998; Torres et al., 2000) repeated O3 exposure for 4 days indicate that
4     there is ongoing cellular damage irrespective of the attenuation of some cellular inflammatory

5     responses of the airways, lung function responses and symptoms.
      January 2005
                              AX6-77
DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.7  EFFECTS ON EXERCISE PERFORMANCE
 2     AX6.7.1  Introduction
 3          In an early epidemiologic study examining race performances in Los Angeles area high
 4     school cross-country runners, Wayne et al. (1967) observed that endurance exercise performance
 5     was depressed by inhalation of ambient oxidant air pollutants. The authors concluded that the
 6     detrimental effects of oxidant air pollutants on race performance might have been related to the
 7     associated discomfort in breathing, thus limiting the runners' motivation to perform at high
 8     levels, although physiologic effects limiting O2 availability could not be ruled out.
 9     Subsequently, the effects of acute O3 inhalation on endurance exercise performance have been
10     examined in numerous controlled laboratory studies. These studies were discussed in the
11     previous O3 criteria document (U.S. Environmental Protection Agency, 1996) in two categories:
12     (1) those that examined the effects of acute O3 inhalation on maximal oxygen uptake (VO2max)
13     and (2) those that examined the effects of acute O3 inhalation on the ability to complete
14     strenuous continuous exercise protocols of up to 1 h in duration. In this section, major
15     observations in these studies are briefly reviewed with emphasis on reexamining the primary
16     mechanisms causing decrements in VO2max and endurance exercise performance consequent to
17     O3 inhalation. A summary of major studies of O3 inhalation effects on endurance exercise
18     performance, together with observed pulmonary function and symptoms of breathing discomfort
19     responses, is given in Table AX6-10.
20
21     AX6.7.2  Effect on Maximal Oxygen Uptake
22          Three early studies (Folinsbee et al.,  1977; Horvath et al., 1979; Savin and Adams, 1979)
23     examining the effects of acute O3 exposures on VO2max were reviewed in an earlier O3 criteria
24     document (U.S. Environmental Protection Agency, 1986).  Briefly, Folinsbee et al. (1977)
25     observed that VO2max was significantly decreased (10.5%) following a 2-h exposure to 0.75 ppm
26     O3 with light (50 Watts) IE. Reduction in VO2max was  accompanied by a decrease in maximal
27     ventilation, maximal heart rate, and a large decrease in maximal tidal volume. In addition, the
28     2-h IE O3 exposure resulted in a 22.3% decrease in FEVj and significant symptoms of cough and
29     chest discomfort. In contrast, Horvath et al. (1979) did not observe a change in VO2max or other
30     maximal cardiopulmonary endpoints in subjects  exposed for 2 h at rest to either 0.50 or


       January 2005                            AX6-78      DRAFT-DO NOT QUOTE OR CITE

-------
Table AX6-10. Ozone Effects on Exercise Performance"
3
C
to
o
o















^>
X
ON
i
VO




O
!>
!-rj
H
6
o
2!
-*— 1
o
H
O
o
NW'
H
W
0
o
HH
H
W


Ozone
Concentration b

ppm
0.06-0.07
0.12-0.13



0.18


0.35


0.12
0.20

0.12
0.18
0.24

0.21

0.20
0.35



0.25
0.50
0.75

0.15
0.30


0.75




Hg/m3
120-140
245-260



353


686


235
392

235
353
470

412

392
686



490
980
1,470

294588



1,470



xauii; r^^vu

Exposure Duration Exposure
and Activity Conditions
CE 23 to 24.5 °C
(VE = 30 to 120 L/min) 50 to 53% RH
16 to 28 min progressive
maximum exercise
protocol
1 h CE or competitive NA
simulation at mean
VE = 94 L/min
50 min CE 22 to 25 °C
VE = 60 L/min 35 to 50% RH

IhCE 31 °C
VE = 89 L/min

1 h competitive 23 to 26 °C
simulation exposures at 45 to 60% RH
mean
VE = 87 L/min
1 h CE at 75% V02max 19 to 21 °C
60 to 70% RH
1 h CE or competitive 23 to 26 °C
simulation at mean 45 to 60% RH
VE = 77.5 L/min


2 h rest NA



—30 min, progressively 23 °C
incremented exercise to 50% RH
voluntary exhaustion

2 h IE NA
(4x 15 min light [50 W]
bicycle ergometry)

'-AV. V^i/LFllC i^llCVia LFll iliACl
Number and
Gender of Subject
Subjects Characteristics
12 M, 12 F Athletic




# not given; Well-trained
all males distance runners

8 M Trained
nonathletes

15M, 2F Highly trained
competitive
cyclists
10 M Highly trained
competitive
cyclists

6 M, 1 F Well-trained
cyclists
10 M Well-trained
distance runners



8M, 5F



9 M Healthy, NS
21 to 44 years old


13 M 4 light S,
9NS


viac i ci iiri UKIIIVI;


Observed Effect(s)
Reduced maximum performance time and increased symptoms
of breathing discomfort during O3 exposure.



Maximal treadmill run time reduced from 71.7 min in FA to
66.2 min during O3 exposure with no decease in arterial O2
saturation.
VT decreased, fB increased with 50-min O3 exposures; decrease
in FVC, FEVj, FEF25.75, performance time, VO2max, V^^, and
HR,,,,,, from FA to 0.35-ppm O3 exposure.
Decrease in VEmax, VO2max, VTmax, workload, ride time, FVC,
and FEVj with 0.20 ppm O3 exposure, but not significant with
0. 12-ppm O3 exposure, as compared to FA exposure.
Decrease in exercise time of 7.7 min and 10. 1 min for subjects
unable to complete the competitive simulation at 0.18 and 0.24
ppm O3, respectively; decrease in FVC and FEVj for 0.18-and
0.24-ppm O3 exposure compared with FA exposure.
Decrease in FVC, FEVj, FEF25.75, and MVV with 0.21 ppm O3
compared with FA exposure.
VT decreased and fB increased with continuous 50-min O3
exposures; decrease in FVC, FEVj, and FEF25.75 from FA to
0.20 ppm and FA to 0.35-ppm O3 exposure in all conditions;
three subjects unable to complete continuous and competitive
protocols at 0.35 ppm O3.
FVC decreased with 0.50- and 0.75-ppm O3 exposure
compared with FA; 4% nonsignificant decrease in mean
VO2max following 0.75 ppm O3 compared with FA exposure.

Exposure to 0. 15 and 0.30 ppm O3 did not decrease maximal
exercise performance or VO2max compared to FA.
No significant pulmonary function or symptom responses were
observed, although a trend (P < .10) was evident.
Decrease in FVC, FEVj, ERV, 1C, and FEF50% after 1-h
0.75-ppm O3 exposure; decrease in VO2max, VTmax, VEmax,
maximal workload, and HRmax following 0.75-ppm O3
exposure compared with FA.



Reference
Linder et al.
(1988)



Folinsbee et al.
(1986)

Foxcroft and
Adams (1986)

Gong et al.
(1986)

Schelegle and
Adams (1986)


Folinsbee et al.
(1984)
Adams and
Schelegle (1983)



Horvath et al.
(1979)


Savin and
Adams (1979)


Folinsbee et al.
(1977)


"See Appendix A for abbreviations and acronyms.
""Listed from
lowest to highest O3 concentration.

-------
 1     0.75 ppm, although FVC was significantly decreased 10% following the latter exposure. Without
 2     preliminary exposure to O3, Savin and Adams (1979) examined the effects of a 30-min exposure
 3     to 0.15 and 0.30 ppm O3 while performing a progressively incremented exercise test to volitional
 4     fatigue (mean = 31.5 min in FA). No significant effect on maximal work time or VO2max was
 5     observed compared to that observed upon FA exposure.  Further, no significant effect on
 6     pulmonary function, maximal heart rate, and maximal tidal volume was observed, although
 7     maximal VE  was significantly reduced 7% in the 0.30 ppm O3 exposure. Results of these early
 8     studies suggest that VO2max is reduced if the incremented maximal exercise test is preceded by an
 9     O3 exposure  of sufficient total inhaled dose of O3 to result in significant pulmonary function
10     decrements and symptoms of breathing discomfort.
11          Using trained nonathletes, Foxcroft and Adams (1986) observed significant (p < 0.05)
12     reductions in rapidly incremented VO2max exercise performance time (-16.7%), VO2max (-6.0%),
13     maximal VE  (-15.0%), and maximal heart rate (-5.6%) immediately following an initial 50-min
14     exposure to 0.35 ppm O3 during heavy CE (VE = 60 L/min).  These decrements were
15     accompanied by a significant reduction in FEVj (-23%) and the occurrence of marked
16     symptoms of breathing discomfort.  Similarly, Gong et al. (1986) found significant reductions in
17     rapidly incremented VO2max exercise performance time (-29.7%), VO2max (-16.4%), maximal VE
18     (-18.5%), and maximal workload (-7.8%) in endurance cyclists immediately following a  1-h
19     exposure to 0.20 ppm O3 with very heavy exercise ( VE 89 L/min), but not following exposure to
20     0.12 ppm. Gong et al. (1986) observed only  a 5.6% FEVj decrement and mild symptoms
21     following exposure to 0.12 ppm, but a large decrement in FEVj (-21.6%) and substantial
22     symptoms of breathing discomfort following the 0.20 ppm exposure, which the authors
23     contended probably limited maximal performance and VO2max.
24
25     AX6.7.3  Effect on Endurance Exercise Performance
26          A number of studies of well trained endurance athletes exposed to O3 have consistently
27     observed an  impairment of 1-h continuous heavy exercise performance of some individuals
28     (Adams and  Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1984; Gong et al.,  1986).  The
29     performance impairment is indicated by an inability to complete the prescribed O3 exposures
30     (even at concentrations as low as 0.16 ppm) that subjects were able to complete in FA (Avol

       January 2005                           AX6-80      DRAFT-DO NOT QUOTE OR CITE

-------
 1      et al., 1984).  Other indications of impaired endurance exercise performance upon exposure to O3
 2      include a -7.7% reduced endurance treadmill running time when exposed to 0.18 ppm O3
 3      (Folinsbee et al., 1986), which was accompanied by significantly decreased FEVj and
 4      significantly elevated symptoms of breathing discomfort. Another study (Schelegle and Adams,
 5      1986) observed the failure of some trained endurance athletes to complete a 1-h competitive
 6      simulation protocol upon exposure to O3 (30 min warm-up, followed immediately by 30 min at
 7      the maximal workload that each subject could just maintain in FA; mean VE = 120 L/min).
 8      In this study, all subjects (n = 10) completed the FA exposure, whereas one, five, and seven
 9      subjects could not complete the 0.12, 0.18, and 0.24 ppm O3 exposures, respectively.  Following
10      the 0.18 ppm and 0.24 ppm O3 exposures, but not the 0.12 ppm exposure, FEVj was reduced
11      significantly and symptoms were significantly increased. Linder et al. (1988) also observed
12      small decrements in performance time (1 to 2 min) during a progressive maximal exercise test
13      (mean = 21.8 min) at O3 concentrations of 0.065 and 0.125 ppm.  These small effects were
14      accompanied by a significant increase in subjective perception of overall effort at 0.125 ppm, but
15      with no significant reduction in FEVj at either O3 concentration.  Collectively, reduced
16      endurance exercise performance and associated pulmonary responses are clearly related to the
17      total inhaled dose of O3 (Adams and Schelegle, 1983; Avol et al., 1984;  Schelegle and Adams,
18      1986).
19          Mechanisms limiting VO2max and maximal exercise performance upon O3 exposure have
20      not been precisely identified. Schelegle and Adams (1986) observed no significant effect of O3
21      on cardiorespiratory responses, and  there was no indirect indication that arterial O2 saturation
22      was affected. The latter is consistent with the observation that measured arterial O2 saturation at
23      the end of a maximal endurance treadmill run was not affected by O3 (Folinsbee et al., 1986).
24      In studies in which O3 inhalation resulted in a significant decrease in VO2max, and/or maximal
25      exercise performance,  significantly decreased FEVj and marked symptoms of breathing
26      discomfort were observed (Adams and Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1977,
27      1984, 1986; Foxcroft and Adams, 1986; Gong et al.,  1986; Schelegle and Adams, 1986).
28      However, Gong et al. (1986) observed rather weak correlations between FEVj impairment and
29      physiological variable responses during maximal exercise (R = 0.26 to 0.44).  Rather, these
30      authors concluded that substantial symptoms of breathing discomfort consequent to 1  h of very
31      heavy exercise while exposed to 0.20 ppm O3, probably limited maximal performance and

        January 2005                           AX6-81     DRAFT-DO NOT QUOTE OR CITE

-------
 1     VO2max either voluntarily or involuntarily (Gong et al., 1986). Strong support for this contention
 2     is provided by the observation of significant increases in VO2max (4.7%) and maximal
 3     performance time (8.8%) following four consecutive days of 1 h exposure to 0.35 ppm O3 with
 4     heavy exercise (VE  =60 L/min) compared to initial O3 exposure (Foxcroft and Adams, 1986).
 5     These improvements, which were not significantly different from those for FA, were
 6     accompanied by a significant reduction in symptoms of breathing discomfort with no significant
 7     attenuation of FEVj and other pulmonary function responses. In this regard, Schelegle et al.
 8     (1987) observed a disparate effect of indomethacin pretreatment (an inhibitor of the cyclo-
 9     oxygenation of arachidonic acid to prostaglandins associated with inflammatory responses) on
10     O3-induced pulmonary function response (significant reduction) and an overall rating of
11     perceived exertion and symptoms of pain on deep inspiration and shortness of breath (no
12     significant effect).
13
14
15     AX6.8   EFFECTS ON AIRWAY RESPONSIVENESS
16           Increased airway responsiveness, also called airway hyperresponsiveness (AHR) or
17     bronchial hyperreactivity, indicates that the airways are more reactive to bronchoconstriction
18     induced by a variety of stimuli  (e.g., specific allergens, exercise, SO2, cold air) than they would
19     be when normoreactive. In order to determine the level of airway responsiveness, airway
20     function (usually assessed by spirometry or plethysmography) is measured after the inhalation of
21     small amounts of an aerosolized specific (e.g., antigen, allergen) or nonspecific (e.g.,
22     methacholine, histamine) bronchoconstrictor agent or measured stimulus (e.g., exercise, cold
23     air). The dose or concentration of the agent or stimulus is increased from a control, baseline
24     level (placebo) until a predetermined degree of airway response, such as a 20% drop in FEVj or
25     a  100% increase in Raw, has occurred (Cropp et al., 1980; Sterk et al.,  1993). The  dose or
26     concentration of the bronchoconstrictor agent that produced the increased responsiveness often is
27     referred to as the "PD^FEVj" or "PC^FEVj" (i.e., the provocative dose or concentration that
28     produced a 20% drop in FEVj) or the "PD100SRaw" (i.e., the provocative dose that  produced a
29     100% increase in SRaw). A high level of bronchial responsiveness is a hallmark of asthma. The
30     range of nonspecific bronchial responsiveness, as expressed by the PD20 for example, is at least
31     1,000-fold from the most sensitive asthmatics to the least sensitive healthy  subjects.

       January 2005                            AX6-82      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Unfortunately, it is difficult to compare the PD2QFEVJ or PD100SRaw across studies because of
 2      the many different ways of presenting dose response to bronchoconstrictor drugs, for example,
 3      by mg/mL, units/mL, and molar solution; or by cumulative dose (CIU or CBU) and doubling
 4      dose (DD).  Other typical bronchial challenge tests with nonspecific bronchoconstrictor stimuli
 5      are based on exercise intensity or temperature of inhaled cold air.
 6           Increases in nonspecific airway responsiveness were previously reported as an important
 7      consequence of exposure to O3 (e.g., Golden et al., 1978; Table AX6-11). Konig et al. (1980)
 8      and Holtzman et al. (1979) found the increased airway responsiveness after O3 exposure in
 9      healthy subjects appeared to be resolved after 24 h. Because atopic subjects had similar
10      increases in responsiveness to histamine after O3 exposure as non-atopic subjects, Holtzman
11      et al. (1979) concluded that the increased nonspecific bronchial responsiveness after O3 exposure
12      was not related to atopy. Folinsbee and Hazucha (1989) showed increased airway
13      responsiveness in 18 female subjects 1 and 18 h after exposure to 0.35 ppm O3. Taken together,
14      these studies suggest that O3-induced increases in  airway responsiveness usually resolve 18 to
15      24 h after exposure, but may persist in some individuals for longer periods.
16           Gong  et al. (1986) found increased nonspecific airway responsiveness in elite cyclists
17      exercising at competitive levels with O3 concentrations as low as 0.12 ppm. Folinsbee et al.
18      (1988) found an approximate doubling of the mean methacholine responsiveness in a group of
19      healthy volunteers exposed for 6.6 h to 0.12 ppm O3.  Horstman et al. (1990) demonstrated
20      significant decreases in the PD100SRaw in 22 healthy subjects immediately after a 6.6-h exposure
21      to concentrations of O3 as low as 0.08 ppm.  No relationship was found between O3-induced
22      changes in airway responsiveness and changes in FVC or FEVj (Folinsbee et al.,  1988; Aris
23      et al., 1995), suggesting that changes in airway responsiveness and spirometric volumes occur by
24      different mechanisms.
25           Dimeo et al. (1981) were the first to investigate attenuation of the O3-induced increases in
26      nonspecific  airway responsiveness after repeated O3 exposure.  Over 3 days of a 2 h/day
27      exposure to  0.40 ppm O3, they found progressive attenuation of the increases in airway
28      responsiveness such that, after the third day of O3  exposure, histamine airway responsiveness
29      was no longer different from the sham exposure levels.  Kulle  et al. (1982) found that there was
30      a significantly enhanced response to methacholine after the first 3 days of exposure, but this
31      response slowly normalized by the end of the fifth day.  Folinsbee et al. (1994) found a more

        January 2005                             AX6-83      DRAFT-DO NOT QUOTE OR CITE

-------
                                     TABLE AX6-11. Airway Responsiveness Following Ozone Exposures"
to
o
o
        Ozone

        Concentrationb
        ppm
ug/m3
Exposure Duration     Exposure      Number and          Subject

   and Activity        Conditions   Gender of Subjects    Characteristics
Observed Effect(s)
                                                                                                                                  Reference
0.125
0.250
0.125
245 3hIE
490 (10 min rest, 15 min
exercise on bicycle)
VE = 30 L/min
245 3hIEx4days
27 °C 5 F, 6 M Mild bronchial
50 % RH 20-53 years old asthma
6F, 16 M
1 9-48 years old Allergic rhinitis
Mean early-phase FEV; response and number
of > 20% reductions in FEV! were
significantly greater after 0.25 ppm O3 or
4 x 0.125 ppm O3. Most of the > 15% late-
phase FEV; responses occurred after
exposure to 4 x 0.125 ppm O3, as well as
significant inflammatory effects, as indicated
by increased sputum eosinophils (asthma and
allergic rhinitis) and increased sputum
lymphocytes, mast cell tryptase, histamine,
and LDH (asthma only).
Holz et al.
(2002)
X
ON

oo
H

6
o


o
H

O
o
HH
H
W
0.4



0.12



0.2





0.4



0.16



784 2 h IE
VE = 20 L/min/m2
BSA

235 45 min IE
1 5 min exercise
VE = 3 x resting

392 4 h IE
VE = 25 L/min/m2
BSA



784 2 h IE
40 min/h @ 50 W


314 7.6 h IE
50 min/h
VE s25 1/min

NA



60% RH for
test
atmospheres

20° C
62% RH




NA



22°C
40% RH


6F
1M
19-26 years old

12 F
5M
19-3 8 years old

4F
8M
23-47 years old



15 healthy subjects ;
9F,6M;31.1±2.1
years old

5F
4M


Stable mild asthma;
no meds 8 h
preexposure

Physician diagnosed
asthma; SO2-induced
airway
hyperreactivity
Healthy nonsmokers





Healthy; nonatopic



Mild atopic
asthma, HMD
sensitive, 20-35 years
old
Increased bronchial responsiveness to
methacholine 16 h after exposure; inhaled
apocynin treatment significantly reduced O3-
induced airway responsiveness
Dietary supplementation with 400 IU Vit E +
500 mg Vit C reduced airway responsiveness
to 0.10 and 0.25 ppm SO2 challenge

Increased sputum total cells, % neurtophils,
IL-6, and IL-8 at 18 h after exposure;
increased airway responsiveness to
methacholine 2 h after postexposure FEV;
returned to 5% of base-line; no anti-
inflammatory effect of azithromycin
Decreased FEV; and FVC; increased
bronchial reactivity to methacholine
4 h post-exposure; no protection from
inhaled corticosteroid, budesonide
Mean 9.1% FEV; decrease 18 h after O3
exposure; provocative dose of dust mite
allergen decreased from 10.3 to 9.7 dose
units.
Peters et al.
(2001)


Trenga et al.
(2001)


Criqui et al.
(2000)




Nightingale
et al. (2000)


Kehrl et al.
(1999)



-------
                           TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures3
to
o
o
X
ON

oo
H

6
o


o
H

O
o
HH
H
W
Ozone
Concentration1"
ppm
0.2





0.12
Air-antigen
0.4


0.2






0.4






0.12

0.25





ug/m3
392





235

784


392






784






236

490





Exposure Duration
and Activity
4hIE
40 min/h @ 50 W




1 h rest

2hIE
VE= 20 L/min/m2
BSA
4hIE
50 min/h
VE = 25 L/min/m2
BSA



3 h/d for 5 days;
alternating 15 min of rest
and exercise at
VE = 32 L/min



Rest

3hIE
VE = 30 L/min
1 5 min ex/
10 min rest/
5 min no O3; every
30 min.
Exposure
Conditions
NA





NA

NA


20° C
50% RH





31°C
35% RH





22 °C
40% RH
27 °C
54% RH
mouthpiece
exposure


Number and
Gender of Subjects
1 0 asthmatic (6 F, 4
M),
26.6 ± 2.3 years old;
1 0 healthy (4 F, 6
M),
27.3 ±1.4 years old.
6F
9M
5F
1M
18-27 years old
6F
12 M
18-36 years old




2F
8M
19-48 years old




5F
10M
24 mild asthmatics
11F/13M
12 allergic rhinitics
6M/6F


Subject
Characteristics
Mild atopic asthma;
non-atopic healthy
subjects; no meds
8 weeks pre-exposure


Mild allergic asthma;
1 8 to 49 years if age
Stable mild asthma;
no meds 8 h
preexposure
Phy sician-diagno sed
mild asthma; no meds
prior to exposure




Mild asthma
requiring only
occasional
bronchodilator
therapy


atopic
asthma
atopic mild asthmatic
NS




Observed Effect(s)
Decreased FEV; in asthmatic (9.3%)
and healthy (6.7%) subjects; increased
sputum neutrophils in both groups (NS);
no change in methacholine airway reactivity
24 h post-exposure

No effect of O3 on airway response to grass
or ragweed allergen.
Increased airway responsiveness to
methacholine 16 h postexposure; no effect
of proteinase inhibitor, rALP
Decreased FEV; and FVC, increased SRaw;
lower respiratory Sx; increased
% neutrophils, total protein, LDH,
fibronectin, IL-8, GM-CSF, and MPO
in BAL. Correlation between pre-exposure
methacholine challenge and O3-induced
SRaw increase.
Significant FEV[ and Sx response on 1 st and
2nd O3 exposure days, then diminishing with
continued exposure; tolerance partially lost
4 and 7 days postexposure; bronchial
reactivity to methacholine peaked on 1 st O3
exposure day, but remained elevated with
continued exposure
No effect of O3 on airway response to grass
allergen.
Increased allergen responsiveness afer O3
exposure.




Reference
Nightingale
etal. (1999)




Hanania
etal. (1998)
Hiltermann
etal.
(1998)
Balmes et al.
(1997);
Scannell
etal. (1996)



Gong et al.
(1997b)





Ball et al.
(1996)
Jorres et al.
(1996)





-------
                           TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures3
X
Oi
oo
Oi
H
6
o

o
H
O
Ozone
Concentrationb
ppm
0.2




0.12






0.12





0.10
0.25
0.40


Air-antigen
0.12 ppm
O3-antigen
0.08
0.10
0.12

ug/m3
392




235






235





196
490
785





157
196
235

Exposure Duration
and Activity
4hIE
50min/10min
exercise/rest each hour


IhR






6.6 h, IE x 5 days
50 min exercise/10 min
rest, 30 min lunch
VE=38.8L/min


1 h light IE
2x15 min on treadmill
VE = 27 L/min


1 h at rest


6.6 h
IE at « 39 L/min


Exposure
Conditions
22 °C
50% RH



Ambient T
&RHfor
exposure;
23°C&
50% RH for
exercise
challenge
18 °C
40% RH




21 °C
40% RH



NA


18 °C
40% RH


Number and Subject
Gender of Subj ects Characteristics
42M/24F 18-50 years
NS healthy



8 F Mild stable asthma
7M
19-45 years old




17 M Healthy nonsmokers
25 ± 4 years old




9 F Stable mild
12 M asthmatics with FEV;
1 9-40 years old > 70% and
methacholine
responsiveness
4 M, 3 F Asthmatic,
21 to 64 years old

22 M Healthy NS,
18 to 32 years old


Observed Effect(s)
FEV^- 18.6%), FVC (- 14.6%), decreased
after O3. Baseline PC100 for methacholine
was not related to changes in FVC, FEVl5
a weak association was seen for PC100 and
increased SRaw.
No significant difference in % fall FEV[ or
V40p; no increase in bronchial responsiveness
to exercise challenge




FEV[ responses were maximal on 1st day of
exposure (-13%), less on second day (-9%),
absent thereafter. Sx responses only the first
2 days. Methacholine airway responsiveness
was at least doubled on all exposure days,
but was highest on the second day of O3.
No significant differences in FEV[ or FVC
were observed for 0.10 and 0.25 ppm O3-FA
exposures or postexposure exercise challenge;
12 subjects exposed to 0.40 ppm O3 showed
significant reduction in FEV[.
Increased bronchoconstrictor response to
inhaled ragweed or grass after O3 exposure
compared to air.
33, 47, and 55% decreases in cumulative dose
of methacholine required to produce a 100%
increase in SRaw after exposure to O3 at 0.08,
0.10, and 0.12 ppm, respectively.
Reference
Aris et al.
(1995)



Fernandes
etal. (1994)





Folinsbee
etal. (1994)




Weymer
etal.
(1994)


Molfino
etal.
(1991)
Horstman
etal. (1990)


o
HH
H
W

-------
3
to
o
o
                           TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures'1
X
Oi

oo
H

6
O


o
H

O
Ozone
Concentrationb
ppm Mg/m3
0.12 ppm
CylOOppb
S02
0.12 ppm
O3-0.12ppm
03
Air-100ppbSO2
0.35 686


0.40 784





0.12 235




0.12 235
0.20 392

0.40 784


0.20 392
0.40 784
0.40 784




— Exposure Duration Exposure
and Activity Conditions
45 min in first atmosphere 22 °C
and 1 5 min in second IE 75% RH





70 min with IE at NA
40 L/min

2 h with IE at 22 °C
VE = 53 to 55 L/min 50% RH




6. 6 h with IE at NA
«25 L/min/m2 BSA



lhat VE= 89 L/min 31 °C
followed by 3 to 4 min 35% RH
at -150 L/min
3 h/day for 5 days in a
row

2 h with IE at 2 x resting 22 °C
2 h with IE at 2 x resting 55% RH
2 h/day for 3 days




Number and Subject
Gender of Subj ects Characteristics
8 M, 5 F Asthmatic,
12 to 18 years old





18 F Healthy NS,
19 to 28 years old

8M, 10 F 9 asthmatics (5 F,
4M),
9 healthy (5 F, 4 M),
18 to 34 years old


10 M Healthy NS,
18 to 33 years old



15M,2F Elite cyclists,
19 to 30 years old

13M, 11 F Healthy NS,
1 9 to 46 years old

12 M, 7 F Healthy NS,
21 to 32 years old





Observed Effect(s)
Greater declines in FEV[ and Vmaj,50o/0 and
greater increase in respiratory resistance after
O3-SO2 than after O3-O3 or air-SO2.




PD100 decreased from 59 CIU after air
exposure to 41 CIU and 45 CIU, 1 and
18 h after O3 exposure, respectively.
Decreased PC100SRsw from 33 mg/mL to
8.5 mg/mL in healthy subjects after O3.
PC100SRllw fell from 0.52 mg/mL to
0.19 mg/mL in asthmatic subjects after
exposure to O3 and from 0.48 mg/mL
to 0.27 mg/mL after exposure to air.
Approximate doubling of mean
methacholine responsiveness after exposure.
On an individual basis, no relationship
between O3-induced changes in airway
responsiveness and FEV; or FVC.
Greater than 20% increase in histamine
responsiveness in one subject at 0.12 ppm
O3 and in nine subjects at 0.20 ppm O3.
Enhanced response to methacholine after
first 3 days, but this response normalized by
Day5.
110% increase in ASRaw to a 10-breath
histamine (1.6%) aerosol challenge after
exposure to O3 at 0.40 ppm, but no change
at 0.20 ppm. Progressive adaptation of this
effect over 3-day exposure.


Reference
Koenig et al.
(1990)





Folinsbee
and Hazucha
(1989)
Kreit et al.
(1989)




Folinsbee
etal. (1988)



Gong et al.
(1986)

Kulle et al.
(1982)

Dimeo et al.
(1981)



o
HH
H
W

-------
                              TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures3
1
to
o
o














1
oo
oo

Ozone
Concentration1"
ppm Mg/m3 and Activity Conditions Gender of Subjects Characteristics
0.10 196 2h NA 14 HealthNS,
0.32 627 24 ± 2 years old
1.00 1,960
0.60 1,176 2 h with IE at 2 x resting 22 °C 1 1 M, 5 F 9 atopic,
5 5 % RH 7 nonatopic ,
NS, 21 to 35 years
old




0.6 1,176 2 hat rest NA 5 M, 3 F Healthy NS,
22 to 30 years old




Observed Effect(s)
Increased airway responsiveness
to methacholine immediately after exposure
at the two highest concentrations of O3.
Ten-breath methacholine or histamine
challenge increased SRaw > 150% in
16 nonasthmatics after O3. On average, the
atopic subjects had greater responses than
the nonatopic subjects. The increased
responsiveness resolved after 24 h. Atropine
premedication blocked the O3-induced
increase in airway responsiveness.
300% increase in histamine-induced ARaw
5 min after O3 exposure; 84 and 50%
increases 24 h and 1 week after exposure
(p > 0.05), respectively. Two subjects had
an increased response to histamine 1 week
after exposure.
Reference
Konig et al.
(1980)

Holtzman
etal. (1979)






Golden et al.
(1978)




H
6
o

o
H
O
       aSee Appendix A for abbreviations and acronyms.

       bListed from lowest to highest O3 concentration.
o
HH
H
W

-------
 1      persistent effect of O3 on airway responsiveness which was only partially attenuated after
 2      5 consecutive days of O3 exposure.
 3           The occurrence and duration of increased nonspecific airway responsiveness following O3
 4      exposure could have important clinical implications for asthmatics.  Kreit et al. (1989)
 5      investigated changes in airway responsiveness to methacholine that occur after O3 exposure in
 6      mild asthmatics. They found that the baseline PC100SRaw declined from 0.52 to 0.19 mg/mL
 7      after a 2-h exposure to 0.40 ppm O3  as compared to a decline from 0.48 to 0.27 mg/mL after air
 8      exposure; however, because of the large variability in responses of the asthmatics, the percent
 9      decrease from baseline in mean PC100SRaw was not statistically different between healthy and
10      asthmatic subjects (74.2 and 63.5%,  respectively).
11           Two studies examined the effects of preexposure to O3 on exacerbation of exercise-induced
12      bronchoconstriction (Fernandes et al., 1994; Weymer et al., 1994). Fernandes et al. (1994)
13      preexposed subjects with stable mild asthma and a history of > 15% decline in FEVj after
14      exercise to 0.12 ppm O3 for 1 h at rest followed by a 6-min exercise challenge test and found no
15      significant effect on either the magnitude or time course of exercise-induced
16      bronchoconstriction. Similarly, Weymer et al. (1994) observed that preexposure to either 0.10 or
17      0.25 ppm O3 for 60 min while performing light IE did not enhance or produce exercise-induced
18      bronchoconstriction in otherwise healthy adult subjects with stable mild asthma. Although the
19      results suggested that preexposure to O3 neither enhances nor produces exercise-induced asthma
20      in asthmatic subjects, the relatively low total inhaled doses of O3 used in these studies limit the
21      ability to draw any definitive conclusions.
22           Gong et al. (1997b) found that subjects with asthma developed tolerance to repeated
23      O3 exposures in a manner similar to normal subjects; however,  there were more persistent effects
24      of O3 on airway responsiveness, which only partially attenuated when compared to filtered air
25      controls.  Volunteer subjects with mild asthma requiring no more than bronchodilator therapy
26      were exposed to filtered air or 0.4 ppm O3, 3 h/d for 5 consecutive days, and follow-up
27      exposures 4 and 7 days later. Symptom and FEVj responses were large on the 1st and 2nd
28      exposure days,  and diminished progressively toward filtered air responses by the 5th exposure
29      day.  A methacholine challenge was  performed when postexposure FEVj  returned to within 10%
30      of preexposure baseline levels. The  first O3 exposure significantly decreased PD^FEVj by an
31      order of magnitude and subsequent exposures resulted in smaller decreases, but they were still

        January 2005                             AX6-89      DRAFT-DO NOT QUOTE OR CITE

-------
 1      significantly different from air control levels. Thus, the effects of consecutive O3 exposures on
 2      bronchial reactivity differ somewhat from the effects on lung function. The same conclusion
 3      was drawn by Folinsbee et al. (1994) after consecutive 5-day O3 exposures in healthy subjects,
 4      despite a much lower bronchial reactivity both before and after O3 exposure.
 5           A larger number of studies examined the effects of O3 on exacerbation of antigen-induced
 6      asthma. Molfmo et al. (1991) were the first to report the effects of a 1-h resting exposure to
 7      0.12 ppm O3 on the response of subjects with mild, stable  atopic asthma to a ragweed or grass
 8      allergen inhalation challenge. Allergen challenges were performed 24 h after air and
 9      O3 exposure. Their findings suggested that allergen-specific airway responsiveness of mild
10      asthmatics is increased after O3 exposure. However, Ball  et al. (1996) and Hanania et al. (1998)
11      were unable to confirm the findings of Molfmo et al. (1991) in a group of grass-sensitive mild
12      allergic asthmatics exposed to 0.12 ppm O3 for 1  h. The differences between Hanania et al.
13      (1998) and Molfmo et al. (1991), both conducted in the same laboratory, were due to better, less
14      variable control of the 1 h 0.12  ppm O3 exposure and better study design by Hanania and
15      colleagues.  In the original, Molfmo et al. (1991) study, the control (air) and experimental (O3)
16      exposures were not randomized after the second subject because of long-lasting (3 months),
17      O3-induced potentiation of airway reactivity  in that subject. For safety reasons,  therefore, the air
18      exposures were performed prior to the O3 exposures for the remaining 5 of 7 subjects being
19      evaluated. It is possible that the first antigen challenge caused the significant increase in the
20      second (post-O3) antigen challenge.
21           Torres et al. (1996) later confirmed that higher O3 concentrations cause increased airway
22      reactivity to specific antigens in subjects with mild allergic asthma, and to a lesser extent in
23      subjects with allergic rhinitis, after exposure to 0.25 ppm O3 for 3 h.  The same laboratory
24      repeated this study in separate groups of subjects with asthma and rhinitis and found similar
25      enhancement of allergen responsiveness after O3  exposure (Holz et al., 2002); however, the
26      effects of a  3-h exposure to 0.25 ppm O3 were more variable, most likely  due to performing the
27      allergen challenges 20 h after exposure, rather than the 3 h used in the first study.
28           The timing of allergen challenges in O3-exposed subjects with allergic asthma is important.
29      Bronchial provocation with allergen, and subsequent binding with IgE antibodies on mast cells
30      in the lungs, triggers the release of histamine and leukotrienes and a prompt early-phase
31      contraction of the smooth muscle cells of the bronchi, causing a narrowing of the lumen of the

        January 2005                             AX6-90      DRAFT-DO NOT QUOTE OR CITE

-------
 1      bronchi and a decrease in bronchial airflow (i.e., decreased FEVj). In many asthma patients,
 2      however, the release of histamine and leukotrienes from the mast cells also attracts an
 3      accumulation of inflammatory cells, especially eosinophils, followed by the production of mucus
 4      and a late-phase decrease in bronchial airflow for 4 to 8 h.
 5           A significant finding from the study by Holz et al. (2002) was that clinically relevant
 6      decreases in FEVj (> 20%) occurred during the early-phase allergen response in subjects with
 7      rhinitis after a consecutive 4-day exposure to 0.125 ppm O3.  Kehrl et al. (1999) previously
 8      found an increased reactivity to house dust mite antigen in asthmatics 16 to  18 h after exposure
 9      to 0.16 ppm O3 for 7.6 hours. These important observations indicate  that O3 not only causes
10      immediate increases in airway-antigen reactivity, but that this effect may persist for at least 18 to
11      20 h.  Ozone exposure, therefore, may be a clinically important co-factor in the response to
12      airborne bronchoconstrictor substances in individuals with pre-existing allergic asthma.  It is
13      plausible that this phenomenon could contribute to increased symptom exacerbations and, even,
14      consequent increased physician or ER visits, and possible hospital admissions (see Chapter 7).
15           A number of human studies, especially more recent ones, have been undertaken to
16      determine various aspects of O3-induced increases in nonspecific airway responsiveness, but
17      most studies have been conducted in laboratory animals (See the toxicology chapter, Section
18      5.3.4.4.). In humans, increased airway permeability (Kehrl et al.,  1987; Molfmo et al., 1992)
19      could play a role in increased airway responsiveness.  Inflammatory cells and mediators also
20      could affect changes in airway responsiveness. The results of a multiphase study (Scannell
21      et al.,  1996; Balmes et al., 1997) showed a correlation between preexposure methacholine
22      responsiveness in healthy subjects and increased SRaw caused by a 4 h exposure to 0.2 ppm O3,
23      but not with O3-induced decreases in FEVj and FVC.  The O3-induced increase in SRaw, in turn,
24      was correlated with O3-induced increases in neutrophils and total protein concentration in BAL
25      fluid.  Subjects with asthma had a significantly greater inflammatory  response to the same O3
26      exposures, but it was not correlated with increased SRaw, and nonspecific airway provocation
27      was not measured.  Therefore, it is difficult to determine from this series of studies if underlying
28      airway inflammation plays a role in increased airway responsiveness  to nonspecific
29      bronchoconstrictors. The study, however,  confirmed an earlier observation (e.g., Balmes et al.,
30      1996) that O3-induced changes in airway inflammation and lung volume measurements are not
31      correlated.

        January 2005                             AX6-91      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Hiltermann et al. (1998) reported that neutrophil-derived serine proteinases associated with
 2      O3-induced inflammation are not important mediators for O3-induced nonspecific airway
 3      hyperresponsiveness. Subjects with mild asthma, prescreened for O3-induced airway
 4      responsiveness to methacholine, were administered an aerosol of recombinant antileukoprotease
 5      (rALP) or placebo at hourly intervals two times before and six times after exposure to filtered air
 6      or 0.4 ppm O3 for 2 h.  Methacholine challenges were performed 16 h after exposure. Treatment
 7      with rALP had no effect on the O3-induced decrease in FEVj or PC^FEVj in response to
 8      methacholine challenge.  The authors speculated that proteinase-mediated tissue injury caused
 9      by O3 may not be important in the development of airway hyperresponsiveness of asthmatics to
10      O3.  In a subsequent study using a similar protocol (Peters et al., 2001), subjects with mild
11      asthma were administered an aerosol of apocynin, an inhibitor of NADPH oxidase present in
12      inflammatory cells such as eosinophils and neutrophils, or a placebo. In this study, methacholine
13      challenge performed 16 h after O3 exposure showed treatment-related effects on PC2QFEVJ,
14      without an effect on FEVj. The authors concluded that apocynin could prevent O3-induced
15      bronchial hyperresponsiveness in subjects with asthma, possibly by preventing superoxide
16      formation by eosinophils and neutrophils in the larger airways.
17           Nightingale et al. (1999) reported that exposures of healthy subjects and subjects with mild
18      atopic asthma to a lower O3 concentration (0.2 ppm) for 4 h caused a similar neutrophilic  lung
19      inflammation in both groups but no changes in airway responsiveness to methacholine measured
20      24 h after O3 exposure in either group. There were, however,  significant decreases in FEVj of
21      6.7 and 9.3% immediately after O3 exposure in both healthy and asthmatic subjects, respectively.
22      In a subsequent study, a significant increase in bronchoresponsiveness to methacholine was
23      reported 4 h after healthy subjects were exposed to 0.4 ppm O3 for 2 h (Nightingale et al., 2000).
24      In the latter study, preexposure treatment with inhaled budesonide (a corticosteroid) did not
25      protect against O3-induced effects on spirometry,  methacholine challenge, or sputum neutrophils.
26      These studies also confirm the earlier reported findings that O3-induced increases in airway
27      responsiveness usually resolve by 24 h after exposure.
28           Ozone-induced airway inflammation and hyperresponsiveness were used by Criqui  et al.
29      (2000) to evaluate anti-inflammatory properties of the macrolide antibiotic, azithromycin. In a
30      double-blind, cross-over study, healthy volunteers were exposed to 0.2 ppm O3 for 4 h after
31      pretreatment with azithromycin or a placebo. Sputum induction 18 h postexposure resulted in

        January 2005                            AX6-92      DRAFT-DO NOT QUOTE OR CITE

-------
 1      significantly increased total cells, percent neutrophils, IL-6, and IL-8 in both azithromycin- and
 2      placebo-treated subjects.  Significant pre- to post-exposure decreases in FEVj and FVC also
 3      were found in both subject groups.  Airway responsiveness to methacholine was not significantly
 4      different between azithromycin-treated and placebo-treated subjects when they were challenged
 5      2 h after postexposure FEVj decrements returned to within 5 % of baseline. Thus, azithromycin
 6      did not have anti-inflammatory effects in this study.
 7          The effects of dietary antioxidants on O3-induced bronchial hyperresponsiveness were
 8      evaluated in adult subjects with asthma by Trenga et al. (2001). Recruited subjects were
 9      pretested for responsiveness to a provocative  SO2 challenge (0.10 and 0.25 ppm) while
10      exercising on a treadmill and selected for study if they experienced a > 8% decrease in FVC.
11      The rationale for this environmental challenge approach is based on previously published work
12      by this laboratory (Koenig et al., 1990).  In a placebo-controlled, double-blind crossover study,
13      subjects took two vitamin supplements (400 IU vitamin E and 500 mg vitamin C) or two
14      placebos once a day for 4 weeks, and were exposed to filtered air and to 0.12 ppm O3 for 45 min
15      during intermittent exercise at three times resting ventilation. Blood samples were used to verify
16      placebo and vitamin treatment levels.  Provocative airway challenges with SO2 were performed
17      immediately after O3 exposure. Ozone exposure potentiated the SO2 challenge in asthmatics,
18      and subjects given antioxidant supplementation responded less severely to the airway  challenge
19      than subjects given the placebo. The protective effect of antioxidants was even more
20      pronounced among subjects with more severe asthma and higher sensitivity to SO2.
21
22
23      AX6.9   EFFECTS ON INFLAMMATION AND HOST DEFENSE
24      AX6.9.1   Introduction
25          In general, inflammation can be considered as the host response to injury, and the
26      induction of inflammation can be accepted as evidence that injury has occurred.  Several
27      outcomes are possible: (1) inflammation can resolve entirely; (2) continued acute inflammation
28      can evolve into a chronic inflammatory state; (3) continued inflammation can alter the structure
29      or function of other pulmonary tissue, leading to diseases such as fibrosis or emphysema;
30      (4) inflammation can alter the body's host defense response to inhaled microorganisms; and
31      (5) inflammation can alter the lung's response to other agents such as allergens or toxins.

        January 2005                            AX6-93      DRAFT-DO NOT QUOTE OR CITE

-------
 1      At present, it is known that short-term exposure of humans to O3 can cause acute inflammation
 2      and that long-term exposure of laboratory animals results in a chronic inflammatory state (see
 3      Chapter 5).  However, the relationship between repetitive bouts of acute inflammation in humans
 4      caused by O3 and the development of chronic respiratory disease is unknown.
 5           Bronchoalveolar lavage (BAL) using fiberoptic bronchoscopy has been utilized to sample
 6      cells and fluids lining the respiratory tract primarily from the alveolar region, although the use of
 7      small volume lavages or balloon catheters permits sampling of the airways.  Cells and fluid can
 8      be retrieved from the nasal passages using nasal lavage (NL) and brush or scrape biopsy.
 9           Several studies have analyzed BAL and NL fluid and cells from O3-exposed humans for
10      markers of inflammation and lung damage (see Tables AX6-12 and AX6-13).  The presence of
11      neutrophils (PMNs) in the lung has long been accepted as a hallmark of inflammation and is an
12      important indicator that O3 causes inflammation in the lungs.  It is apparent, however, that
13      inflammation within airway tissues may persist beyond the point that inflammatory cells are
14      found in BAL fluid.  Soluble mediators of inflammation such as the cytokines (IL-6, IL-8) and
15      arachidonic acid metabolites (e.g., PGE2, PGF2a, thromboxane, and leukotrienes [LTs] such as
16      LTB4) have been measured in the BAL fluid of humans exposed to O3.  In addition to their role
17      in inflammation, many of these compounds have bronchoconstrictive properties and may be
18      involved in increased airway responsiveness following O3 exposure.
19           Some recent evidence suggests that changes in small airways function may provide a
20      sensitive indicator of O3 exposure and effect (see Section AX6.2.5), despite the fact that inherent
21      variability in their measurement by  standard spirometric approaches make their assessment
22      difficult.  Observations of increased functional responsiveness of these areas relative to the more
23      central airways, and of persistent effects following repeated exposure, may indicate that further
24      investigation of inflammatory processes in these regions is warranted.
25           Under normal circumstances, the epithelia lining the large and small airways develop tight
26      junctions and restrict the penetration of exogenous particles and macromolecules from the
27      airway lumen into the interstitium and blood, as well as restrict the flow of plasma components
28      into the airway lumen.  O3 disrupts the integrity of the epithelial cell barrier in human airways, as
29      measured by markers of plasma influx such as albumin, immunoglobulin, and other proteins into
30      the airways.  Markers of epithelial cell damage such as lactate dehydrogenase  (LDH) also have
31      been measured in the BAL fluid of humans exposed to O3. Other soluble factors that have been

        January 2005                            AX6-94      DRAFT-DO NOT QUOTE OR CITE

-------
             Table AX6-12. Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
Ozone Concentration11
ppm |ig/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Upper Airway Studies
0.4 784 2 h


0.2 392 2h
0.4 980 2h
0.12 235 1.5 h
0.24 470
0.5 980 4h
0.4 784 2 h

0.5 980 4hon
2 consecutive days
At rest


IE
(15 min/30 min);
(VE) « 20
L/min/m2 BSA
At rest
IE
(20 L/min) at
15 -min intervals
Resting
IE
(70 L/min) at
15 -min intervals
Resting
12 mild,
asymptomatic dust
mite-sensitive
asthmatics;
18-35 years of age

8 M, 5 F healthy NS
20-3 1 years of age
10 mild NS
asthmatics
18-3 5 years old
5 M, 5 F, asthmatic;
4 M, 4 F,
nonasthmatic;
18 to 41 years old
6 M, 6 F,
allergic rhinitics,
31.4 ± 2.0 (SD)
years old
11 M,
18 to 35 years old
41 M
(21 O3-exposed,
20 air-exposed),
18 to 35 years old
Release of early-onset mast cell-derived mediators into NL in
response to allergen not enhanced following O3 exposure.
Neutrophil and eosinophil inflammatory mediators were not
increased after O3 exposure or enhanced after allergen
challenge. O3 increased eosinophil influx following allergen
exposure.
No neutrophilia in NL samples by 1 .5 h post exposure.
Depletion of uric acid in NL fluid by 30% during h 2 of
exposure with increase in plasma uric acid levels.
No depletion of ascorbic acid, reduced glutathione,
extracellular superoxide dismutase.
Response to allergen increased (NS). PMN and
eosinophils increased after O3 plus allergen challenge.
Ozone alone increased inflammation in the nose.
NL done immediately and 24 h after exposure.
Increased number of PMNs at both times in asthmatic subjects
exposed to 0.24 ppm O3; no change in nonasthmatic subjects.
No change in lung or nasal function.
NL done immediately after exposure. Increased upper
and lower respiratory symptoms and increased levels of PMNs,
eosinophils, and albumin in NL fluid.
NL done immediately before, immediately after, and 22 h after
exposure. Increased numbers of PMNs at both times after
exposure; increased levels of tryptase, a marker of mast cell
degranulation, immediately after exposure; increased levels
of albumin 22 h after exposure.
NL done immediately before and after each exposure
and 22 h after the second exposure. Increased levels of PMNs
at all times after the first exposure, with peak values occurring
immediately prior to the second exposure.
Michelson
etal. (1999)

Mudway et al.
(1999)
Peden et al.
(1995)
McBride et al.
(1994)
Bascom et al.
(1990)
Graham and
Koren(1990)
Koren et al.
(1990)
Graham et al.
(1988)

-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
Oi
Ozone Concentrationb
ppm ug/m3 Duration (VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies
0.2 392 2h IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA

0.1 196 2h mild IE





0.2 392 2h IE
(15 min/30 min);
(VE) - 20
L/min/m2 body
surface area
0.27 529 2h IE
(20 min/ 60 min);
(VE) - 25
L/min/m2 BSA

0.22 431 4 h IE
(15 min/30 min);
(VE) = 25
L/min/m2 BSA





6M, 6F healthy,
nonatopic and
9 M, 6F mild
asthmatic subjects,
19-48 years of age
12 M, 10 F
healthy subjects
mean age -30
years


6M, 9F healthy
subjects and
9 M, 6F mild
asthmatics

12 subjects with
intermittent-mild
asthma exhibiting a
dual response; 18-37
years of age
12 nonsmoker,
nonresponders;
1 3 nonsmoker,
responders;
13 smokers;
1 8-40 years of age



Significantly higher baseline expression of IL-4 and IL-5 in
bronchial mucosal biopsies from asthmatic vs. healthy subjects
6 h post-exposure. Following O3 exposure, epithelial
expression of IL-5, GM-CSF, ENA-78, and IL-8 increased
significantly in asthmatics, as compared to healthy subjects.
Markers of exposure in exhaled breath condensate, including
increased 8-isoprostane, TEARS and LTB-4, and a marker of
ROS-DNA interaction in peripheral blood leukocytes
(8-OHdG), were increased in a sub-set of subjects bearing the
wild genotype for NAD(P)H:quinone oxidoreductase and the
null genotype for glutathione-S-transferase Ml .
No evidence seen for increased responsiveness to the
inflammatory effects of O3 in mild asthmatics versus healthy
subjects at 6 h following exposure. Used neutrophil
recruitment and exacerbation of pre-existing inflammation.

Exposure to O3 24 h following allergen challenge resulted in a
significant decrease in FEV1, FVC and VC and increase in
symptom scores compared to air exposure. The percentage of
eosinophils, but not neutrophils, in induced sputum was higher
6 h after O3 than after air.
Recovery of AM was approximately 3-fold higher in BAL
from smokers versus nonsmokers. Unstimulated AM from
smokers released ~2-fold greater amounts of superoxide anion
than from nonsmokers at 30 min and 18 h post-exposure, but
release was not further enhanced by stimulation of the cells.
ROS generation by AM from non-smokers decreased following
exposure at 18 h; markers of epithelial permeability increased.
No relationship was found between measures of ROS
production and lung function responsiveness to O3.
Bosson et al.
(2003)



Corradi et al.
(2002)




Stenfors et al.
(2002)



Vagaggini
et al. (2002)



Voter et al.
(2001)








-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
Ozone Concentration11
ppm ug/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.2 392 2h





0.4 784 2 h



0.4 784 2 h





0.4 784 1 h




0.2 392 2h

0.4 784 2 h/day for 5 days,
2 h either 10 or
20 days later



IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA


IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
At rest





Continuous
exercise;
(VE) - 30
L/min/m2 BSA



IE
(40 L/min) at
15 -min intervals



8M, 5F healthy
nonsmokers;
20-3 1 years of age



10M, 6F subjects
with intermittent
asthma;
19-35 years of age
12 mild,
asymptomatic dust
mite-sensitive
asthmatics;
18-35 years of age

4 healthy subjects




15 healthy
nonsmokers
16 M; 18 to 35 years
of age




Early (1 .5 h post-exposure) increase in adhesion molecule
expression, submucosal mast cell numbers and alterations in
lining fluid redox status. No clear relationship between early
markers of response and lung function deficits. 2.5-fold
increase in % human leukocyte antigen (HLA)-DR+ alveolar
macrophages in BAL.
In a cross-over study, levels of eosinophil cationic protein, IL-8
and percentage eosinophils were found to be highly correlated
in induced sputum and BAL 16 h following O3 exposure.

Release of early-onset mast cell-derived mediators into NL in
response to allergen not enhanced following O3 exposure.
Neutrophil and eosinophil inflammatory mediators were not
increased after O3 exposure or enhanced after allergen
challenge. O3 increased eosinophil influx following allergen
exposure.
Apoptosis was observed in cells obtained by airway lavage 6 h
following exposure. AM obtained by BAL showed the
presence of a 4-hydroxynonenal (HNE) protein adduct and the
stress proteins, 72-kD heat shock protein and ferritin. These
effects were replicated by in vitro exposure of AM to HNE.
Increased numbers of CD3+, CD4+, and CD8+ T lymphocyte
subsets, in addition to neutrophils, in BAL 6 h post-exposure.
BAL done immediately after fifth day of exposure and again
after exposure 10 or 20 days later. Most markers of
inflammation (PMNs, IL-6, PGE2, LDH, elastase, fibronectin)
showed complete attenuation; markers of damage did not.
Reversal of attenuation was not complete for some markers,
even after 20 days.
Blomberg
etal. (1999)




Hiltermann
etal. (1999)


Michelson
etal. (1999)




Hamilton et al.
(1998)



Blomberg
etal. (1997)
Devlin et al.
(1997)





-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
oo
Ozone Concentrationb
ppm |ig/m3
Exposure
Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.22 431







0.12 235



0.16 314


0.2 392


0.4 784



0.4 784






0.2 392


4h







2h



7.6 h


4h
T = 20 °C
RH = 50%
2h
T = 22 °C
RH = 50%

2h
1 5 min,
ex/ 15 min, rest




4h
T = 20 °C
RH = 50%
IE
20 min ex/1 9 min
rest
(VE) * 39-45 1/min




IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
IE
50 min/hr
(VE) = 25L/min
IE
(50 min/60 min);
(VE)«44L/min
1 5 min rest
1 5 min exercise
cycle ergometer
(VE) « 55 1/min
(VE) = 66 1/min






IE
(50 min/60 min);
(VE)«44L/min
31M,7F
smokers and
nonsmokers





9M, 3F healthy
nonsmokers; mean
age ~ 28 years

8 asthmatics sensitive
to dust mites

14 M, 6 F
healthy NS

1 1 healthy
nonsmokers;
18-3 5 years

8M
healthy nonsmokers





17 M, 6F
mild asthmatics

Post O3 exposure FEV; in 3 groups: Smokers (-13.9%);
non-responders (- 1 .4%) and responders (-28.5%). PMN's
increased immediately and at 18 h in all groups. Eosinophils
and lymphocytes increased after O3. IL-6 increased more in
non-smokers. No relationship of symptoms with inflammation,
lung function changes not related to inflammation. Nasal
lavage indicators did not predict bronchial or alveolar
inflammation.
Increase in the percentage of vessels expressing P-selectin in
bronchial biopsies at 1 .5 h post-exposure. No changes in
FEVj, FVC, inflammatory cells or markers in BAL, or vessels
expressing VCAM-1, E-selectin or ICAM-1 in biopsies.
Increased numbers of eosinophils in BAL after O3 exposure.


Ozone increased PMN, protein, IL-8, for all subjects.
No relationship of inflammation with spirometric responses.

Mean FEVl5 change = - 10%. BAL occurred at 0, 2, or 4 h
post-exposure. Small n limits statistical inference. Trend for
PMN's to be highest at 4 h. LTC4 increased at all time points.
No change in PGE2 or thromboxane.
Comparison of BAL at 1 h post-exposure vs. 18 h post-
exposure. At 1 h, PMN's, total protein, LDH, al-antitrypsin,
fibronectin, PGE2, thromboxane B2, C3a, tissue factor, and
clotting factor VII were increased. IL-6 and PGE2 were higher
after 1 h than 18 h. Fibronectin and tissue plasminogen
activator higher after 1 8 h. No time differences for PMN and
protein.
Increased PMN, protein, IL-8, LDH, in BAL. Inflammatory
responses were greater than a group of non-asthmatics
(Balmesetal., 1996)
Frampton
etal. (1997a)
Torres et al.
(1997)




Krishna et al.
(1997b)


Peden et al.
(1997)

Balmes et al.
(1996)

Coffey et al.
(1996)


Devlin et al.
(1996)
(compare with
Koren et al.
(1989a)


Scannell et al.
(1996)


-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
Ozone Concentrationb
ppm ug/m3
Exposure
Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.4 784
0.2 392
0.08 157
0.10 196
0.4 784
0.3 588
0.40 784
0.40 784
0.4 784
2 h mouthpiece
exposure
20 °C
42% RH
4h
6.6 h
2h
Ih
(mouth-piece)
2h
2h
2h
1 5 min exercise
1 5 min rest
(VE) « 40 1/min
IE
(50 min/60 min);
(VE) = 40 L/min
IE
(50 min/60 min) +
35 min lunch;
(VE) = 40 L/min
IE
(15 min/30min);
(VE) = 70 L/min
CE (60 L/min)
IE
(15 min/30 min);
(VE) = 70 L/min
IE
(15 min/30 min);
(VE) = 70 L/min
IE
(15 min/30 min);
(VE) = 70 L/min
5M, 5F
healthy; age ~ 30
15 M, 13 F,
21 to 39 years old
18 M,
18 to 35 years of age
10 M,
18 to 35 years old
5M
11M, 18 to
35 years old
11 M,
18 to 35 years old
11 M,
18 to 35 years old
Sputum induction 4 h after O3 exposure 3-fold increase in
neutrophils and a decrease in macrophages after O3 exposure.
IL-6, IL-8, and myeleperoxidase increased after O3. Possible
relationship of IL-8 and PMN levels.
Bronchial lavage, bronchial biopsies, and BAL done 1 8 h
after exposure. BAL shows changes similar to other studies.
Airway lavage shows increased cells, LDH, IL-8. Biopsies
show increased number of PMNs.
BAL fluid 18 h after exposure to 0.1 ppm O3 had significant
increases in PMNs, protein, PGE2, fibronectin, IL-6, lactate
dehydrogenase, and a-1 antitrypsin compared with the same
subjects exposed to FA. Similar but smaller increases in all
mediators after exposure to 0.08 ppm O3 except for protein and
fibronectin. Decreased phagocytosis of yeast by alveolar
macrophages was noted at both concentrations.
BAL fluid 1 h after exposure to 0.4 ppm O3 had significant
increases in PMNs, protein, PGE2, TXB2, IL-6, LDH, a-1
antitrypsin, and tissue factor compared with the same subjects
exposed to FA. Decreased phagocytosis of yeast by alveolar
macrophages.
Significantly elevated PMNs in the BAL fluid 1, 6, and 24 h
after exposure, with peak increases at 6 h.
Macrophages removed 18 h after exposure had changes in the
rate of synthesis of 123 different proteins as assayed by
computerized densitometry of two-dimensional gel protein
profiles.
BAL fluid 18 h after exposure contained increased levels of the
coagulation factors, tissue factor, and factor VII. Macrophages
in the BAL fluid had elevated tissue factor mRNA.
BAL fluid 18 h after exposure had significant increases in
PMNs, protein, albumin, IgG, PGE2, plasminogen activator,
elastase, complement C3a, and fibronectin.
Fahy et al.
(1995)
Aris et al.
(1993)
Devlin et al.
(1990,1991)
Koren et al.
(1991)
Koren et al.
(1991)
Schelegle
etal. (1991)
Devlin and
Koren (1990)
McGee et al.
(1990)
Koren et al.
(1989a,b)

-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
O
O
Ozone Concentrationb
ppm ug/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Repeated Exposure Studies
0.125 245 3 h exposures to
0.25 490 both O3 cones, and
to FA;
3 h on four
consecutive days to
0.125; study arms
separated by
>4 wks
0.25 490 2 h on four
consecutive days;
O3 and FA exposure
study arms
separated by >3
wks


0.2 392 single,
4 h exposures to O3
and to FA;
4 h on four
consecutive days to
O3; study arms
separated by
>4 wks

0.2 392 single, 4 h
exposure;
4 h exposures on
four consecutive
days; study arms
separated by
>4 wks
IE
(15 min/30 min)






IE
(30 min/60 min);
(VE) ~ 8 times the
FVC/min




IE
(15 min/30 min);
(MeanVE) =
14.8 L/min/m2
BSA




IE
(30 min/60 min);
(MeanVE) = 25
L/min/m2 BSA



5M, 6F allergic
asthmatic and
16M, 6F allergic
rhinitic subjects;
19-53 years of age



5M, 3F healthy
subjects;
25-31 years of age





15M, 8F healthy
subjects;
21-35 years of age






9M, 6F healthy NS
23-37 years of age





All subjects underwent 4 exposure arms and were challenged
with allergen 20 h following the last exposure in each. Sputum
was induced 6-7 h later. In rhinitics, but not asthmatics, the
incidence and magnitude of early phase FEV; decrements to
Ag were greater after 0.25 and 4x 0. 125 ppm O3. Repeated
exposure caused increases in neutrophil and eosinophil
numbers in both subject groups, as well as increased
percentage and number of lymphocytes in the asthmatics.
Maximal mean reductions in FEV; and FVC were observed on
day 2, and became negligible by day 4. FEF25.75, VmaxSO, and
Vmax75 were combined into a single value representing small
airway function (SAWgrp). This variable was the only one to
show persistent depression of the 24 h post-exposure baseline
from day 2 to day 5 measurements. Numbers of PMNs in BAL
fluid on day 5 were significantly higher in subjects following
O3, compared to air, exposures.
All subjects underwent 3 exposure arms with BAL and
bronchial mucosal biopsies performed 20 h following the last
exposure in each. After repeated exposure, functional and
BAL cellular responses were not different from those after FA,
whereas total protein, IL-6, IL-8, reduced glutathione and
ortho-tyrosine remained elevated. Also at this time,
macroscopic scores of inflammation and tissue neutrophils
were increased in mucosal biopsies. IL-10 was detected only
in BAL fluid following repeated O3 exposure.
Subjects were randomly assigned to each of the exposure
regimens in a crossover design. Compared to single exposure,
repeated exposure resulted in an initial progression followed by
an attenuation of decrements in FEVj, FVC and specific
airways resistance by day 4. Bronchial and BAL washings
showed decreases in the numbers of PMNs and fibronectin
levels and IL-6 was decreased in BAL fluid on day 4.
Holz et al,
(2002)






Frank et al.
(2001)






Jorres et al.
(2000)







Christian et al.
(1998)






-------
to
o
o
            Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
Ozone Concentration11
ppm |ig/m3 Duration
Repeated Exposure Studies (cont'd)
0.4 784 2 h/day for 5 days,
2 h either 10 or
20 days later
Activity Level
(VE)
IE
(60 L/min) at
1 5 -min intervals
Number and Gender
of Subjects
16 M;
18 to 35 years of age
Observed Effect(s)
BAL done immediately after fifth day of exposure and again
after exposure 10 or 20 days later. Most markers of
inflammation (PMNs, IL-6, PGE2, fibronectin) showed
Reference
Devlin et al.
(1997)
                                                                                   complete attenuation; markers of damage (LDH, IL-8, protein,

                                                                                   al-antitrypsin, elastase) did not. Reversal of attenuation was

                                                                                   not complete for some markers, even after 20 days.
X
0.40 784 2 h
0.60


IE
(83 W for women,
1 00 W for men)
at 1 5-min intervals
7M, 3F
23 to 41 years of age


BAL fluid 3 h after exposure had significant increases
in PMNs, PGE2, TXB2, and PGF2ct at both O3 concentrations.


Seltzer et al.
(1986)


        * See Appendix A for abbreviations and acronyms.

        b Listed from lowest to highest O3 concentration.
H

b
o


o
H

O

O
H
W

O


O
HH
H
W

-------
         Table AX6-13. Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to Controlled

                                                    Human Exposure to Ozone"
to
o
o
X
O
to
H

6
o


o
H

O
Ozone Concentration13
ppm |ig/m3 Duration

Activity Level Number and
(VE) Gender of Subjects

Observed Effect(s)

Reference
Host Defense
0.2 392 2 h




0.3 588 6 h/day for
5
consecutive
days


0.2 382 2 h






IE(15min/30 4M, 5F
min); mild atopic
(VE) ~ 20 asthmatics;
L/min/m2 BSA 2 1 -42 years of age

IE (light treadmill) 24 M
(12 O3, 12 air)




IE 10M,2F
(15 min/30 min); healthy NS
( VE) ~ 30 L/min mean ~ 28 years
of age



A significant decline in FEV! and VC immediately following
exposure. A 2-fold increase in percent PMNs, with no changes in
other biomarkers, was observed at 6 h post exposure. By 24 h post-
exposure, PMNs had decreased, but albumin, total protein,
myeloperoxidase and eosinophil cationic protein had increased.
Subjects inoculated with type 39 rhinovirus prior to exposure.
NL was performed on the morning of Days 1 to 5, 8, 15, and 30.
No difference in virus liters in NL fluid of air and O3-exposed
subjects at any time tested. No difference in PMNs or interferon
gamma in NL fluid, or in blood lymphocyte proliferative response
to viral antigen.
Subjects were exposed to O3 and FA in a cross-over design and
underwent BAL 6 h post-exposure. O3 exposure induced a 3-fold
increase in % PMNs and epithelial cells, and increased IL-8, Gro-a,
and total protein in BAL fluid. % PMNs correlated positively with
chemokine levels. Exposure also resulted in a significant decrease in
the CD4+/CD8+ ratio and the % of activated CD4+ and CD8+ T
cells in BAL fluid.
Newson et al.
(2000)



Henderson et al.
(1988)




Krishna et al.
(1998)





Host Defense - Mucous Clearance
0.4 784 1 h


0.20 392 2 h
0.40 784



CE (40 L/min) 1 5 healthy NS
18 to 35 years old

IE (light treadmill) 7 M,
27.2 ± 6.0 (SD)
years old


Subjects inhaled radiolabeled iron oxide particles 2 h after exposure.
No significant O3 -induced effect on clearance of particles during the
next 3 h or the following morning.
Subjects inhaled radiolabeled iron oxide particles immediately
before exposure. Concentration-dependent increase in rate of
particle clearance 2 h after exposure, although clearance was
confined primarily to the peripheral airways at the lower O3
concentration.
Gerrity et al.
(1993)

Foster et al.
(1987)



o
HH
H
W

-------
to
o
o
             Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
                                                             Controlled Human Exposure to Ozone"
 Ozone Concentration11
	   Exposure                        Number and Gender
  ppm      Hg/m3      Duration     Activity Level (VE)       of Subjects
                                                                                                        Observed Effect(s)
    Reference
         Host Defense - Epithelial Permeability
0.15
0.35


0.5


0.4


294 130 min IE
686 10 exercise/
10 rest
(VE) = 8xFVC
784 2.25 h IE
(70 L/min) at
15-min intervals
784 2 h IE
(70 L/min) at
15-min intervals
8M,1F
NS


16 M,
20 to 30 years old

8M,
20 to 30 years old

Subjects inhaled 99mTc-DTPA 19 h after exposure to O3. Clearance was
increased in the lung periphery. Clearance was not related to spirometry.


Similar design and results as earlier study (Kehrl et al., 1987). For the
combined studies the average rate of clearance was 60% faster in O3-
exposed subjects.
Subjects inhaled 99mTc-DTPA 75 min after exposure. Significantly
increased clearance of 99mTc-DTPA from the lung in O3-exposed
subjects. Subjects had expected changes in FVC and SRaw.
Foster and
Stetkiewicz
(1996)

Kehrl etal. (1989)


Kehrl etal. (1987)


         Drug Effects on Inflammation
X
I
o
H
6
o
o
H
O
o
HH
H
W
            0.4
                      784
                                   2h
                                                      23 healthy adults      Subjects were exposed to O3 following random selection for a 2 wk daily
                                                                          regimen of antioxidants, including vegetable juice high in the carotenoid,
                                                                          lycopene, or placebo. Concentrations of lycopene in the lungs of
                                                                          supplemented subjects increased by 12% following treatment.
                                                                          Supplemented subjects showed a 20% decrease in epithelial cell DNA
                                                                          damage as assessed by the Comet Assay. Effects attributable to lycopene
                                                                          could not be separated from those of other antioxidants.
Arab etal. (2002)
0.0 0 2 h IE 4 M, 5 F
0.4 784 20 min
mild-mod.
exercise,
10 min rest
0.2 392 2h IE
All exposures (15 min/30 min);
separated by (VE) ~ 20 L/min/m2
at least 2 wks BSA
(mean ~ 30d)
Healthy NS
30 ± 3 years old
Healthy (6 M, 9 F)
and mild asthmatic
(9 M, 6 F) subjects
Subjects previously in Nightingale et al. (2000) study. Placebo-control:
Immediately postexposure decrements in FVC (9%) and FEVj (14%)
relative to pre-exposure values. FEVj decrement only 9% at 1 hr
postexposure. By 3 h postexposure, recovery in FVC to 97% and FEVj
to 98% of preexposure values. Significant increases in 8-isoprostane at 4
h postexposure. Budesonide for 2 wk prior to exposure did not affect
responses.
Comparison was made of responses in healthy subjects, who had higher
basal ascorbate (ASC) levels and lower glutathione disulfide (GSSG)
levels than those of asthmatics. 6 h after exposure, ASC levels were
decreased and GSSG levels were increased in BAL fluid of normals, but
not asthmatics. Despite these differences in basal antioxidant levels and
response to O3, decrements in FEVj and neutrophil influx did not differ
in the two subject groups.
Montuschi et al.
(2002)
Mud way et al.
(2001)

-------
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozone"
to
o
o
X

 I

o
H

6
o


o
H

O
o
HH
H
W
Oz one Concentration11
ppm
ug/m3 Duration
Activity Level ( VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Drug Effects on Inflammation (cont'd)
0.4





0.27




0.4





0.4





0.0
0.4









784 2h





529 2h
All exposures
separated by
at least 1 wk
(mean ~ 14 d)
784 2h





784 2h





784 2 h IE
4x15 min at
VE=18
L/min/m2
BSA

2 exposures:
25% subjects
exposed to
air-air, 75%
to O3-O3
IE
15-min intervals;
VE ~ 20 L/min/m2
BSA


Continuous
exercise;
(VE) ~ 25 L/min/m2
BSA

IE
15-min intervals
VE min ~ 30 L/min



IE
(20 min/30 min);
workload @
50 watts


21 °C
40% RH









Placebo group
15 M, 1 F
Antioxidant group
13M,2F
Mean age 27 years

7 M, F subjects with
mild asthma; 20-50
years of age


5 M, 4 F healthy
6 M, 7 F asthmatics




6M, 9 F healthy NS
mean ~ 3 1 years of
age



Weak responders
7 M, 13F

Strong responders
21 M, 21 F

Healthy NS
20 to 59 years old



All subjects were exposed to FA and then entered a 2 wk regimen
of placebo or 250 mg Vit C, 50IU a-tocopherol, and 12 oz veg.
cocktail/day prior to O3 exposure. O3-induced decrements in FEVj and
FVC were 30% and 24% less, respectively, in supplemented subjects.
Percent neutrophils and IL-6 levels in BAL fluid obtained 1 h post
exposure were not different in the two treatment groups.
Subjects were randomly exposed to FA and to O3 before and after 4 wks
of treatment with 400 ug budesonide, b.i.d. Budesonide did not inhibit
the decrement in FEV; or increase in symptom scores, but significantly
reduced the increase in % neutrophils and IL-8 in sputum induced 6 h
post-exposure.
Subjects were pretreated for 3 days prior to exposure with indomethacin
(75 mg/day) or placebo. Similar reductions in FEV; and FVC were seen
in both groups following placebo, whereas mid-flows showed greater
decline in asthmatics than normals. Indomethacin attenuated decrements
in FEVj and FVC in normals, but not asthamtics. Attenuation of
decrements was seen for FEF60p in asthmatics and for FEF50 in normals.
Subjects were randomly exposed to FA and to O3 before and after 2 wks
of treatment with 800 ug budesonide, b.i.d. O3 caused significant
decrements in FEVj and FVC immediately following exposure, and a
small increase in Mch-reactivity and increases in neutrophils and
myeloperoxidase in sputum induced at 4 h post-exposure. No differences
were detected between responses in the two treatment groups.
Significant Cyinduced decrements in spirometric lung function.
Young adults (< 35 years) were significantly more responsive than older
individuals (> 35 years). Sufentanil, a narcotic analgesic, largely
abolished symptom responses and improved FEVj in strong responders.
Naloxone, an opioid antagonist, did not affect O3 effects in weak
responders. See Section AX6.2. 5.1





Sametetal. (2001)
Stech-Scott
et al. (2004)



Vagaggini, et al.
(2001)



Alexis et al. (2000)





Nightingale etal.
(2000)




Passannante etal.
(1998)










-------
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozone"
to
o
o
X

 I

o
H

6
o


o
H

O
o
HH
H
W
Ozone Concentration11
ppm Hg/m3 Duration

Activity Level ( VE)

Number and Gender
of Subjects

Observed Effect(s)

Reference
Drug Effects on Inflammation (cont'd)
0.4 784 2 h





0.4 784 2 h



0.35 686 1 h



IE
(60 L/min) at
15-min intervals



IE
(15 min/ 30 min);
( VE) = 30 L/min/m2
BSA
Continuous
exercise;
(VE) = 60 L/min

10 M





1 3 healthy male
subjects


14 healthy college-
age males


Subjects given 800 mg ibuprofen or placebo 90 min before exposure.
Subjects given ibuprofen had less of a decrease in FEVj after O3
exposure. BAL fluid 1 h after exposure contained similar levels of
PMNs, protein, fibronectin, LDH, a-1 antitrypsin, LTB4, and C3a in both
ibuprofen and placebo groups. However, subjects given ibuprofen had
decreased levels of IL-6, TXB2, and PGE2.
Four days prior to O3 exposure, subjects received either no treatment,
placebo or 150 mg indomethacin/day. Indomethacin treatment
attenuated the O3-induced decrease in FEVj, but had no effect on the O3-
induced increase in Mch responsiveness.
In a placebo- and air-controlled random design, subjects were treated
with 75 mg indomethacin every 12 h for 5 days prior to exposure.
Indomethacin significantly reduced O3-induced decrements in FEVj and
FVC.
Hazuchaetal.
(1996)




Yingetal. (1990)



Schelegle et al.
(1987)


Supportive In Vitro Studies
0.01 to 19.6 to 6h
0.10 196



0.1 196 24 h



0.2 392 3 h


1 1,690 4h



bronchial
epithelial cells



Nasal mucosa



Nasal epithelial
cells and airway
epithelial cell line
Macrophage-like
THP-1 cells


Nonatopic,
nonasthmatic and
atopic, mild
asthmatic bronchial
biopsy samples
Allergic and
nonallergic patients









Exposure to 0.01-0.10 ppm O3 significantly decreased the electrical
resistance of cells from asthmatic sources, compared to nonasthmatic
sources. This range of O3 concentrations also increased the
movement of 14C-BSA across the confluent cultures of "asthmatic"
cells to an extent that was greater than that in "nonasthmatic" cells.
Increased concentrations of neurokinin A and substance P in
medium following O3 exposure. Levels of release of both
neuropeptides were higher from tissues derived from allergic
compared to nonallergic patients.
Synergistic effect of O3 exposure on rhino virus- induced release
of IL-8 at 24 h through mechanisms abrogated by antioxidant
pretreatment. Additive enhancement of 1C AM- 1 expression.
THP-1 cells were treated with samples of human surfactant protein
A (SP-A) genetic variants (SP-A1 and SP-A2) that had been
previously exposed to O3. O3-exposed variants differed in then-
ability to stimulate the production of TNFa and IL-8 by these cells.
Bayram et al.
(2002)



Schierhorn et al.
(2002)


Spannhake et al.
(2002)

Wang et al.
(2002)



-------
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozone"
to
o
o
X
Oi
O
Oi
H

6
o


o
H

O
Ozone Concentration5
ppm
ug/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects Observed Effect(s)
Reference
Supportive In Vitro Studies (cont'd)
0.01 to
0.10




0.12
0.24
0.50

0.06 to
0.20



0.5





0.5




0.4




19.6 to 6h
196




235 3h
470
980

118 to 24 h
392



980 Ih





980 Ih




784 Ih




bronchial
epithelial cells




Nasal epithelial
cells


Nasal mucosa




Lung fibroblast
cell line with an
airway epithelial
cell line


tracheal epithelial
cells



Lung fibroblasts;
airway epithelial
cell line


Nonatopic, No difference in constitutive release of IL-8, GM-CSF, si CAM- 1
nonasthmatic and and RANTES from cells from nonasthmatic and asthmatic sources,
atopic, mild except for detection of RANTES in latter cells only. Increased
asthmatic bronchial release of all mediators 24 h after 0.05 to 0.10 ppm O3 in
biopsy samples "asthmatic" cells, but only IL-8 and si CAM- 1 in "nonasthmatic"
cells.
Small dose-response activation of NF-KB coinciding with
O3-induced production of free radicals assessed by electron spin
resonance. Increased TNFa at two higher concentrations of O3 at
16 h post-expo sure.
105 surgical Increased histamine release correlated with mast cell degranulation.
samples from Increased release of IL-1, IL-6, IL-8 and TNFa following O3
atopic and exposure at 0. 10 ppm. Release of IL-4, IL-6, IL-8 and TNFa at
nonatopic patients this concentration was significantly greater from tissues from atopic
versus nonatopic patients.
BEAS-2B cells in the presence or absence of HFL-1 cells were
exposed and incubated for 11 or 23 h. Steady-state mRNA levels
of alpha 1 procollagens type I and II, as well as TGF(31, were
increased in O3-exposed co-cultured fibroblasts compared to air
controls. Data support interactions between the cell types in the
presence and the absence of O3 -expo sure.
O3 exposure caused an increase in ROS formation and a decline in
PGE2 production. No differences in mRNA and protein levels of
prostaglandin endoperoxide G/H synthase 2 (PGHS-2) or the rate
of its synthesis were detected, suggesting a direct effect of O3-
generated oxidants on PGHS-2 activity.
Cells incubated with O3-exposed arachidonic acid (AA) were found
to contain DNA single strand breaks. Pretreatment of the exposed
AA solution with catalase eliminated the effect on DNA, indicating
its dependence on H2O2 production. The effect was potentiated by
the non-carbony 1 component of ozonized AA.
Bayram et al.
(2001)




Nichols et al.
(2001)


Schierhorn et al.
(1999)



Langetal. (1998)





Alpert et al.
(1997)



Kozumbo et al.
(1996)



o
HH
H
W

-------
3
Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to
                                    Controlled Human Exposure to Ozone"
<<
to
o
o




X
Oi
i
O
O
H
6
o
0
H
O
O
0
o
H
W
Ozone Concentration13
ppm |ig/m3 Duration
Supportive In Vitro Studies (cont'd)
0.25 490 6 h
0.50 980
0.25 490 1 h
0.50 980
1.00 1,960
0.20 to 392 to 2hor4h
1.0 1960

0.25 490 1 h
0.50 980
1.00 1,960
0.30 588 1 h
1.00 1,960

Activity Level
(VE)

Human nasal
epithelial cells
Airway epithelial
cell line and
alveolar
macrophages
Airway epithelial
cell line

Airway epithelial
cell line
Alveolar
macrophages

Number and
Gender of Subjects Observed Effect(s)

Increased in ICAM-1, IL-6, IL-1, and TNF expression at 0.5 ppm.
No increase in IL-8 expression. No increases at 0.25 ppm.
Increased secretion of IL-6, IL-8, and fibronectin by epithelial cells,
even at lowest O3 concentration. No O3-induced secretion of these
compounds by macrophages.
O3 caused a dose-related loss in cellular replicative activity at
exposure levels that caused minimal cytotoxicity. DNAsingle
strand breaks were not detected. These effects were different from
those of H2O2 and, thus, not likely related to production of this
oxidant within the cells.
Concentration-dependent increased secretion of PGE2, TXB2,
PGF2ct, LTB4, and LTD4. More secretion basolaterally than apically.
Concentration-dependent increases in PGE2 production, and
decreases in phagocytosis of sheep erythrocytes. No O3-induced
secretion of IL-1, TNF, or IL-6.

Reference

Beck etal. (1994)
Devlin et al.
(1994)
Gabrielson et al.
(1994).

McKinnon et al.
(1993)
Becker et al.
(1991)

"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.

























-------
 1      studied include those involved with fibrin deposition and degradation (Tissue Factor, Factor VII,
 2      and plasminogen activator), potential markers of fibrogenesis (fibronectin, platelet derived
 3      growth factor), and components of the complement cascade (C3a).
 4           Inflammatory cells of the lung such as alveolar macrophages (AMs), monocytes, and
 5      PMNs also constitute an important component of the pulmonary host defense system.  Upon
 6      activation, they are capable of generating free radicals and enzymes with microbicidal
 7      capabilities, but they also have the potential to damage nearby cells.  More recently published
 8      studies since the last literature review (U.S. Environmental Protection Agency, 1996) observed
 9      changes in T lymphocyte subsets in the airways following exposure to O3 that suggest
10      components of the immune host defense also may be affected.
11
12      AX6.9.2  Inflammatory Responses in the  Upper Respiratory Tract
13           The nasal passages constitute the primary portal for inspired air at rest and, therefore, the
14      first region of the respiratory tract to come in contact with airborne pollutants. Nikasinovic et al.
15      (2003) recently reviewed the literature of laboratory-based nasal inflammatory studies  published
16      since 1985.  Nasal lavage (NL) has provided a useful tool for assessing O3-induced inflammation
17      in the nasopharynx.  Nasal lavage is simple and rapid to perform, is noninvasive, and allows
18      collection of multiple sequential samples.  Graham  et al. (1988) reported increased levels of
19      PMNs in the NL fluid of humans exposed to 0.5 ppm O3 at rest for 4 h on 2 consecutive days,
20      with NL performed immediately before and after each exposure, as well as 22 h after the second
21      exposure. Nasal lavage fluid contained elevated numbers of PMNs at all postexposure times
22      tested, with peak values occurring immediately prior to the second day of exposure.  Bascom
23      et al.  (1990) exposed subjects with allergic rhinitis to 0.5 ppm O3 at rest for 4 h, and found
24      increases in PMNs, eosinophils, and mononuclear cells following O3 exposure.  Graham and
25      Koren (1990) compared inflammatory mediators present in both the NL and BAL fluids of
26      humans exposed to 0.4 ppm O3 for 2 h. Increases in NL and BAL PMNs were similar  (6.6- and
27      eightfold, respectively), suggesting a qualitative correlation between inflammatory changes in
28      the lower airways (BAL)  and the upper respiratory tract (NL), although the PMN increase in NL
29      could not quantitatively predict the PMN increase in BAL.  Torres et al. (1997) compared NL
30      and BAL in smokers and nonsmokers exposed to 0.22 ppm O3 for 4 h.  In contrast to Graham
31      and Koren (1990), they did not find a relationship between numbers or percentages of

        January 2005                           AX6-108     DRAFT-DO NOT QUOTE OR CITE

-------
 1      inflammatory cells (PMNs) in the nose and the lung, perhaps in part due to the variability
 2      observed in their NL recoveries. Albumin, a marker of epithelial cell permeability, was
 3      increased 18 h later, but not immediately after exposure, as seen by Bascom et al. (1990).
 4      Tryptase, a constituent of mast cells, was also elevated after O3 exposure at 0.4 ppm for 2 h
 5      (Koren et al., 1990). McBride et al. (1994) reported that asthmatic subjects were more sensitive
 6      than non-asthmatics to upper airway inflammation at an O3 concentration (0.24 ppm (1.5 h)) that
 7      did not affect lung or nasal function or biochemical mediators.  A significant increase in the
 8      number of PMNs in NL fluid was detected in the asthmatic subjects both immediately and 24 h
 9      after exposure. Peden et al. (1995) also found that O3 at a concentration of 0.4 ppm had a direct
10      nasal inflammatory effect, and reported a priming effect on the response to nasal allergen
11      challenge, as well.  A subsequent study in  dust mite-sensitive asthmatic subjects indicated that
12      O3 at this concentration enhanced eosinophil influx in response to allergen, but did not promote
13      early mediator release or enhance the nasal response to allergen (Michelson et al., 1999).
14      Similar to observations made in the lower  airways, the presence of O3 molecular "targets" in
15      nasal lining fluid is likely to provide some level of local protection against exposure.  In a study
16      of healthy subjects exposed to 0.2 ppm O3  for 2 h, Mudway and colleagues (1999) observed a
17      significant depletion of uric acid in NL fluid at 1.5 h following exposure.
18           An increasing number of studies have taken advantage of advances in cell and tissue
19      culture techniques to examine the role of upper and lower airway epithelial cells and mucosa in
20      transducing the effects of O3 exposure. Many of these studies have provided important insight
21      into the basis of observations made in vivo. One of the methods used enables the cells or tissue
22      samples to be cultured at the air-liquid interface (ALI), allowing cells to establish apical and
23      basal polarity, and both cells and tissue samples to undergo exposure to O3 at the apical surfaces
24      as would occur in vivo. Nichols and colleagues (2001) examined human nasal epithelial cells
25      grown at the ALI for changes in free radical production, based on electron spin resonance, and
26      activation of the NF-KB transcription factor following exposure to O3 at 0.12 to  0.5 ppm for 3 h.
27      They found a dose-related activation of NF-KB within the cells that coincided with O3-induced
28      free radical production and increased release of TNFa at levels above 0.24 ppm. These data
29      confirm the importance of this oxidant stress-associated pathway in transducing the O3 signal
30      within nasal  epithelial cells, and suggest its role in directing the inflammatory response.  In a
31      study of nasal mucosal biopsy plugs, Schierhorn, et al. (1999) found that tissues exposed to O3 at

        January 2005                           AX6-109     DRAFT-DO NOT QUOTE OR CITE

-------
 1      a concentration of 0.1 ppm induced release of IL-4, IL-6, IL-8, and TNFa that was significantly
 2      greater from tissues from atopic patients compared to nonatopic controls. In a subsequent study,
 3      this same exposure regimen caused the release of significantly greater amounts of the
 4      neuropeptides, neurokinin A and substance P, from allergic patients, compared to nonallergic
 5      controls, suggesting increased activation of sensory nerves by O3 in the allergic tissues
 6      (Schierhorn et al., 2002).
 7
 8      AX6.9.3  Inflammatory Responses in the Lower Respiratory Tract
 9           Seltzer et al. (1986) were the first to demonstrate that exposure of humans to O3 resulted in
10      inflammation in the lung. Bronchoalveolar lavage fluid (3 h post-exposure) from subjects
11      exposed to O3 contained increased PMNs as well as increased levels of PGE2, PGF2a, and TXB2
12      compared to fluid from air-exposed subjects. Koren et al. (1989a,b) described inflammatory
13      changes 18 h after O3 exposure. In addition to an eightfold increase in PMNs, Koren et al.
14      reported a two-fold increase in  BAL fluid protein, albumin, and immunoglobulin G (IgG) levels,
15      suggestive of increased epithelial  cell permeability. There was a 12-fold increase in IL-6 levels,
16      a two-fold increase in PGE2, and a two-fold increase in the complement component, C3a.
17      Evidence for stimulation of fibrogenic processes in the lung was shown by significant increases
18      in coagulation components, Tissue Factor and Factor VII (McGee et al., 1990), urokinase
19      plasminogen activator and fibronectin (Koren et al., 1989a).  Subsequent studies by Lang, et al.
20      (1998), using co-cultures of cells of the BEAS-2B bronchial epithelial line and of the HFL-1
21      lung fibroblast line, provided additional information about O3-induced fibrogenic processes.
22      They demonstrated that steady-state mRNA levels of both alpha 1 and procollagens type I and
23      III in the fibroblasts were increased following O3 exposure and that this effect was mediated by
24      the O3-exposed epithelial cells. This group of studies demonstrated that exposure to O3 results in
25      an inflammatory reaction in the lung, as evidenced by increases in PMNs and proinflammatory
26      compounds. Furthermore, they demonstrated that cells and mediators capable of damaging
27      pulmonary tissue are increased after O3 exposure, and provided early suggestion of the potential
28      importance of the epithelial cell-myofibroblast "axis" in modulating fibrotic and fibrinolytic
29      processes in the airways.
30           Isolated lavage of the mainstream bronchus using balloon catheters or BAL using small
31      volumes of saline have been used to assess O3-induced changes in the large airways. Studies

        January 2005                           AX6-110     DRAFT-DO NOT QUOTE OR CITE

-------
 1      collecting lavage fluid from isolated airway segments after O3 exposure indicate increased
 2      neutrophils in the airways (Aris et al., 1993; Balmes et al., 1996; Scannell et al., 1996).  Other
 3      evidence of airway neutrophil increase comes from studies in which the initial lavage fraction
 4      ("bronchial fraction") showed increased levels of neutrophils (Schelegle et al., 1991; Peden
 5      et al., 1997; Balmes et al., 1996; Torres et al., 1997).  Bronchial biopsies show increased PMNs
 6      in airway tissue (Aris et al.,1993) and, in sputum collected after O3 exposure, neutrophil numbers
 7      are elevated (Fahy et al., 1995).
 8           Increased BAL protein, suggesting O3-induced changes in epithelial permeability (Koren
 9      et al., 1989a, 1991 and Devlin et al., 1991) supports earlier work in which increased epithelial
10      permeability, as measured by increased clearance of radiolabled diethylene triamine pentaacetic
11      acid (99mTc-DTPA) from the lungs of humans exposed to O3, was demonstrated (Kehrl et al.,
12      1987).  In addition, Foster and Stetkiewicz (1996) have shown that increased permeability
13      persists for at least 18-20 h and the effect is greater at the lung apices than at the base. In a study
14      of mild atopic asthmatics exposed to 0.2 ppm O3 for 2 h,  Newson, et al. (2000) observed a 2-fold
15      increase in the percentage of PMNs present at 6 hours post exposure, with no change in  markers
16      of increased permeability as assessed by sputum induction. By 24 h, the neutrophilia was seen
17      to subside while levels of albumin, total protein, myeloperoxidase, and eosinophil cationic
18      protein increased significantly.  It was concluded that the transient PMN influx induced  by acute
19      exposure of these asthmatic subjects was followed by plasma extravasation and the activation of
20      both PMNs and eosinophils within the airway tissues.  Such changes in permeability associated
21      with acute inflammation may provide better access  of inhaled antigens, particulates, and other
22      substances to the submucosal region.
23           Devlin et al. (1991) reported an inflammatory response in subjects exposed to 0.08 and
24      0.10 ppm O3 for 6.6 h. Increased numbers of PMNs and  levels of IL-6 were found at both
25      O3 concentrations, suggesting that lung inflammation  from O3 can occur as a consequence of
26      prolonged exposure to ambient levels while exercising.  Interestingly, those individuals who had
27      the largest increases in inflammatory mediators in this study did not necessarily have the largest
28      decrements in pulmonary function, suggesting that separate mechanisms underlie these two
29      responses. The absence of a relationship between spirometric responses and inflammatory cells
30      and markers has been reported in several studies, including Balmes et al., 1996; Schelegle et al.,
31      1991; Torres et al., 1997; Hazucha et al., 1996; Blomberg et al., 1999. These observations relate

        January 2005                            AX6-111      DRAFT-DO NOT QUOTE OR CITE

-------
 1      largely to disparities in the times of onset and duration following single exposures. The
 2      relationship between inflammatory and residual functional responses following repeated or
 3      chronic exposures may represent a somewhat different case (see Section AX6.9.4).
 4           As indicated above, a variety of potent proinflammatory mediators have been reported to
 5      be released into the airway lumen following O3 exposure. Studies of human alveolar
 6      macrophages (AM) and airway epithelial cells exposed to O3 in vitro suggest that most mediators
 7      found in the BAL fluid of O3-exposed humans are produced by  epithelial cells.  Macrophages
 8      exposed to O3 in vitro showed only small increases in PGE2  (Becker et al., 1991). In contrast,
 9      airway epithelial cells exposed in vitro to O3 showed large concentration-dependent increases in
10      PGE2, TXB2, LTB4, LTC4, and LTD4 (McKinnon et al., 1993) and increases in IL-6, IL-8, and
11      fibronectin at O3 concentrations as low as 0.1 ppm (Devlin et al., 1994).  Macrophages lavaged
12      from subjects exposed to 0.4 ppm (Koren et al., 1989a) showed changes in the rate of synthesis
13      of 123 different proteins, whereas AMs exposed to O3 in vitro showed changes in only six
14      proteins, suggesting that macrophage function was altered by mediators released from other
15      cells. Furthermore, recent evidence suggests that the release of mediators from AMs may be
16      modulated by the products of O3-induced oxidation of airway lining fluid components, such as
17      human surfactant protein A (Wang et al., 2002).
18           Although the release of mediators has been demonstrated to occur at exposure
19      concentrations and times that are minimally cytotoxic to airway cells, potentially detrimental
20      latent effects have been demonstrated in the absence of cytotoxicity. These include the
21      generation of DNA single strand breaks (Kozumbo et al., 1996) and the loss of cellular
22      replicative activity (Gabrielson et al.,  1994) in bronchial epithelial cells exposed in vitro, and the
23      formation of protein and DNA adducts.  A highly toxic aldehyde formed during O3-induced lipid
24      peroxidation is 4-hydroxynonenal  (HNE).  Healthy human subjects exposed to 0.4 ppm O3 for 1
25      h underwent BAL 6 h later. Analysis  of lavaged alveolar macrophages by Western blot
26      indicated increased levels of a 32-kDa HNE-protein adduct,  as well as 72-kDa heat shock protein
27      and ferritin, in O3- versus air-exposed subjects (Hamilton et  al., 1998). In a recent study of
28      healthy subjects exposed to 0.1 ppm O3 for 2 h (Corradi et al., 2002), formation of 8-hydroxy-2'-
29      deoxyguanosine (8-OHdG), a biomarker of reactive oxidant species (ROS)-DNA interaction,
30      was measured in peripheral blood lymphocytes. At 18 h post exposure, 8-OHdG was
31      significantly increased in cells compared to pre-exposure levels, presumably linked to concurrent

        January 2005                            AX6-112     DRAFT-DO NOT QUOTE OR CITE

-------
 1      increases in chemical markers of ROS.  Of interest, the increase in 8-OHdG was only significant
 2      in a subgroup of subjects with the wild genotype for NAD(P)H:quinone oxidoreductase and the
 3      null genotype for glutathione-S-transferase Ml, suggesting that polymorphisms in redox
 4      enzymes may confer "susceptibility' to O3 in some individuals.  The generation of ROS
 5      following exposure to O3 has been shown to be associated with a wide range of responses. In a
 6      recent study, ROS production by alveolar macrophages lavaged from subjects exposed to
 7      0.22 ppm for 4 h was assessed by flow cytometry (Voter et al., 2001).  Levels were found to be
 8      significantly elevated 18 h post exposure and associated with several markers of increased
 9      permeability.  An in vitro study of human tracheal epithelial  cells exposed to O3 indicated that
10      generation of ROS resulted in decrease in synthesis of the bronchodilatory prostaglandin, PGE2,
11      as a result of inactivation of prostaglandin endoperoxide G/H synthase 2 (Alpert et al.,  1997).
12      These and similar studies indicate that the responses to products of O3 exposure in the airways
13      encompass a broad range of both stimulatory and inhibitory activities, many of which may be
14      modulated by susceptibility factors upstream in the exposure process, at the level of
15      compensating for the imposed oxidant stress.
16           The inflammatory responses to O3 exposure also have been studied in asthmatic subjects
17      (Basha et al., 1994; Scannell et al., 1996; Peden et al., 1997). In these studies, asthmatics
18      showed significantly more neutrophils in the BAL (18 h post-exposure) than similarly exposed
19      healthy individuals.  In one of these studies (Peden et al., 1997), which included only allergic
20      asthmatics who tested positive for Dematophagoides farinae  antigen, there was an eosinophilic
21      inflammation (2-fold increase), as well as neutrophilic inflammation (3-fold increase).  In a
22      study of subjects with intermittent asthma that utilized a 2-fold higher concentration of O3 (0.4
23      ppm) for 2 h, increases in eosinophil cationic protein, neutrophil elastase and IL-8 were found to
24      be significantly increased 16 h post-exposure and comparable in induced sputum and BAL fluid
25      (Hiltermann et al,  1999). In two  studies (Basha et al., 1994;  Scannell et al.,  1996), IL-8 was
26      significantly higher in post-O3 exposure BAL in asthmatics compared to non-asthmatics,
27      suggesting a possible mediator for the increased neutrophilic inflammation in those subjects.
28      In a recent study comparing the neutrophil response to O3 at  a concentration and exposure time
29      similar to those of the latter three studies, Stenfors and colleagues (2002) were unable to detect a
30      difference in the increased neutrophil numbers between 15 mild asthmatic and 15 healthy
31      subjects by bronchial wash at the 6 h post-exposure time point.  These results suggest that, at

        January 2005                           AX6-113     DRAFT-DO NOT QUOTE OR CITE

-------
 1      least with regard to neutrophil influx, differences between healthy and asthmatic individuals
 2      develop gradually following exposure and may not become evident until later in the process.
 3      In another study, mild asthmatics who exhibited a late phase underwent allergen challenge 24 hrs
 4      before a 2  h exposure to 0.27 ppm O3 or filtered air in a cross-over design (Vagaggini et al.,
 5      2002).  At 6 h post-exposure, eosinophil numbers in induced sputum were found to be
 6      significantly greater after O3 than after air. Studies such as these suggest that the time course of
 7      eosinophil and neutrophil influx following O3 exposure can occur to levels detectable within the
 8      airway  lumen by as early as 6 h. They also suggest that the previous or concurrent activation of
 9      proinflammatory pathways within the airway epithelium may enhance the inflammatory effects
10      of O3. For example, in an in vitro study of epithelial cells from the upper and lower respiratory
11      tract, cytokine production induced by rhinovirus infection was enhanced synergistically by
12      concurrent exposure to O3 at 0.2 ppm for 3 h (Spannhake et al, 2002).  The use of bronchial
13      mucosal biopsies has  also provided important insight into the modulation by O3 of existing
14      inflammatory processes within asthmatics. In a study of healthy and allergic asthmatic subjects
15      exposed to 0.2 ppm O3 or filtered air for 2 h, biopsies were performed 6 hr following exposure
16      (Bosson et al., 2003).  Monoclonal antibodies were used to assess epithelial expression of a
17      variety  of cytokines and chemokines.  At baseline (air exposure), asthmatic subjects showed
18      significantly higher expression of interleukins (IL)-4 and -5. Following O3 exposure, the
19      epithelial expression of IL-5, IL-8, granulocyte-macrophage colony-stimulating factor (GM-
20      CSF) and epithelial cell-derived neutrophil-activating peptide 78 (ENA-78) was significantly
21      greater  in asthmatic subjects, as compared to healthy subjects. In vitro studies of bronchial
22      epithelial cells derived by biopsy from nonatopic, nonasthmatic subjects and asthmatic subjects
23      also demonstrated the preferential release of GM-CSF and also of regulated on activation,
24      normal  T cell-expressed and -secreted (RANTES) from asthmatic cells following O3 exposure.
25           The time course of the inflammatory response to O3 in humans has not been explored fully.
26      Nevertheless,  studies  in which BAL was performed 1-3 h (Devlin et al., 1990; Koren et al.,
27      1991; Seltzer et al., 1986) after exposure to 0.4 ppm O3 demonstrated that the inflammatory
28      response is quickly initiated, and other studies (Koren et al.,  1989a,b; Torres et al., 1997;
29      Scannell et al., 1996;  Balmes et al., 1996) indicated that,  even 18 h after exposure, inflammatory
30      mediators  such as  IL-6 and PMNs were still elevated. However, different markers show peak
31      responses at different times. Ozone-induced increases in IL-8, IL-6, and PGE2 are greater

        January 2005                            AX6-114     DRAFT-DO NOT QUOTE OR CITE

-------
 1     immediately after O3 exposure, whereas BAL levels of fibronectin and plasminogen activator are
 2     greater after 18 h. PMNs and some products (protein, Tissue Factor) are similarly elevated both
 3     1 and 18 h after O3 exposure (Devlin et al., 1996; Torres et al., 1997). Schelegle et al. (1991)
 4     found increased PMNs in the "proximal airway" lavage at 1, 6, and 24 h after O3 exposure, with
 5     a peak response at 6 h. In a typical BAL sample, PMNs were elevated only at the later time
 6     points.  This is consistent with the greater increase 18 h after exposure seen by Torres et al.
 7     (1997). In addition to the influx of PMNs and (in allergic asthmatics) eosinophils, lymphocyte
 8     numbers in BAL were also seen to be elevated significantly at 6 h following exposure of healthy
 9     subjects to 0.2 ppm O3 for 2 h (Blomberg et al., 1997). Analysis of these cells by flow cytometry
10     indicated the increased presence of CD3+, CD4+ and CD8+ T cell subsets. This same laboratory
11     later demonstrated that within 1.5 h following exposure of healthy subjects to the same O3
12     regimen, expression of human leukocyte antigen (HLA)-DR on lavaged macrophages underwent
13     a significant, 2.5-fold increase (Blomberg et al., 1999). The significance of these alterations in
14     immune system components and those in IL-4 and IL-5 expression described above in the
15     studies of Bosson et al. (2003) has not been fully explored and may suggest a role for O3 in the
16     modulation of immune inflammatory processes.
17
18     AX6.9.4  Effects of Repeated Exposures and Adaptation of Responses
19          Residents of areas with high oxidant concentrations tend to have somewhat blunted
20     pulmonary function responses and symptoms to O3 exposure (Hackney et al.,  1976, 1977b, 1989;
21     Avol et al., 1988; Linn et al., 1988). Animal studies suggest that while inflammation may be
22     diminished with repeated exposure, underlying damage to lung epithelial cells continues (Tepper
23     et al., 1989). Devlin et al. (1997) examined the inflammatory responces of humans repeatedly
24     exposed to 0.4 ppm O3 for 5 consecutive days.  Several indicators of inflammation (e.g., PMN
25     influx, IL-6, PGE2, fibronectin, macrophage phagocytosis) were attenuated after 5  days of
26     exposure (i.e.,  values were not different from FA). Several  markers (LDH, IL-8, total protein,
27     epithelial cells) did not show attenuation, indicating that tissue damage probably continues to
28     occur during repeated exposure. The recovery of the inflammatory response occurred for some
29     markers after 10 days, but some responses were not normalized even after 20 days. The
30     continued presence of markers of cellular injury indicates a  persistent but not necessarily
31     perceived response to O3. Christian et al. (1998) randomly subjected heathy subjects to a single

       January 2005                           AX6-115    DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposure and to 4 consecutive days of exposure to 0.2 ppm O3 for 4 h. As reported by others,
 2      they found an attenuation of FEVl3 FVC and specific airway resistance when comparing the
 3      single exposure with day 4 of the multiday exposure regimen. Similarly, both "bronchial" and
 4      "alveolar" fractions of the BAL showed decreased numbers of PMNs and fibronectin
 5      concentration at day 4 versus the single exposure, and a decrease in IL-6 levels in the alveolar
 6      fraction. Following a similar study design and exposure parameters, but with single day filtered
 7      air controls, Torres et al. (2000) found a decrease in  FEVj and increases in the percentages of
 8      neutrophils and lymphocytes, in concentrations of total protein, IL-6, IL-8, reduced glutathione,
 9      ortho-tyrosine and urate in BAL fluid, but no changes in bronchial biopsy histology following
10      the single exposure.  Twenty hours after the day 4 exposure, both functional and BAL cellular
11      responses to O3 were abolished. However, levels of total protein, IL-6, IL-8, reduced glutathione
12      and ortho-tyrosine were still increased significantly. In addition, following the day 4 exposure,
13      visual scores for bronchitis, erythema and the numbers of neutrophils in the mucosal biopsies
14      were increased. Their results indicate that, despite reduction of some markers of inflammation
15      in BAL and measures of large airway function, inflammation within the airways persists
16      following repeated exposure to O3. In another study, Frank and colleagues (2001) exposed
17      healthy subjects to filtered air and to  O3 (0.25 ppm,  2 h) on 4 consecutive days each, with
18      pulmonary function measurements being made prior to and following each exposure. BAL was
19      performed on day 5, 24 h following the last exposure.  On day 5,  PMN numbers remained
20      significantly higher in the  O3  arm compared to air control. Of particular note in this study was
21      the observation that small  airway function, assessed by grouping values for isovolumetric
22      FEF25.75, VmaxSO and Vmax75 into a single value, showed persistent reduction from day 2
23      through day 5. These data suggest that methods to more effectively monitor function in the most
24      peripheral airway regions, which are  known to be the primary sites of O3 deposition in the lung,
25      may provide important information regarding the cumulative effects of O3 exposure.  It is
26      interesting to note that Alexis et al. (2000) reported  that, following exposure of normals and
27      asthmatics to 0.4 ppm O3 for 2 h, variables representing small airways function (viz., FEF25, FEF
28      50, FEF 60P, FEF75) demonstrated the greatest O3-induced decline in the asthmatic subjects.  Holz,
29      et al. (2002) made a comparison of early and late responses to allergen challenge following O3 in
30      subjects with allergic rhinitis  or allergic asthma. With some variation, both early and late FEVj
        January 2005                            AX6-116     DRAFT-DO NOT QUOTE OR CITE

-------
 1      and cellular responses in the two subject groups were significantly enhanced by 4 consecutive
 2      days of exposure to 0.125 ppm O3 for 3 h.
 3
 4      AX6.9.5  Effect of Anti-Inflammatory and other Mitigating Agents
 5           Studies have shown that indomethacin, a non-steroidal anti-inflammatory agent (NSAID)
 6      that inhibits the production of cyclooxygenase products of arachidonic acid metabolism, is
 7      capable of blunting the well-documented decrements in pulmonary function observed in humans
 8      exposed to O3 (Schelegle et al., 1987; Ying et al., 1990). In the latter study, indomethacin did
 9      not alter the O3-induced increase in bronchial responsiveness to methacholine.  Pretreatment of
10      healthy subjects and asthmatics with indomethacin prior to exposure to 0.4 ppm for 2 h
11      significantly attenuated decreases in FVC and FEVj in normals, but not asthmatics (Alexis et al.,
12      2000). Subjects have also been given ibuprofen, another NSAID agent that blocks
13      cyclooxygenase metabolism, prior to O3 exposure. Ibuprofen blunted decrements in lung
14      function following O3  exposure (Hazucha et al., 1996).  Subjects given ibuprofen also had
15      reduced BAL levels of the cyclooxygenase product PGE2 and thromboxane B2, as well as IL-6,
16      but no decreases were observed in PMNs, fibronectin, permeability, LDH activity, or
17      macrophage phagocytic function. These studies suggest that NSAIDs can blunt O3-induced
18      decrements in FEVj with selective (perhaps drug-specific) affects on mediator release and other
19      markers of inflammation.
20           At least two studies have looked at the effects of the inhaled corticosteroid, budesonide, on
21      the effects of O3, with differing outcome perhaps associated with the presence of preexistent
22      disease. Nightingale and colleagues  (2000) exposed healthy nonsmokers to 0.4 ppm O3 for 2 h
23      following 2 wk of treatment with budesonide (800 micrograms, twice daily) or placebo in a
24      blinded, randomized cross-over study.  This relatively high exposure resulted in significant
25      decreases in spirometric measures  and increases in methacholine reactivity  and neutrophils and
26      myeloperoxidase in induced sputum. No significant differences were observed in any of these
27      endpoints following budesonide treatment versus placebo.  In contrast, Vagaggini et al. (2001)
28      compared the effects of treatment with budesonide (400 micrograms, twice  daily) for 4 wk on
29      the responses of mild asthmatic subjects to exposure to 0.27 ppm O3for 2 h. Prior to exposure,
30      at the midpoint and end of exposure, and at 6 h post exposure, FEVj was measured and a
31      symptom questionnaire was administered;  at 6 h post exposure, sputum was induced.

        January 2005                           AX6-117     DRAFT-DO NOT QUOTE OR CITE

-------
 1      Budesonide treatment did not inhibit the decrement in FEVj or alter symptom score, but
 2      significantly blunted the increase in percent PMNs and concentration of IL-8 in the sputum.  The
 3      difference in subject health status between the two studies (healthy versus mild asthmatic) may
 4      suggest a basis for the differing outcomes; however, because of differences in the corticosteroid
 5      dosage and O3 exposure levels, that basis remains unclear.
 6           Because the O3 exerts its actions in the respiratory tract by virtue of its strong oxidant
 7      activity, it is reasonable to assume that molecules that can act as surrogate targets in the airways,
 8      as constituents of either extracellular fluids or the intracellular milieus, could abrogate the effects
 9      of O3. Some studies have examined the ability of dietary "antioxidant" supplements to reduce
10      the risk of exposure of the lung to oxidant exposure. In a study of healthy, nonsmoking adults,
11      Samet and colleagues (2001) restricted dietary ascorbate and randomly treated subjects for
12      2 weeks with a mixture of vitamin C, a-tocopherol and vegetable  cocktail high in carrot and
13      tomato juices or placebo. Responses to 0.4 ppm O3 for 2 h were assessed in both groups at the
14      end of treatment.  O3 -induced decrements in FEVj and FVC were significantly reduced in the
15      supplemented group, whereas the inflammatory response, as assessed by percentage neutrophils
16      and levels of IL-6 in BAL fluid, were unaffected by antioxidant supplementation.  In a  study that
17      focused on supplementation with a commercial vegetable cocktail high in the carotenoid,
18      lycopene, healthy subjects were exposed for 2 h to 0.4 ppm O3 after 2 wk of antioxidant
19      supplementation or placebo (Arab et al., 2002). These investigators observed that lung epithelial
20      cell DNA damage, as measured by the Comet Assay, decreased by 20 % in supplemented
21      subjects.  However, the relationships between the types  and levels of antioxidants in airway
22      lining fluid and responsiveness to O3 exposure is likely to be complex. In a study in which
23      differences in  ascorbate and glutathione concentrations between healthy and mild asthmatic
24      subjects were exploited, no relationship between antioxidant levels and spirometric or cellular
25      responses could be detected (Mudway, et al., 2001).
26
27      AX6.9.6  Changes in Host Defense Capability Following  Ozone Exposure
28           Concern about the effect of O3 on human host defense capability derives from numerous
29      animal studies demonstrating that acute exposure to as little as 0.08 ppm O3 causes decrements
30      in antibacterial host defenses (see Chapter 5). A study of experimental rhinovirus  infection in
31      susceptible human volunteers failed to  show any effect of 5 consecutive days of O3 exposure on

        January 2005                            AX6-118      DRAFT-DO NOT QUOTE OR CITE

-------
 1      the clinical evolution of, or host response to, a viral challenge (Henderson et al., 1988).  Healthy
 2      men were nasally inoculated with type 39 rhinovirus (103 TCID50). There was no difference
 3      between the O3-exposed and control groups in rhinovirus liters in nasal secretions, in levels of
 4      interferon gamma or PMNs in NL fluid, or in blood lymphocyte proliferative response to
 5      rhinovirus antigen.  However, subsequent findings that rhinovirus can attach to the intracellular
 6      adhesion molecule (ICAM)-l receptor on respiratory tract epithelial cells (Greve et al., 1989)
 7      and that O3 can up-regulate the ICAM-1 receptor on nasal epithelial cells (Beck et al., 1994)
 8      suggest that more studies are  needed to explore the possibility that prior O3 exposure can
 9      enhance rhinovirus binding to, and infection of, the nasal epithelium.
10           In a single study, human AM host defense capacity was measured in vitro in AMs removed
11      from subjects exposed to 0.08 and 0.10 ppm O3 for 6.6 h while undergoing moderate exercise.
12      Alveolar macrophages from O3-exposed subjects had significant decrements in complement-
13      receptor-mediated phagocytosis of Candida albicans (Devlin et al., 1991).  The impairment of
14      AM host defense capability could potentially result in decreased ability to phagocytose and kill
15      inhaled microorganisms in vivo. A concentration-dependent decrease in phagocytosis of AMs
16      exposed to 0.1 to 1.0 ppm O3  in vitro has also been shown Becker et al. (1991).  Although the
17      evidence is inconclusive at present, there is a concern that O3 may render humans and animals
18      more susceptible to a subsequent bacterial challenge.
19           Only two studies (Foster et al., 1987; Gerrity et al., 1993) have investigated the effect of
20      O3 exposure on mucociliary particle clearance in  humans. Foster et al. (1987) had seven healthy
21      subjects inhale radiolabeled particles (5 jim MMAD) and then exposed these subjects to FA or
22      O3 (0.2 and 0.4 ppm) during light IE for 2 h. Gerrity et al. (1993) exposed 15 healthy subjects to
23      FA or 0.4 ppm O3 during CE  (40 L/min) for  1 h; at 2 h post O3 exposure, subjects then inhaled
24      radiolabeled particles (5 jim MMAD).  Subjects in both studies had similar pulmonary function
25      responses (average FVC decrease of 11 to 12%) immediately post exposure to 0.4 ppm O3.  The
26      Foster et al. (1987) study suggested there  is a stimulatory affect of O3 on mucociliary clearance;
27      whereas, Gerrity et al. (1993) found that in the recovery period following O3 exposure, mucus
28      clearance is similar to control, i.e., following a FA exposure. The clearance findings in these
29      studies are complementary not conflicting.  Investigators in both studies suggested that
30      O3-induced increases in mucociliary clearance could be mediated by cholinergic receptors.
31      Gerrity et al. (1993) further suggested that transient clearance increases might be coincident to

        January 2005                             AX6-119     DRAFT-DO NOT QUOTE OR CITE

-------
 1      pulmonary function responses; this supposition based on the return of sRaw to baseline and the
 2      recovery of FVC to within 5% of baseline (versus an 11% decrement immediately postexposure)
 3      prior to clearance measurements.
 4           Insofar as the airway epithelial surface provides a barrier to entry of biological, chemical
 5      and particulate contaminants into the submucosal region, the maintenance of barrier integrity
 6      represents a component of host defense.  Many of the studies of upper and lower respiratory
 7      responses to O3 exposure previously cited above have reported increases in markers of airway
 8      permeability after both acute exposures and repeated exposures. These findings suggest that O3
 9      may increase access of airborne agents. In a study of bronchial epithelial cells obtained from
10      nonatopic and mild atopic asthmatic subjects (Bayram et al., 2002), cells were grown to
11      confluence and transferred to porous membranes. When the cultures again  reached confluence,
12      they were exposed to 0.01-0.1 ppm O3 or air and their permeability was assessed by measuring
13      the paracellular flux of 14C-BSA. The increase in permeability 24 h following O3 exposure was
14      observed to be significantly greater in  cultures of cells derived from asthmatics, compared to
15      healthy subjects.  Thus, the late increase  in airway  permeability following exposure of asthmatic
16      subjects to O3, of the sort described by Newson et al. (2000), may be related to an inherent
17      susceptibility of 'asthmatic' cells to the barrier-reducing effects of O3.
18           As referenced in Section 6.9.3, the  O3-induced increase in the numbers of CD8+ T
19      lymphocytes in the airways of healthy subjects reported by Blomberg, et al. (1997) poses several
20      interesting questions regarding possible alterations in immune surveillance  processes following
21      exposure. In a subsequent study from  the same group, Krishna et al. (1998) exposed healthy
22      subjects to 0.2 ppm O3 or filtered air for 2 h followed by BAL at 6 h. In addition to increased
23      PMNs and other typical markers of inflammation, they found a significant decrease in the
24      CD4+/CD8+ T lymphocyte ratio and in the proportion of activated CD4+ and CD8+ cells.
25      Studies relating to the effects of low-level O3 exposure on the influx and activity of immuno-
26      competent cells in the upper and lower respiratory tracts may shed additional light on
27      modulation of this important area of host defense.
28
29
        January 2005                            AX6-120      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.10  EXTRAPULMONARY EFFECTS OF OZONE
 2           Ozone reacts rapidly on contact with respiratory system tissue and is not absorbed or
 3      transported to extrapulmonary sites to any significant degree as such.  Laboratory animal studies
 4      suggest that reaction products formed by the interaction of O3 with respiratory system fluids or
 5      tissues may produce effects measured outside the respiratory tract—either in the blood, as
 6      changes in circulating blood lymphocytes, erythrocytes, and serum, or as changes in the structure
 7      or function of other organs, such as the parathyroid gland, the heart, the liver, and the central
 8      nervous system. Very little is known, however, about the mechanisms by which O3 could cause
 9      these extrapulmonary effects. (See Section 5.4 for a discussion of the systemic effects of
10      O3 observed in laboratory animals^)
11           The results from human exposure studies discussed in the previous criteria documents
12      (U.S. Environmental Protection Agency, 1986, 1996) failed to demonstrate any consistent
13      extrapulmonary effects. Early studies on peripheral blood lymphocytes collected from human
14      volunteers did not find any significant genotoxic or functional changes at O3 exposures of 0.4 to
15      0.6 ppm for up to 4 h/day. Limited data on human subjects indicated that 0.5 ppm O3 exposure
16      for over 2 h caused transient changes in blood erythrocytes and sera (e.g., erythrocyte fragility
17      and enzyme activities), but the physiological significance of these studies remains questionable.
18      The conclusions drawn from these early studies raise doubt that cellular damage or altered
19      function is occurring to circulating cells at O3 exposures under 0.5 ppm.
20           Other human exposure studies have attempted to identify specific markers of exposure to
21      O3 in blood. For example, Schelegle et al. (1989) showed that PGF2a was elevated after O3
22      exposure (0.35 ppm); however, no increase in a-1 protease inhibitor was observed by Johnson
23      et al.  (1986). Foster et al. (1996) found a reduction in the serum levels of the free radical
24      scavenger a-tocopherol after O3 exposure.  Vender et al. (1994) failed to find any changes in
25      indices of red blood cell antioxidant capacity (GSH, CAT) in healthy male subjects exposed to
26      0.16 ppm O3 for 7.5 h while intermittently exercising.  Liu et al. (1997, 1999) used a salicylate
27      metabolite, 2,3, dehydroxybenzoic acid (DFffiA), to indicate increased levels of hydroxyl radical
28      which hydroxylates salicylate to DHBA. Increased DHBA levels after exposure to 0.12 and
29      0.40 ppm suggest that O3 increases production of hydroxyl radical.  The levels of DHBA were
30      correlated with changes in spirometry.
        January 2005                            AX6-121      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Only a few experimental human studies have examined O3 effects in other non-pulmonary
 2      organ systems besides blood.  Early studies on the central nervous system (Gliner et al., 1979,
 3      1980) were not able to find significant effects on motor activity or behavior (vigilance and
 4      psychomotor performance) from O3 exposures at rest up to 0.75 ppm (U.S. Environmental
 5      Protection Agency, 1986).  Drechsler-Parks et al. (1995) monitored ECG, HR, cardiac output,
 6      stroke volume, and systolic time intervals in healthy, older subjects (56 to 85 years of age)
 7      exposed to 0.45 ppm O3 using a noninvasive impedance cardiographic method. No changes
 8      were found at this high O3 concentration after 2 h of exposure while the subjects exercised
 9      intermittently at 25 L/min.  Gong et al. (1998) monitored ECG, HR, cardiac output, blood
10      pressure, oxygen saturation, and chemistries, as well as calculating other  hemodynamic variables
11      (e.g., stroke volume,  vascular resistance, rate-pressure products) in both healthy and
12      hypertensive adult males, 41 to 78 years of age. No major acute cardiovascular effects were
13      found in either the normal or hypertensive subjects after exposure to 0.3 ppm O3 for 3 h with
14      intermittent exercise  at 30 L/min.  Statistically significant O3 effects for both groups combined
15      were a decrease in FEVl3 and  increases in HR, rate-pressure product, and the alveolar-to-arterial
16      PO2 gradient, which might be more important in some patients with severe cardiovascular
17      disease.  [See Section AX6.3 for a more  detailed discussion of the effects ofO3 exposure in
18      subjects with preexisting disease. ]
19
20
21      AX6.11   OZONE  MIXED WITH OTHER POLLUTANTS
22           Controlled laboratory studies simulating conditions of ambient exposures have failed for
23      the most part to demonstrate significant adverse effects either in healthy subjects, atopic
24      individuals, or in young and middle-aged asthmatics.
25
26      AX6.11.1  Ozone and Sulfur Oxides
27           The difference  in solubilities and other chemical properties of O3 and SOX seems to limit
28      chemical  interaction and formation of related species in the mixture of these pollutants either in
29      liquid or gaseous phase. Laboratory studies reviewed in the previous O3 criteria document
30      (Table AX6-14) reported,  except for one study (Linn et al., 1994), no significant effects on
31      healthy individuals exposed to mixtures of O3 and SO2 or H2SO4 aerosol.  In the study of Linn

        January 2005                            AX6-122      DRAFT-DO NOT QUOTE OR CITE

-------
                                         Table AX6-14. Ozone Mixed with Other Pollutants3
to
o
o
X

 I

to
H

6
o


o
H

/O
o
HH
H
W
Concentration11
ppm Hg/in3
Pollutant
Exposure Duration Exposure
and Activity Conditions'
Number and
Gender of Subject
Subjects Characteristics
Observed Effect(s)
Reference
Sulfur-Containing Pollutants
0.0
0.1 +
0.1 +





0.2
0.3

0.12
0.30

0.05


0.12









0.08
0.12
0.18







0
196" +
262" +
101"




392
564

235
564
70



235
100








157
235
353
100
100





Air
03 +
S02 +
H2S04




03
NO2
H2SO4
03
N02
H2SO4
HNO3


03
H2SO4








03
03
03
NaCl
H2S04





4 h 25 °C
IE 15' ex/ 15' rest 50% RH
VE=22L/min





90 min. 21 °C
VE ~ 32 L/min 50% RH
IE 3 x 15 min
1. 5 h with IE for 22 °C
2 consecutive days; 65% RH
VE = 23.2L/min



6.5 h 21 °C
2 consecutive days 50% RH
50 min exercise/h
VE = 29 L/min






3-h exposure to aerosol, 21 °C
followed 24 h later by a -40% RH
3-h exposure to O3. IE
(10 min per half hour)
VE = 4 times resting
(30 to 364 min)




8 M, 7 F Healthy
1 M, 4 F Asthmatic
10M, 11F Allergic
All NS,
9 to 12 yrs. old



24 Asthmatic NS,
(17 M, 7F) 11 to 18 years old

22 completed Asthmatic NS,
study; adolescents; NS, 12 to
15M, 7F 19 years old



8 M, 7 F Nonasthmatic NS,
22 to 4 1 years old

13M, 17 F Asthmatic NS,
18 to 50 years old





16 M, 14 F Nonasthmatic NS,
18 to 45 years old
10 M, 20 F Asthmatic NS,
21 to 42 years old






Spirometry, PEER and subjective
symptoms score showed no meaningful
changes between any condition for a
total study population. The symptoms
score reported by a subset of
asthmatics/allergies were positively
associated with inhaled concentration
ofH2SO4(p = 0.01).
H2SO4/O3/NO2, O3/NO2 and clean air
produced similar responses

No significant pulmonary function
changes following any exposure
compared to response to clean air.
Six additional subjects started the
study, but dropped out due to
uncomfortable symptoms.
Exposure to O3 or O3 + H2SO4 induced
significant decrements in
forced expiratory function. Differences
between O3 and O3 + H2SO4 were, at
best, marginally significant. O3 is the
more important pollutant for inducing
respiratory effects. A few asthmatic
and nonasthmatic subjects were more
responsive to O3 + H2SO4 than to
O3 alone.
No significant changes in symptoms or
lung function with any aerosol/O3
combination in the healthy group.
In asthmatics, H2SO4 preexposure
enhanced the small decrements in FVC
that occurred following exposure to
0.18ppmO3. Asthmatics had no
significant changes on FEV; with any
O3 exposures, but symptoms were
greater.
Linn et al.
(1997)






Linnetal.
(1995)

Koenig et al.
(1994)




Linn et al.
(1994)








Utell et al.
(1994)
Frampton
etal. (1995)







-------
                                      Table AX6-14 (cont'd). Ozone Mixed with Other Pollutants3
to
o
o
X

 I

to
H

6
o


o
H

/O
Concentration11

ppm

Hg/m3

Pollutant
Number and
Exposure Duration Exposure Gender of
and Activity Conditions' Subjects

Subject
Characteristics


Observed Effect(s)


Reference
Sulfur-Containing Pollutants (cont'd)
0.12
0.10



0.25



235
262



490
1,200
to
1,600
03
SO2



03
H2SO4


1 h (mouthpiece) 22 °C 8 M, 5 F
IE VE ~ 30 L/min 75% RH
45-min exposure to air or
O3, followed by 1 5-min
exposure to O3 or SO2
2h 35 °C 9M
IE 83% RH
VE = 30 to 32 L/min

Allergic asthmatics,
12 to 18 years old,
medications withheld
for at least 4 h before
exposures
Healthy NS,
19 to 29 years old


Prior exposure to O3 potentiated
pulmonary function responses to SO2;
decrements in FEVj were -3, -2, and
-8% for the air/O3, O3/O3, and O3/SO2
exposures, respectively.
No significant effects of exposure to O3
alone or combined with H2SO4 aerosol.


Koenig et al.
(1990)



Horvath
etal. (1987)


Nitrogen-Containing Pollutants
0.0
0.2
0.4
0.2+0.4

0.1
0.2
0.1+0.2




0.0
0.36
0.36
0.75
0.36

0.45
0.60


0
392
752


196
376





0
706
677
1,411
943

883
1,129


Air
03
NO2
O3+NO2

03
N02
03+N02




Air
03
NO2
NO2
SO2

03
NO2


3 h 9 M, 2 F
IE 10' ex/ 20' rest
VE =32 L/min


6h
IE 10' ex/ 20' rest
VE =32 L/min
T = 25°C
RH = 50%


2 h Head only 6 M, 6 F
rest exposure




2-h random exposures to 23.6 °C 6 M
FA, O3, NO2, and O3 + 62% RH 2 F
NO2; IE; VE = 26-29
L/min
Atopic asthmatics
22 to 41 yrs. old










Healthy NS
19 to 33 yrs. old




Healthy, NS, 56 to
85 years old


Exposure to NO2 alone had minimal
effects on FEVj. However, O3 alone or
in combination elicited significantly
greater decline in FEVj in a short (3 h)
exposure (higher concentrations) than a
long (6 h) exposure where the effects
were nonsignificant. Allergen
challenge inhalation significantly
reduced PD20 FEV; in all short but not
the long exposures. No additive or
potentiating effects have been
observed.
For NO2 and SO2 the absorbed fraction
of O3 increased relative (to baseline)
whereas after O3 exposure it decreased.
The differences explained by an
increased production of O3-reactive
substrate in ELF due to inflammation.
Exercise-induced cardiac output was
smaller with O3 + NO2 exposure
compared to FA or O3 alone.

Jenkins et al.
(1999)










Rigas et al.
(1997)




Drechsler-
Parks et al.
(1995)

o
HH
H
W

-------
                                                 Table AX6-14 (cont'd). Ozone Mixed with Other Pollutants3
to
o
o
X
 I
to
H
6
o
o
H
/O
o
HH
H
W
Concentration11
ppm
(ig/m3 Pollutant
Exposure Duration
and Activity
Exposure
Conditions'
Number and
Gender of
Subjects
Subject
Characteristics
Observed Effect(s)
Reference
         Nitrogen-Containing Pollutants (cont'd)
0.30
0.60









0.2





0.0
0.12
0.30
0.12+0.30

589 O3
1,129 N02









392 O3
500 HNO3
H20



0 Air
235 03
564 N02
03 + N02

2-h exposure to NO2 or 21 °C
FA, followed 3 h later by 40% RH
2-h exposure to O3, IE
VE = 20 L/min/m2 BSA







5 h 20 °C
IE (50 min/h exercise) 5% RH
VE ~ 40 L/min
2 h HNO3 or H2O fog or
air, followed by 1-h break,
followed by 3 h O3
1 h (mouthpiece) 22 °C
IE 75% RH
VE = 33 L/min
VE = 35 L/min

21 F Healthy NS,
18 to 34 years old









6 M, 4 F Healthy NS,
minimum of 1 0%
decrement in FEV;
after 3 h exposure to
0.20 ppm O3 with
50 min exercise/h
5 M, 7 F Healthy NS,
12 to 17 years old

9 M, 3 F Asthmatic
13 to 18 years old
No significant effect of NO2 exposures
on any measured parameter. Sequential
exposure of NO2 followed by O3
induced small but significantly larger
decrements in FEVj and FEF25.75 than
FA/O3 sequence. Subjects had
increased airway responsiveness to
methacholine after both exposures, with
significantly greater responsiveness
after the NO2/O3 sequences than after
the FA/O3 sequence.
Exposure to HNO3 or H2O fog followed
by O3 induced smaller pulmonary
function decrements than air followed
by03.


Findings inconsistent across cohorts
and atmospheres. No significant
differences in FEV; and RT between
asthmatics and healthy, or between
atmospheres and cohorts.
Hazucha
etal. (1994)









Aris etal.
(1991)




Koenig et al.
(1988)



          0.30        589    O3          1 h (mouthpiece)
          0.60        1,129   N02        CE
                                        VE ~ 70 L/min for men
                                        VB ~ 50 L/min for women
20 M, 20 F
Healthy NS,
21.4±1.5(SD)
years old for F,
22.7 ± 3.3 (SD) years
old for M
No differences between responses to
O3 and NO2 + O3 for spirometric
parameters. Increase in SRaw with
NO2 + O3 was significantly less than for
O, alone.
Adams etal.
(1987)
0.30
0.30

0.15
0.15

0.15
0.15
0.15

589
564
200
294
284
200
294
282
393
200
03
N02
H2S04
03
NO2
H2SO4
03
N02
S02
H2S04
2 hCE for 20 min 28 to 29 °C 6M
V = 25 L/min 50 to 60% RH

2 h, 60 min 6 M
total exercise
V ~ 25 L/min
2 h, 60 min 3 M
total exercise
V = 25 L/min

Healthy subjects, Possible small decrease in SGaw
some smokers

Possible small decrease in SGaw


Possible small decrease in FEV;



Kagawa
(1986)









-------
3
C
^
to
o
o











>
X
ON
i
to
ON



O
^>
H
1
O
o
0
H
O
xauii; r^^w
Concentration11
Exposure Duration
ppm Hg/in3 Pollutant and Activity
Peroxyacetyl Nitrate
0.45 883 O3 2h
0.60 1,129 N02 IE
0.13 644 PAN VE = 25 L/min
0.45 883 O3 2h
0.30 1,485 PAN IE
VE = 27 L/min
0.485 952 O3 2h
0.27 1,337 PAN IE
VE = 25 L/min

Particle-Containing Pollutants
0.0 0 Air 2-2. 5 h
0.12 235b+ O3+ rest
153b PM25





"See Appendix A for abbreviations and acronyms.
'Grouped by pollutant mixture.
CWBGT = 0.7 Twetbult + 0.3 T^hulhagl
-------
 1      et al. (1994), which was a repeated 6.5 h exposure protocol, O3 alone and O3 + H2SO4 induced
 2      significant spirometric decrements in healthy adults and asthmatics, but the magnitude of effects
 3      between exposure atmospheres was not significant. Asthmatic and atopic subjects showed
 4      somewhat enhanced or potentiated response to mixtures or sequential exposure, respectively;
 5      however, the observed effects were almost entirely attributable to O3 (U.S. Environmental
 6      Protection Agency, 1996). Thus, in both healthy and asthmatic subjects, the interactive effects
 7      of O3 and other pollutants were marginal and the response was dominated by O3.
 8           Since 1994, the only laboratory study that examined the health effects of a mixture of O3
 9      and sulfur oxides (SO2 and H2SO4) has been that of Linn et al.  (1997). In this study, the
10      investigators closely simulated ambient summer haze air pollution conditions in Uniontown, PA
11      as well as controlled the selection of study subjects with the objective to corroborate earlier
12      reported findings of an epidemiologic study of Neas et al. (1995). The subjects were 41 children
13      (22F/19M) 9 to 12 yrs old. Of these, 26 children had history of asthma or allergy. During a
14      14-day study period, children were exposed on the 4th and 11th day for 4 hrs (IE, 15 min @ avg.
15      VE 22 L/min) in random order to air and a mixture of 0.10 ppm O3, 0.10 ppm SO2 and 42 to
16      198 mg/m3 H2SO4 (mean cone. 101 mg/m3, 0.6 mm MMAD).  The effects of controlled
17      exposures were assessed by spirometry. Except for exposure days, children used diaries to
18      record activity, respiratory symptoms, location, and PEFR. Thus, every exposure day was
19      bracketed by 3 days of monitoring.  Spirometry, PEFR, and respiratory symptoms score showed
20      no meaningful changes between any condition for a total study population.  The symptoms score
21      reported by a subset of asthmatic/allergic subjects was positively associated with  the inhaled
22      concentration of H2SO4 (p = 0.01). However, the reported symptoms were different from the
23      ones reported in the Uniontown study (Neas et al., 1995). Although retrospective statistical
24      power calculations using these study observations for the symptoms score, PEFR, and
25      spirometric endpoints were sufficient to detect with > 80% probability the same magnitude of
26      changes  as observed in Uniontown, the effects were minimal and not significant.  The divergent
27      observations of the two studies have been explained by the presence of an unidentified
28      environmental factor in Uniontown, differences in physico-chemical properties of acid,
29      differences in time course of exposure  and history of previous  exposure of children to pollutants,
30      psychological and physiological factors related to chamber exposures, and by other conjectures.
31

        January 2005                            AX6-127      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.11.2  Ozone and Nitrogen-Containing Pollutants
 2           Nitrogen dioxide is a key component of the photooxidation cycle and formation of O3.
 3      Both gases are almost invariably present in ambient atmosphere. Compared to O3, NOX species
 4      have limited solubility and moderate oxidizing capability. Both O3 and NO2 are irritants and
 5      tissue oxidants and exert their toxic actions through many common mechanisms. The regional
 6      dosimetry and the primary sites of action of O3 and NO2 overlap but are not the same. Since
 7      these gases are relatively insoluble in water, they will likely penetrate into the peripheral airways
 8      that are more sensitive to damage than better protected conducting airways. The controlled
 9      studies reviewed in the previous O3 criteria document (Table AX6-14) generally reported only
10      small pulmonary function changes after combined exposures of NO2 or nitric acid (HNO3) with
11      O3, regardless if the interactive effects were potentiating or additive.  In two of these studies, the
12      effects reached statistical significance, but they were not coherent. Preexposure with NO2
13      potentiated both spirometric and nonspecific airway reactivity response following subsequent O3
14      exposure (Hazucha et al., 1994); however, exposure to NO2 + O3 mixture blunted SRaw increase
15      as compared to O3 alone (Adams et al.,1987). As with O3 and SOX mixtures, the effects have
16      been dominated by O3 (U.S. Environmental Protection Agency, 1996).
17           Combined exposure to O3 and NO2 also blunted the exercise-induced increase in cardiac
18      output found with FA and O3 exposures alone (Drechsler-Parks, 1995). Eight healthy older
19      subjects (56 to 85 years of age) were exposed for 2 h to FA, 0.60 ppm NO2, 0.45 ppm O3, and to
20      0.60 ppm NO2 + 0.45 ppm O3 while alternating 20-min periods of rest and exercise.  Cardiac
21      output, HR, stroke volume, and systolic time intervals were measured by noninvasive impedance
22      cardiography at the beginning of each exposure, while the subjects were at rest, and again during
23      the last 5 min of exercise. Metabolic exercise data (VE, VO2, fB) also were measured.  There
24      were no statistically significant differences between exposures for FIR, VE, VO2, fB,  stroke
25      volume, or systolic time intervals.  Exercise increased cardiac output after all exposures;
26      however, the incremental increase over rest was significantly smaller for the combined O3 and
27      NO2 exposures. The authors speculated that nitrate and nitrite reaction products from the
28      interaction of O3 and NO2 cross the air/blood interface in the lungs, causing peripheral
29      vasodilation and a subsequent drop in cardiac output. No major cardiovascular effects of O3
30      only exposures have been reported in human subjects (see Section AX6.10).
        January 2005                            AX6-128      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Despite suggested potentiation of O3 response by NO2 in healthy subjects, it is unclear
 2      what response, and at what dose, either sequential or combined gas exposures will induce in
 3      asthmatics.  Jenkins et al. (1999) exposed 11 atopic asthmatics in random order to air, 0.1 ppm
 4      O3, 0.2 ppm NO2, and 0.1 ppm O3 + 0.2 ppm NO2 for 6 h (IE for 10 min @ 32 L/min every
 5      40 min). Two weeks later, 10 of these subjects were exposed for 3 h to doubled concentrations
 6      of these gases (i.e., 0.2 ppm O3,  0.4 ppm NO2, and 0.2 ppm O3 + 0.4 ppm NO2) employing the
 7      same exercise regimen. Immediately following each exposure, subjects were challenged with
 8      allergen (D. pteronyssinus) and PD20 FEVj was determined. Exposure to NO2 alone had
 9      minimal effects on FEVj or airway responsiveness. However, O3 alone or in combination with
10      NO2 elicited a significantly (p < 0.05) greater decline in FEVj in a short (3 h) exposure (higher
11      concentrations) than the long (6 h) exposure, where the effects were not significant. Allergen
12      challenge inhalation significantly (p = 0.018 to 0.002) reduced PD20 FEVj in all short, but not the
13      long, exposures. No associations were observed between pollutant concentrations and
14      physiologic endpoints.  The statistical analyses of these data suggest that the combined effect
15      (O3 + NO2) on lung function (FVC, FEVj) was not significantly greater than the effect of
16      individual gases for 6-h exposures, thus no additive or potentiating effects have been observed.
17      Shorter 3-h exposures using twice as high NO2 concentrations, however, showed significant
18      FEVj decrements following exposures to atmospheres containing O3. The analysis also suggests
19      that it is the inhaled concentration, rather than total dose, that determines lung airway
20      responsiveness to allergen.
21           The potential for interaction between O3 and other gas mixtures was studied by Rigas  et al.
22      (1997). They used an O3 bolus absorption technique to determine how exposures to O3, NO2,
23      and SO2 will affect distribution of O3 adsorption by airway mucosa.  The selected O3 bolus
24      volume was set to reach lower conducting airways. Healthy young nonsmokers (6F/6M) were
25      exposed on separate days at rest in a head dome to 0.36 ppm O3, 0.36 ppm NO2, 0.75 ppm NO2
26      and 0.75 ppm SO2 for 2 h.  The rationale for the selection of these gases was their differential
27      absorption.  Because O3 and NO2 are much less soluble in liquid (i.e., ELF) than SO2, they are
28      expected to penetrate deeper into the lung than SO2 which is absorbed more quickly in the
29      epithelial lining fluid of the upper airways. The actual experimental measurements have shown
30      that during continuous NO2 and SO2 exposure the absorbed fraction  of an O3 bolus in lower
31      conducting airways increased relative to baseline, whereas during continuous O3 exposure the O3

        January 2005                            AX6-129     DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
bolus fraction in lower conducting airways decreased. The authors attempted to explain the
differences by suggesting that there may be increased production of an O3-reactive substrate in
epithelial lining fluid due to airway inflammation. As interpreted by the investigators, during
NO2 and SO2 exposures the substrate was not depleted by these gases and so could react with the
O3 bolus, whereas during O3 exposure the substrate was depleted, causing the fractional
absorption of the O3 bolus to decrease. Greater absorption in males than females for all gases
was attributed to anatomical differences in the bronchial tree.
 9      AX6.11.3  Ozone and Other Pollutant Mixtures Including Particulate Matter
10           Almost all of the studies published over the last twenty years investigating the health
11      effects of mixtures of O3 with other air pollutants involved peroxyacetyl nitrate (PAN).  These
12      studies on healthy individuals exposed under laboratory conditions came from the Horvath
13      laboratory at UC Santa Barbara (Table AX6-13). In the last of this series of studies, Drechsler-
14      Parks and colleagues (1989) found the same equivocal interaction of O3 and PAN as in previous
15      studies, which is attributable to O3 exposure alone (U.S. Environmental Protection Agency,
16      1996). Subsequently, only a couple of studies have investigated the effects of more complex air
17      pollutant mixtures on human pathophysiology under controlled conditions.
18           It is not only the interaction between air pollutants in ambient air; but, as Rigas et al.
19      (1997) has found, an uneven distribution of O3,  SO2, and NO2 absorption in the lower conducting
20      airways of young healthy subjects may modulate pathophysiologic response as well.  Exposure
21      to SO2 and NO2 increased, while exposure to O3 decreased, the absorbing capacity of the airways
22      for O3. The authors have suggested that SO2 or NO2 -inflamed airways release additional
23      substrates into the epithelial lining fluid that react with O3, thus progressively removing O3 from
24      the airway lumen.  This mechanism may explain findings  of antagonistic response (e.g., Adams
25      et al., 1987; Dreschler-Parks, 1995) when the two gases are combined in an exposure
26      atmosphere.
27           The mechanisms by which inhalation exposure to other complex ambient atmospheres
28      containing particulate matter (PM) and O3 induce cardiac  events frequently reported in
29      epidemiologic studies are rarely studied in human subjects under laboratory conditions.
30      Recently, Brook et al. (2002) have reported changes in brachial artery tone and reactivity in
31      healthy nonsmokers following 2-h exposures to a mixture of 0.12 ppm O3 and 153 |ig/m3 of

        January 2005                            AX6-130      DRAFT-DO NOT QUOTE OR CITE

-------
 1      concentrated ambient PM25, and a control atmosphere of filtered air with a trace of O3
 2      administered in random order. Neither systolic nor diastolic pressure was affected by pollutant
 3      exposure despite a significant brachial artery constriction and a reduction in arterial diameter
 4      when compared to filtered air (p = 0.03).  The authors postulate that changes in arterial tone may
 5      be a plausible mechanism of air pollution-induced cardiac events.  However, the observations of
 6      no changes in blood pressure, and an absence of flow- and nitroglycerin- mediated brachial
 7      artery dilatation, cast some doubt on the plausibility of this mechanism. A number of other
 8      proposed mechanisms advanced to establish a link between cardiac events due to pollution and
 9      changes in vasomotor tone based on the findings of this study are purely speculative.
10
11
12      AX6.12   CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES
13          A large amount of informative O3 exposure-effects data has been obtained in controlled
14      laboratory exposure studies under a variety of different experimental conditions.  However,
15      laboratory simulation of the variable pollutant mixtures present in ambient air is not practical.
16      Thus, the exposure effects of one or several artificially generated pollutants (i.e., a simple
17      mixture) on pulmonary function and symptoms may not explain responses to ambient air where
18      complex pollutant mixtures exist. Epidemiologic studies, which do investigate ambient air
19      exposures, do not typically provide  the level of control and monitoring necessary to adequately
20      characterize short term responses. Thus, controlled exposures to ambient air using limited
21      numbers of volunteers have been used to try and bridge the gap between laboratory and
22      community exposures.
23
24      AX6.12.1   Mobile Laboratory Studies
25          As presented in previous criteria documents (U.S. Environmental Protection Agency, 1986;
26      1996), quantitatively useful information on the effects of acute exposure to photochemical
27      oxidants on pulmonary function and symptoms responses originated from field studies using a
28      mobile laboratory. These field studies involved subjects exposed to ambient air, FA without
29      pollutants, or FA containing artificially generated concentrations of O3 that are comparable to
30      those measured in the ambient environment. As a result, measured pulmonary responses in
31      ambient air can be directly compared to those found in more artificial or controlled conditions.

        January 2005                            AX6-131      DRAFT-DO NOT  QUOTE OR CITE

-------
 1      However, the mobile laboratory shares some of the same limitations of stationary exposure
 2      laboratories (e.g., limited number of both subjects and artificially generated pollutants for
 3      testing). Further, mobile laboratory ambient air studies are dependent on ambient outdoor
 4      conditions which can be unpredictable, uncontrollable, and not completely characterizable.
 5           As summarized in Table AX6-15, investigators in California used a mobile laboratory and
 6      demonstrated that pulmonary effects of ambient air in Los Angeles residents are related to O3
 7      concentration and level of exercise (Avol et al., 1983, 1984, 1985a,b,c, 1987; Linn et al., 1980,
 8      1983).  Avol et al. (1987) observed no significant pulmonary function or symptoms responses in
 9      children (8 to 11 years) engaged in moderate continuous exercise for 1 h while breathing
10      ambient air with an O3 concentration of 0.113 ppm.  However, significant pulmonary function
11      decrements and increased symptoms of breathing discomfort were observed in healthy
12      exercising (1 h continuous) adolescents (Avol et al., 1985a,b), athletes, (Avol et al., 1984, 1985c)
13      and lightly exercising asthmatic subjects (Linn et al., 1980, 1983) at O3 concentrations averaging
14      from 0.144 to 0.174 ppm. Many of the healthy subjects with a history of allergy appeared to be
15      more responsive to  O3 than "nonallergic" subjects (Linn et al., 1980, 1983), although a
16      standardized evaluation of atopic status was not performed.  Comparative studies of exercising
17      athletes (Avol et al., 1984, 1985c) with chamber exposures to oxidant-polluted ambient air
18      (mean O3 concentration of 0.153  ppm) and purified air containing a controlled concentration of
19      generated O3 at 0.16 ppm showed similar pulmonary function responses and symptoms,
20      suggesting that acute exposures to coexisting ambient pollutants had minimal contribution to
21      these responses under the typical summer ambient conditions in Southern California.  This
22      contention is similar to, but extends, the laboratory finding of no significant difference in
23      pulmonary function effects between O3 and O3 plus PAN exposures (Drechsler-Parks, 1987b).
24      Additional supporting evidence is provided in Section AX6.11.
25
26      AX6.12.2  Aircraft Cabin Studies
27           Respiratory symptoms and pulmonary function effects resulting from exposure to  O3 in
28      commercial aircraft flying at high altitudes, and in altitude-simulation studies, have been
29      reviewed elsewhere (U.S. Environmental Protection Agency, 1986, 1996). Flight attendants,
30      because of their physical activities at altitude, tend to receive higher exposures. In a series of
31      hypobaric chamber studies of nonsmoking  subjects exposed to  1,829 m (6,000 ft) and O3 at

        January 2005                           AX6-132    DRAFT-DO NOT QUOTE  OR CITE

-------
fa
to
O
o













X
ON

OJ
OJ




O
3>
'•Tj
H
1
O
o
2|
0
h-j

o
Table AX6-15. Acute Effects of Ozone in Ambient Air in Field S
Mean Ozone
Concentration" Ambient
ppm MS/m3 (°C) Duration (VE) Number of Subjects
0.1 13. ±.033 221 ±65 33 ± 1 Ih CE (22 L/min) 66 healthy children,
8 to 1 1 years old


0.144 ±.043 282 ±84 32 ± 1 Ih CE (32 L/min) 59 healthy
adolescents,
12 to 15 years old


0.153 ±.025 300 ±49 32 ±2 Ih CE (53 L/min) 50 healthy adults
(competitive
bicyclists)



0.156 ±.055 306 ±107 33 ±4 Ih CE (38 L/min) 48 healthy adults,
50 asthmatic adults



0.165 ±.059 323 ±115 33 ± 3 Ih CE (42 L/min) 60 "healthy" adults
(7 were asthmatic)




0.174 ±.068 341 ±133 33 ±2 2h IE (2 times resting) 34 "healthy" adults,
at 15-min intervals 30 asthmatic adults




studies With a Mobile Laboratory"
Observed Effect(s)
No significant changes in forced expiratory
function and symptoms of breathing
discomfort after exposure to 0. 1 1 3 ppm O3 in
ambient air.
Small significant decreases in FVC (-2.1%),
FEV075 (-4.0%), FEV; (-4.2%), andPEFR
(-4.4%) relative to control with no recovery
during a 1-h postexposure rest; no significant
increases in symptoms.
Mild increases in symptoms scores and
significant decreases in FEV; (-5.3%) and
FVC; mean changes in ambient air were not
statistically different from those in purified air
containing 0.16 ppm O3.

No significant changes for total symptom
score or forced expiratory performance in
normals or asthmatics; however, FEV;
remained low or decreased further (-3%) 3 h
after ambient air exposure in asthmatics.
Small significant decreases in FEV[ (-3.3%)
and FVC with no recovery during a 1-h
postexposure rest; TLC decreased and AN2
increased slightly.


Increased symptom scores and small
significant decreases in FEV[ (-2.4%), FVC,
PEFR, and TLC in both asthmatic and healthy
subjects; however, 25/34 healthy subjects were
allergic and "atypically" reactive to polluted
ambient air.

Reference
Avol et al.
(1987)


Avol et al.
(1985a,b)



Avol et al.
(1984, 1985c)




Linn et al.
(1983)
Avol et al.
(1983)

Linn et al.
(1983)
Avol et al.
(1983)


Linn et al.
(1980, 1983)




O
HH
H
W
        aSee Appendix A for abbreviations and acronyms.
        bRanked by lowest level of O3 in ambient air, presented as the mean ± SD.
        cMean±SD.

-------
 1      concentrations of 0.2 and 0.3 ppm for 3 or 4 h (Lategola et al., 1980a,b), increased symptoms
 2      and pulmonary function decrements occurred at 0.3 ppm but not at 0.2 ppm.
 3           Commercial aircraft cabin O3 levels were reported to be very low (average concentration
 4      0.01 to 0.02 ppm) during 92 randomly selected smoking and nonsmoking flights in 1989 (Nagda
 5      et al., 1989). None of these flights recorded O3 concentrations exceeding the 3-h time-weighted
 6      average (TWA) standard of 0.10 ppm promulgated by the Federal Aviation Administration
 7      (FAA, 1980), probably due to the use of O3-scrubbing catalytic filters (Melton, 1990).  However,
 8      in-flight O3 exposure can still occur because catalytic filters are not necessarily in continuous use
 9      during flight. Other factors to consider in aircraft cabins, however, are erratic temperature
10      changes, lower barometric pressure and oxygen pressure, and lower humidity, often reaching
11      levels between 4 and 17% (Rayman,  2002).
12           Ozone contamination aboard high-altitude aircraft also has been an interest to the U.S. Air
13      Force because of complaints by crew members of frequent symptoms of dryness and irritation of
14      the eyes, nose, and throat and an occasional cough (Hetrick et al., 2000).  Despite the lack of
15      ventilation system modifications as used in commercial aircraft, the O3 concentrations  never
16      exceeded the FAA ceiling limit of 0.25 ppm and exceeded the 3-h TWA of 0.10 ppm only 7% of
17      the total monitored flight time (43 h). The authors concluded that extremely low average
18      relative humidity (12%) during flight operations was most likely responsible for the reported
19      symptoms.
20
21
        January 2005                           AX6-134     DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Adams, W. C. (2000a) Feasibility study of prolonged ozone inhalation exposure via face mask. Inhalation Toxicol.
  3             12:299-313.
  4       Adams, W. C. (2000b) Ozone dose-response effects of varied equivalent minute ventilation rates. J. Exposure Anal.
  5             Environ. Epidemiol. 10: 217-226.
  6       Adams, W. C. (2002) Comparison of chamber and face-mask 6.6-hour exposures to ozone on pulmonary function
  7             and symptoms responses. Inhalation Toxicol. 14: 745-764.
  8       Adams, W. C. (2003a) Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm ozone via
  9             square-wave and triangular profiles on pulmonary responses. Inhalation Toxicol. 15: 265-281.
10       Adams, W. C. (2003b) Relation of pulmonary responses induced by 6.6-h exposures to 0.08 ppm ozone and 2-h
11             exposures to 0.30 ppm ozone via chamber and face-mask inhalation. Inhalation Toxicol. 15: 745-759.
12       Adams, W. C.; Ollison, W. M. (1997) Effects of prolonged simulated ambient ozone dosing patterns on human
13             pulmonary function and symptomatology. Presented at: 90th annual meeting of the Air & Waste Management
14             Association; June; Toronto, Ontario,  Canada. Pittsburgh, PA: Air & Waste Management Association; paper
15             no. 97-MP9.02.
16       Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation effects on pulmonary function and endurance
17             performance. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.  55: 805-812.
18       Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during continuous exercise via the
19             effective dose concept. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 51: 415-422.
20       Adams, W. C.; Brookes, K.  A.; Schelegle, E. S. (1987) Effects  of NO2 alone and in combination with O3 on young
21             men and women. J. Appl. Physiol. 62: 1698-1704.
22       Alexis, N.; Urch, B.; Tarlo,  S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000) Cyclooxygenase
23             metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
24             normals. Inhalation Toxicol. 12: 1205-1224.
25       Alexis, N. E.; Becker, S.; Bromberg, P. A.; Devlin, R.; Peden, D. B. (2004) Circulating GDI Ib expression correlates
26             with the neutrophil response and airway mCD14 expression is enhanced following ozone exposure in
27             humans. Clin. Immunol. Ill: 126-131.
28       Alpert, S. E.; Walenga, R. W.; Jaspers, I.; Qu, Q.; Chen, L. C. (1997) Ozone inactivates Cyclooxygenase in human
29             tracheal epithelial cells without altering PGHS-2 mRNA or protein. Am. J. Physiol. 272: L879-L887.
30       Arab, L.; Steck-Scott, S.; Fleishauer, A. T. (2002) Lycopene and the lung. Exp. Biol. Med. 227: 894-899.
31       Aris, R.; Christian, D.; Sheppard, D.; Balmes, J. R. (1991) The  effects of sequential exposure to acidic fog and
32             ozone on pulmonary function in exercising subjects. Am. Rev. Respir. Dis. 143: 85-91.
33       Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner,  W. E.; Balmes, J. R. (1993) Ozone-induced airway
34             inflammation in human subjects as determined by airway lavage and biopsy. Am. Rev. Respir. Dis. 148:
35             1363-1372.
36       Aris, R. M.; Tager, L; Christian, D.; Kelly, T.; Balmes, J. R.  (1995) Methacholine responsiveness is not associated
37             with O3-induced decreases inFEVj. Chest 107: 621-628.
38       Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N.  (2000) Ozone, but not nitrogen dioxide,
39             exposure  decreases glutathione peroxidases in epithelial lining fluid of human lung. Am. J. Respir. Crit. Care
40             Med. 162: 1342-1347.
41       Avol, E. L.; Linn, W. S.;  Shamoo, D. A.; Venet, T. G.; Hackney, J. D. (1983) Acute respiratory effects of
42             Los Angeles smog in continuously exercising  adults. J. Air Pollut.  Control Assoc. 33:  1055-1060.
43       Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative respiratory effects of
44             ozone and ambient oxidant pollution exposure during heavy exercise. J. Air Pollut. Control Assoc.
45             34:804-809.
46       Avol, E. L.; Linn, W. S.;  Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Venet, T. G.;  Hackney, J. D. (1985a)
47             Respiratory effects of photochemical oxidant air pollution in exercising adolescents. Am. Rev. Respir. Dis.
48             132:619-622.
49       Avol, E. L.; Linn, W. S.;  Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Hackney, J. D. (1985b) Short-term health
50             effects of ambient air pollution in adolescents. In: Lee, S. D., ed. Evaluation of the scientific basis for
51             ozone/oxidants standards: transactions of an APCA international specialty conference; November 1984;
52             Houston,  TX.  Pittsburgh, PA: Air Pollution Control Association; pp. 329-336. (APCA international specialty
53             conference transactions: TR-4).
         January 2005                                AX6-13 5      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J. D. (1985c) Comparative effects of
  2             laboratory generated ozone and ambient oxidant exposure in continuously exercising subjects. In: Lee, S. D.,
  3             ed. Evaluation of the scientific basis for ozone/oxidants standards: proceedings of an APCA international
  4             specialty conference; November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp.
  5             216-225. (APCA international specialty conference transactions: TR-4).
  6       Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M; Venet, T. G.; Trim, S. C.; Hackney, J. D.
  7             (1987) Short-term respiratory effects of photochemical oxidant exposure in exercising children. JAPCA
  8             37: 158-162.
  9       Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Hackney, J. D. (1988) Seasonal ozone reactivity  in Los Angeles residents.
10             Presented at: 81st annual meeting of the Air Pollution Control Association; June; Dallas, TX. Pittsburgh, PA:
11             Air Pollution Control Association; paper no. 88-122.6.
12       Ball, B. A.; Folinsbee, L. J.; Peden, D. B.; Kehrl, H. R. (1996) Allergen bronchoprovocation of patients with mild
13             allergic asthma after ozone exposure. J. Allergy Clin. Immunol. 98:  563-572.
14       Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, I.; Christian, D.; Hearne, P. Q.; Kelly, T.; Aris, R. M. (1996)
15             Ozone-induced decrements in FEVj and FVC do not correlate with measures of inflammation. Am. J. Respir.
16             Crit. Care Med. 153: 904-909.
17       Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.;  Christian, D.; Kelly, T.; Hearne,
18             P. Q.; Ferrando, R.; Welch, B. (1997) Effects of ozone on normal and potentially sensitive human subjects.
19             part I: airway inflammation and responsiveness to ozone in normal and asthmatic subjects. Cambridge, MA:
20             Health Effects Institute. Research report no. 78; pp 1-37, 81-99.
21       Bascom, R.; Naclerio, R. M.; Fitzgerald, T. K.; Kagey-Sobotka, A.; Proud, D. (1990) Effect of ozone inhalation on
22             the response to nasal challenge with antigen of allergic subjects. Am. Rev. Respir. Dis. 142: 594-601.
23       Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, J., Jr. (1994) Bronchoalveolar lavage
24             neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
25             Chest 106: 1757-1765.
26       Bates, D. V; Bell, G. M.; Burnham, C. D.; Hazucha, M.; Mantha, J.; Pengelly, L. D.; Silverman, F. (1972)
27             Short-term effects of ozone on the lung. J. Appl. Physiol. 32: 176-181.
28       Bayram, H.; Sapsford, R. J.; Abdelaziz, M. M.; Khair, O. A. (2001) Effect of ozone and nitrogen dioxide on the
29             release of proinflammatory mediators from bronchial epithelial cells of nonatopic nonasthmatic subjects and
30             atopic asthmatic patients in vitro. J. Allergy Clin. Immunol. 107: 287-294.
31       Bayram, H.; Rusznak, C.; Khair, O. A.; Sapsford, R. J.; Abdelaziz, M. M. (2002) Effect of ozone and nitrogen
32             dioxide on the permeability of bronchial epithelial cell cultures of non-asthmatic and asthmatic subjects.
33             Clin. Exp. Allergy 32: 1285-1292.
34       Beck, N.; Koenig, J. Q.; Luchtel, D. L.; Altaian, L. C.; Orsborn, M. T.; Kenney, J.  S. (1994) Ozone can increase the
3 5             expression of intercellular adhesion molecule-1 and the synthesis of cytokines by human nasal epithelial cells.
36             InhalationToxicol. 6: 345-357.
37       Becker, S.; Madden, M. C.; Newman, S. L.; Devlin, R. B.; Koren, H. S. (1991) Modulation of human alveolar
38             macrophage properties by ozone exposure in vitro. Toxicol. Appl. Pharmacol. 110:  403-415.
39       Bedi, J. F.; Drechsler-Parks, D. M.; Horvath, S. M. (1985) Duration of increased pulmonary function sensitivity to
40             an initial ozone exposure. Am. Ind. Hyg. Assoc. J. 46: 731-734.
41       Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1988) Reproducibility of the pulmonary function response of
42             older men and women to a 2-hour ozone exposure. JAPCA 38: 1016-1019.
43       Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1989) Adaptation by older individuals repeatedly exposed to
44             0.45 parts per million ozone for two hours. JAPCA 39: 194-199.
45       Bergamaschi, E.; De Palma, G.; Mozzoni, P.; Vanni, S.; Vettori, M. V.; Broeckaert, F.; Bernard, A.; Mutti, A.
46             (2001) Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects.
47             Am. J. Respir. Crit. Care Med. 163: 1426-1431.
48       Bhalla, D. K. (2002) Interactive effects of cigarette smoke and ozone in the induction of lung injury. Toxicol. Sci.
49             65: 1-3.
50       Blomberg, A.; Helleday, R.; Pourazar, J.; Stenfors, N.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.; Sandstrom, T. (1997)
51             Early airway and peripheral blood cell responses to 0.20 ppm ozone in healthy human subjects. Eur. Respir. J.
52             10(suppl 25): 274S.
53       Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.;
54             Sandstrom, T. (1999) Ozone-induced lung function decrements do not correlate with early airway
55             inflammatory or antioxidant responses. Eur. Respir. J. 13: 1418-1428.
         January 2005                                 AX6-136      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Blomberg, A.; Mudway, I.; Svensson, M; Hagenbjork-Gustafsson, A.; Thomasson, L.; Helleday, R.; Dumont, X.;
  2             Forsberg, B.; Nordberg, G.; Bernard, A. (2003) Clara cell protein as a biomarker for ozone-induced lung
  3             injury in humans. Eur. Respir. J. 22: 883-888.
  4       Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T.; Kelly, f. J.; Sandstrom, T.; Wilson, S.;
  5             Frew, A. J.; Blomberg, A. (2003) Ozone-induced bronchial epithelial cytokine expression differs between
  6             healthy and asthmatic subjects. Clin. Exp. Allergy 33: 777-782.
  7       Brook, R. D.; Brook, J. R.; Urch, B.; Vincent, R.; Rajagopalan, S.; Silverman, F. (2002) Inhalation of fine paniculate
  8             air pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation 105:  1534-1536.
  9       Brookes, K. A.; Adams, W. C.; Schelegle, E.  S. (1989) 0.35 ppm O3 exposure induces hyperresponsiveness on 24-h
10             reexposure to 0.20 ppm O3. J. Appl. Physiol. 66: 2756-2762.
11       Brummer, M. E. G.; Schwartz, L. W.; McQuillen, N. K. (1977) A quantitative study of lung damage by scanning
12             electron microscopy. Inflammatory cell response to high-ambient levels of ozone. In: Johari,  O.; Becker,
13             R. P., eds. Scanning electron microscopy/1977/11 - biological applications of the SEM: proceedings of the
14             workshops on advances inbiomedical applications of the SEM & STEM; March-April; Chicago, IL. Chicago,
15             IL: IIT Research Institute; pp. 513-518.
16       Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996) Longitudinal distribution of O3
17             absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
18       Centers for Disease Control and Prevention. (2003) Key clinical activities for quality asthma care: recommendations
19             of the National Asthma Education and Prevention Program. Morb. Mortal. Wkly. Rep. 52(RR-6): 1-8.
20             Available: http://www.cdc.gov/mmwr/PDF/rr/rr5206.pdf (11 April 2003).
21       Chen, L. L.; Tager, I. B.; Peden, D. B.; Christian, D. L.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (2004) Effect
22             of ozone exposure on airway responses to inhaled allergen in asthmatic subjects. Chest 125: 2328-2335.
23       Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998) Ozone-induced
24             inflammation is attenuated with multiday exposure. Am. J. Respir. Crit. Care Med. 158: 532-537.
25       Coffey, M. J.; Wheeler, C. S.; Gross, K. B.; Eschenbacher, W. L.; Sporn, P. H. S.; Peters-Golden, M. (1996)
26             Increased 5-lipoxygenase metabolism in the lungs of human subjects exposed to ozone. Toxicology
27             114: 187-197.
28       Coleridge, J. C. G.; Coleridge,  H. M.; Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of ozone stimulates
29             bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
30       Collins, D. V.; Cutillo, A. G.; Armstrong, J. D.; Crapo, R. O.; Kanner, R. E.; Tocino, I.; Renzetti, A. D., Jr. (1986)
31             Large airway size, lung  size, and maximal expiratory flow in healthy nonsmokers. Am. Res. Respir. Dis.
32             134:951-955.
33       Corradi, M.; Alinovi, R.; Goldoni, M.; Vettori, M.; Folesani, G.; Mozzoni, P.; Cavazzini, S.; Bergamaschi, E.; Rossi,
34             L.; Mutti, A. (2002) Biomarkers of oxidative stress after controlled human exposure to ozone. Toxicol. Lett.
35             134:219-225.
36       Criqui, G. I.; Solomon, C.; Welch, B. S.; Ferrando, R.  E.; Boushey, H. A.; Balmes, J. R. (2000) Effects of
37             azithromycin on ozone-induced airway neutrophilia and cytokine release. Eur. Respir. J. 15: 856-862.
38       Cropp, G. J. A.; Bernstein, I. L.; Boushey, H. A., Jr.; Hyde, R. W.; Rosenthal, R. R.; Specter, S. L.; Townley, R. G.
39             (1980) Guidelines for bronchial inhalation challenges with pharmacologic and antigenic agents. ATS News
40             (Spring): 11-19.
41       Devlin, R. B.; Koren, H. S. (1990) The use of quantitative two-dimensional  gel electrophoresis to analyze changes in
42             alveolar macrophage proteins in humans exposed to ozone. Am. J. Respir. Cell Mol. Biol. 2:  281-288.
43       Devlin, R. B.; McDonnell, W. F.; Koren, H. S.; Becker, S. (1990) Prolonged exposure of humans to 0.10 and
44             0.08 ppm ozone results in inflammation in the lung. Presented at: 83rd annual meeting of the Air & Waste
45             Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management Association;
46             paper no. 90-150.2.
47       Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H.  S. (1991)
48             Exposure of humans to ambient levels  of ozone for 6.6 hours causes cellular and biochemical changes in the
49             lung. Am. J. Respir. Cell Mol. Biol. 4:  72-81.
50       Devlin, R. B.; McKinnon, K. P.; Noah, T.; Becker, S.; Koren, H. S. (1994) Ozone-induced release of cytokines and
51             fibronectinby alveolar macrophages and airway epithelial cells. Am. J. Physiol. 266: L612-L619.
52       Devlin, R. B.; McDonnell, W. F.; Becker, S.;  Madden, M. C.; McGee, M. P.; Perez, R.; Hatch, G.; House, D. E.;
53             Koren, H.  S. (1996) Time-dependent changes of inflammatory mediators in the lungs of humans exposed to
54             0.4 ppm ozone for 2 hr:  a comparison of mediators found in bronchoalveolar lavage fluid 1 and 18  hr after
55             exposure. Toxicol. Appl. Pharmacol. 138: 176-185.
         January 2005                                 AX6-137      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Devlin, R. B.; Folinsbee, L. I; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997)
  2             Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhalation Toxicol.
  3             9:211-235.
  4       Dimeo, M. I; Glenn, M. G.; Holtzman, M. I; Sheller, J. R.; Nadel, J. A.; Boushey, H. A. (1981) Threshold
  5             concentration of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated
  6             exposures. Am. Rev. Respir. Dis. 124: 245-248.
  7       Drechsler-Parks, D. M. (1995) Cardiac output effects  of O3 and NO2 exposure in healthy older adults. Toxicol. Ind.
  8             Health 11: 99-109.
  9       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1984) Interaction of peroxyacetyl nitrate and ozone on
10             pulmonary functions. Am. Rev. Respir. Dis. 130: 1033-1037.
11       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1987a) Pulmonary function responses of older men and women
12             to ozone exposure. Exp. Gerontol. 22: 91-101.
13       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1987b) Pulmonary function desensitization on repeated
14             exposures to the combination of peroxyacetyl nitrate and ozone. JAPCA 37: 1199-1201.
15       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1989) Pulmonary function responses of young and older adults
16             to mixtures of O3, NO2 and PAN. Toxicol. Ind. Health 5: 505-517.
17       Drechsler-Parks, D. M.; Horvath, S. M.; Bedi, J. F. (1990) The "effective dose" concept in older adults exposed to
18             ozone. Exp. Gerontol. 25: 107-115.
19       Eschenbacher, W. L.; Ying, R. L.; Kreit, J. W.; Gross, K. B. (1989) Ozone-induced lung function changes in normal
20             and asthmatic subjects and the effect of indomethacin. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;
21             Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
22             international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
23             Science Publishers; pp. 493-499. (Studies in environmental science 35).
24       Fahy, J.  V.; Wong, H. H.; Liu, J. T.; Boushey, H. A. (1995) Analysis of induced sputum after air and ozone
25             exposures in healthy subjects. Environ. Res. 70: 77-83.
26       Federal Aviation Administration. (1980) Airplane cabin ozone contamination. F. R. (January 21) 45: 3880-3883.
27       Fernandes, A. L. G.; Molfino, N. A.; McClean, P. A.; Silverman, F.; Tarlo, S.; Raizenne, M.; Slutsky,  A. S.;
28             Zamel, N. (1994) The effect of pre-exposure to 0.12 ppm of ozone on exercise-induced  asthma. Chest
29             106: 1077-1082.
30       Folinsbee, L. J.; Hazucha, M.  J. (1989) Persistence of ozone-induced changes in lung function and airway
31             responsiveness. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
32             and its policy  implications: proceedings of the  3rd US-Dutch international symposium; May 1988; Nijmegen,
33             The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 483-492. (Studies in
34             environmental science 35).
35       Folinsbee, L. J.; Hazucha, M.  J. (2000) Time course of response to ozone exposure in healthy adult females.
36             Inhalation Toxicol.  12: 151-167.
37       Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure. Aviat. Space Environ.
38             Med.  57: 1136-1143.
39       Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1975) Exercise responses following ozone exposure. J. Appl.
40             Physiol. 38: 996-1001.
41       Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1977) Decrease of maximum work performance following ozone
42             exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 42: 531-536.
43       Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978) The influence of exercise on the pulmonary
44             function changes due to exposure to low concentrations of ozone. In: Folinsbee, L. J.; Wagner, J. A.; Borgia,
45             J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F., eds. Environmental  stress: individual human adaptations.
46             New York, NY: Academic Press; pp.  125-145.
47       Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly exposed to low
48             concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
49       Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1984) Pulmonary function changes after 1 h continuous heavy exercise
50             in 0.21 ppm ozone. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol. 57: 984-988.
51       Folinsbee, L. J.; Horstman, D. H.; Vorona, R. D.; Prince, J. M.; Berry, M.  (1986) Determinants of endurance
52             performance during ozone inhalation. Presented at:  30th International Union of Physiological Sciences
53             congress; July; Vancouver, BC,  Canada. Proc.  Int. Union Physiol. Sci. 16: 176.
54       Folinsbee, L. J.; McDonnell, W. F.; Horstman, D. H. (1988) Pulmonary function and symptom responses after
55             6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38: 28-35.
         January 2005                                 AX6-13 8       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Folinsbee, L. I; Devlin, R. B.; Abdul-Salaam, S.; Koren, H. S. (1993) Repeated severe ozone exposure causes
  2             depressed baseline spirometry. Am. Rev. Respir. Dis. 147: A638.
  3       Folinsbee, L. I; Horstman, D. H.; Kehrl, H. R.; Harder, S.; Abdul-Salaam, S.; Ives, P. J. (1994) Respiratory
  4             responses to repeated prolonged exposure to 0.12 ppm ozone. Am. J. Respir. Crit. Care Med. 149: 98-105.
  5       Folinsbee, L. J.; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam,  S.; Koren, H. S. (1998) Repeated
  6             exposure of humans to ozone: pulmonary function and symptom responses. Research Triangle Park, NC:
  7             U.S. Environmental Protection Agency; National Center for Environmental Assessment; unpublished data.
  8       Foster, W. M.; Stetkiewicz, P. T. (1996) Regional clearance of solute from the respiratory epithelia: 18-20 h
  9             postexposure to ozone. J. Appl. Physiol. 81:  1143-1149.
10       Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters tracheobronchial mucociliary function
11             in humans. J. Appl. Physiol. 63: 996-1002.
12       Foster, W. M.; Wills-Karp, M.; Tankersley, C. G.; Chen, X.; Paquette, N. C. (1996) Bloodborne markers in humans
13             during multiday exposure to ozone. J. Appl. Physiol. 81: 794-800.
14       Foster, W. M.; Brown, R. H.; Macri, K.; Mitchell, C. S. (2000) Bronchial reactivity of healthy subjects: 18-20 h
15             postexposure to ozone. J. Appl. Physiol. 89:  1804-1810.
16       Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993) Enhanced response to ozone exposure during the
17             follicular phase of the menstrual cycle. Environ. Health Perspect. 101: 242-244.
18       Foxcroft, W. J.; Adams, W. C. (1986) Effects of ozone exposure on four consecutive days on work performance and
19             VO2max.  J. Appl. Physiol. 61: 960-966.
20       Frampton, M. W.; Morrow, P. E.; Cox, C.; Levy, P. C.; Condemi, J. J.;  Speers, D.; Gibb, F. R.; Utell, M. J. (1995)
21             Sulfuric acid aerosol followed by ozone exposure in healthy and asthmatic  subjects. Environ. Res. 69: 1-14.
22       Frampton, M. W.; Morrow, P. E.; Torres, A.;  Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.; Speers, D. M. (1997a)
23             Ozone responsiveness in smokers and nonsmokers. Am. J. Respir. Crit. Care Med. 155: 116-121.
24       Frampton, M. W.; Balmes, J. R.;  Cox, C.; Krein, P. M.; Speers, D. M.; Tsai, Y.; Utell, M. J. (1997b) Effects of
25             ozone on normal and potentially sensitive human subjects. Part III: mediators of inflammation in
26             bronchoalveolar lavage fluid from nonsmokers, smokers, and asthmatic subjects exposed to ozone: a
27             collaborative study. Cambridge, MA: Health Effects Institute. Research report 78; pp. 73-99, 81-99.
28       Frank, R.; Liu, M. C.; Spannhake, E. W.; Mlynarek, S.; Macri, K.; Weinmann, G. G.  (2001) Repetitive ozone
29             exposure of young adults:  evidence of persistent small airway dysfunction. Am. J. Respir. Crit. Care Med.
30             164: 1253-1260.
31       Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex modulates airways
32             reactivity in rabbits. Respir. Physiol. 105: 95-102.
33       Gabrielson, E.  W.; Yu, X.-Y.; Spannhake, E. W. (1994) Comparison of the toxic effects of hydrogen peroxide and
34             ozone on cultured human bronchial epithelial cells. Environ. Health Perspect. 102: 972-974.
35       Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled particles measured
36             at 2 h after ozone exposure in humans. J. Appl. Physiol. 74: 2984-2989.
37       Gliner, J. A.; Matsen-Twisdale, J. A.; Horvath, S. M. (1979) Auditory and visual sustained attention during ozone
38             exposure. Aviat.  Space Environ. Med. 50: 906-910.
39       Gliner, J. A.; Horvath, S. M.; Sorich, R. A.; Hanley, J. (1980) Psychomotor performance during ozone exposure:
40             spectral and discriminant function analysis of EEG. Aviat. Space Environ. Med. 51: 344-351.
41       Gliner, J. A.; Horvath, S. M.; Folinsbee, L. J.  (1983) Preexposure to low ozone  concentrations does not diminish the
42             pulmonary function response on exposure to higher ozone concentrations. Am. Rev. Respir. Dis. 127:  51-55.
43       Golden, J. A.; Nadel, J. A.; Boushey, H. A. (1978) Bronchial hyperirritability in healthy subjects after exposure to
44             ozone. Am. Rev. Respir. Dis.  118: 287-294.
45       Gomez, F. P.; Rodriguez-Roisin,  R. (2002) Global initiative for chronic obstructive lung disease (GOLD) guidelines
46             for chronic obstructive pulmonary disease. Curr.  Opin. Pulm. Med. 8: 81-86.
47       Gong, H., Jr.; Tierney, D. F. (1995) Evaluation of chronic obstructive pulmonary disease patients for ozone
48             sensitivity:  validation of health advisories. Sacramento, CA.: California  State Air Resources Board, Research
49             Division; CARB-A133-123. Available from: NTIS, Springfield, VA; PB98-153620.
50       Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired  exercise performance and pulmonary
51             function in elite cyclists during low-level ozone exposure in a hot environment. Am. Rev. Respir. Dis.
52             134:726-733.
53       Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997a) Responses  of older men with and without
54             chronic  obstructive pulmonary disease to prolonged ozone exposure. Arch. Environ. Health 52: 18-25.
55       Gong, H., Jr.; McManus, M. S.; Linn, W. S. (1997b) Attenuated response to repeated daily ozone exposures in
56             asthmatic subjects. Arch. Environ. Health 52: 34-41.


         January 2005                                 AX6-139       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B.
  2             (1998) Cardiovascular effects of ozone exposure in human volunteers. Am. J. Respir. Crit. Care Med.
  3             158: 538-546.
  4       Graham, D. E.; Koren, H. S. (1990) Biomarkers of inflammation in ozone-exposed humans: comparison of the nasal
  5             and bronchoalveolar lavage. Am. Rev. Respir. Dis. 142: 152-156.
  6       Graham, D.; Henderson, F.; House, D.  (1988) Neutrophil influx measured in nasal lavages of humans exposed to
  7             ozone. Arch. Environ. Health 43: 228-233.
  8       Greve, J.  M; Davis, G.; Meyer, A. M; Forte, C. P.; Yost, S. C; Marlor, C. W.; Kamarck, M. E.; McClelland, A.
  9             (1989) The major human rhinovirus receptor is ICAM-1. Cell 56: 839-847.
10       Grievink, L.; Jansen, S. M. A.; Van't Veer, P.; Brunekreef, B. (1998) Acute effects of ozone on pulmonary function
11             of cyclists receiving antioxidant supplements. Occup. Environ. Med. 55: 13-17.
12       Grievink, L.; Zijlstra, A. G.; Ke, X.; Brunekreef, B. (1999) Double-blind intervention trial on modulation of ozone
13             effects on pulmonary function by antioxidant supplements. Am. J. Epidemiol. 149: 306-314.
14       Gwynn, R. C.; Thurston, G. D. (2001) The burden of air pollution: impacts among racial minorities. Environ. Health
15             Perspect. 109(suppl.  4): 501-506.
16       Hackney, J. D.; Linn, W. S.; Buckley, R. D.; Hislop, H. J. (1976) Studies in adaptation to ambient oxidant air
17             pollution: effects of ozone exposure in Los Angeles residents vs. new arrivals. Environ. Health Perspect.
18             18: 141-146.
19       Hackney, J. D.; Linn, W. S.; Mohler, J. G.;  Collier, C. R. (1977a) Adaptation to short-term respiratory effects of
20             ozone in men exposed repeatedly. J.  Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 82-85.
21       Hackney, J. D.; Linn, W. S.; Karuza, S. K.; Buckley, R. D.; Law, D. C.; Bates, D. V.; Hazucha, M.; Pengelly, L. D.;
22             Silverman, F. (1977b) Effects of ozone exposure in Canadians and southern Californians: evidence for
23             adaptation? Arch. Environ. Health 32: 110-116.
24       Hackney, J. D.; Linn, W. S.; Shamoo, D. A.; Avol, E. L. (1989) Responses of selected reactive and nonreactive
25             volunteers to ozone exposure in high and low pollution seasons. In: Schneider, T.; Lee, S. D.; Welters, G. J.
26             R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
27             US-Dutch international symposium;  May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands:
28             Elsevier Science Publishers; pp. 311-318. (Studies in environmental science 35).
29       Hamilton, R. F.; Li, L.; Eschenbacher,  W. L.; Szweda, L.; Holian, A. (1998) Potential involvement of
30             4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274: L8-L16.
31       Hanania,  N.  A.; Tarlo, S. M.; Silverman, F.; Urch, B.; Senathirajah, N.; Zamel, N.; Corey, P. (1998) Effect of
32             exposure to low levels of ozone  on the response to inhaled allergen in allergic asthmatic patients. Chest
33             114:752-756.
34       Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993) Ozone-induced increases in substance P and
3 5             8-epi-prostaglandin F2a in the  airways of human subjects. Am. J. Respir. Cell Mol. Biol. 9: 568-572.
36       Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
37             62: 1671-1680.
38       Hazucha, M. J. (1993) Meta-analysis and "effective dose" revisited. In: Mohr, U., ed. Proceedings of the 3rd
39             International Inhalation Symposium on Advances in Controlled Clinical Inhalation Studies; October,  1991.
40             Berlin, Germany: Springer-Verlag; pp. 247-256.
41       Hazucha, M. J.; Sant'Ambrogio, G. (1993) Effects of ozone on the activity of slowly (SAR) and rapidly adapting
42             (RAR) receptors in cats. FASEB J. 7: 407A.
43       Hazucha, M.; Silverman, F.; Parent, C.; Field, S.; Bates, D. V. (1973) Pulmonary function in man after short-term
44             exposure to ozone. Arch. Environ. Health 27: 183-188.
45       Hazucha, M. J.; Bates, D. V.; Bromberg, P. A. (1989) Mechanism of action of ozone on the human lung. J. Appl.
46             Physiol. 67: 1535-1541.
47       Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone concentration profiles
48             on pulmonary function. Am. Rev.  Respir. Dis. 146: 1487-1493.
49       Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P. A. (1994) Lung function response of healthy women after
50             sequential exposures to NO2 and O3. Am. J. Respir. Crit. Care Med. 150: 642-647.
51       Hazucha, M. J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.; Bromberg, P. A. (1996)
52             Effects of cyclo-oxygenase inhibition on ozone-induced respiratory inflammation  and lung function changes.
53             Eur. J. Appl. Physiol. Occup. Med. 73: 17-27.
54       Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003)  Distribution and reproducibility of spirometric response to
55             ozone by gender and age. J. Appl. Physiol. 95: 1917-1925.
         January 2005                                 AX6-140       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Henderson, F. W.; Dubovi, E. I; Harder, S.; Seal, E., Jr.; Graham, D. (1988) Experimental rhinovirus infection in
  2             human volunteers exposed to ozone. Am. Rev. Respir. Dis. 137: 1124-1128.
  3       Hetrick, S. M; Gould, W. D.; Christensen, D. E. (2000) Inflight cabin ozone aboard long duration C-5 airlift
  4             missions: a historical issue revisited. Aviat. Space Environ. Med. 71: 408-414.
  5       Hiltermann, T. J. N.; Stolk, I; Hiemstra, P. S.; Fokkens, P. H. B.; Rombout, P. J. A.; Sont, J. K.; Sterk, P. I;
  6             Dijkman, J. H. (1995) Effect of ozone exposure on maximal airway narrowing in non-asthmatic and asthmatic
  7             subjects. Clin. Sci. 89: 619-624.
  8       Hiltermann, T. J. N.; de Bruijne, C. R.; Stolk, J.; Zwinderman, A. H.; Spieksma, F. Th. M.; Roemer, W.;
  9             Steerenberg, P. A.; Fischer, P. H.; van Bree, L.; Hiemstra, P. S. (1997) Effects of photochemical air pollution
10             and allergen exposure on upper respiratory tract inflammation in asthmatics. Am. J. Respir. Crit. Care Med.
11             156:1765-1772.
12       Hiltermann, T. J. N.; Peters, E. A.; Alberts, B.; Kwikkers, K.;  Borggreven, P. A.; Hiemstra, P. S.; Dijkman, J. H.;
13             van Bree, L. A.; Stolk, J. (1998) Ozone-induced airway hyperresponsiveness in patients with asthma: role of
14             neutrophil-derived serine proteinases. Free Radical Biol. Med. 24: 952-958.
15       Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.; Sterk, P. J.;
16             Hiemstra, P. S.; Stolk, J.  (1999) Ozone-induced inflammation assessed in sputum and bronchial lavage fluid
17             from asthmatics: a new noninvasive tool in epidemiologic studies on air pollution and asthma. Free Radical
18             Biol. Med. 27: 1448-1454.
19       Holtzman, M. J.; Cunningham, J. H.; Sheller, J. R.; Irsigler, G. B.; Nadel, J. A.; Boushey, H. A.  (1979) Effect of
20             ozone on bronchial reactivity in atopic and nonatopic subjects. Am. Rev. Respir. Dis. 120: 1059-1067.
21       Holz, O.; Torres, R. A.; Timm, P.; Mticke, M.; Richter, K.; Koschyk, S.; Magnussen, H. (1999) Ozone-induced
22             airway inflammatory changes differ between individuals and are reproducible. Am. J. Respir. Crit. Care Med.
23             159:776-784.
24       Holz, O.; Mticke, M.; Paasch, K.; Bohme, S.; Timm, P.; Richter, K.; Magnussen, H.; Torres, R. A. (2002) Repeated
25             ozone exposures enhance bronchial allergen responses in subjects with rhinitis or asthma. Clin. Exp. Allergy.
26             32:681-689.
27       Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone concentration and
28             pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10,
29             and 0.12 ppm. Am. Rev.  Respir. Dis.  142: 1158-1163.
30       Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T.; Folinsbee, L. J. (1995) Comparison of pulmonary responses of
31             asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.
32             Toxicol. Ind. Health 11: 369-385.
33       Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979) Pulmonary function and maximum exercise responses
34             following acute ozone exposure. Aviat. Space Environ. Med. 50: 901-905.
35       Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of effect. Am. Rev. Respir. Dis.
36             123:496-499.
37       Horvath, S. M.; Bedi, J. F.; Drechsler-Parks, D. M. (1986) Effects of peroxyacetyl nitrate alone and in combination
3 8             with ozone in healthy young women.  J. Air Pollut. Control Assoc. 36: 265-270.
39       Horvath, S. M.; Folinsbee, L. J.; Bedi, J. F. (1987) Combined  effect of ozone and sulfuric acid on pulmonary
40             function in man. Am. Ind. Hyg. Assoc. J. 48: 94-98.
41       Horvath, S. M.; Bedi, J. F.; Drechsler-Parks, D. M.; Williams, R. E. (1991) Alterations in pulmonary function
42             parameters during exposure to 80 ppb ozone for 6.6 hours in healthy middle aged individuals. In: Berglund,
43             R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric  ozone and the environment: papers from an
44             international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
45             Association; pp. 59-70. (A&WMA transactions series: no. TR-19).
46       Housley, D. G.; Eccles, R.; Richards, R. J. (1996) Gender difference in the concentration of the antioxidant uric acid
47             in human nasal lavage. Acta Oto-Laryngol. 116:  751-754.
48       Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
49             respiratory flow. J. Appl. Physiol. 77: 574-583.
50       Jenkins, H. S.; Devalia, J. L.; Mister, R. L.; Bevan, A. M.; Rusznak, C.; Davies, R. J. (1999) The effect of exposure
51             to ozone and nitrogen dioxide on the airway response of atopic asthmatics to inhaled allergen: dose- and
52             time-dependent effects. Am. J. Respir. Crit. Care Med.  160: 33-39.
53       Joad, J. P.; Kott, K. S.; Brie, J. M. (1996)  The local C-fiber contribution to ozone-induced effects on the isolated
54             guinea pig lung. Toxicol. Appl. Pharmacol. 141: 561-567.
         January 2005                                 AX6-141      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Johnson, D. A.; Winters, R. S.; Woolley, T.; Graham, D.; Henderson, F. W. (1986) Ozone effects on
  2             alpha-1-proteinase inhibitor in vivo: blood plasma inhibitory activity is unchanged. Exp. Lung Res.
  3             11:95-103.
  4       Torres, R.; Nowak, D.; Magnussen, H.; Speckin, P.; Koschyk, S. (1996) The effect of ozone exposure on allergen
  5             responsiveness in subjects with asthma or rhinitis. Am. J. Respir. Crit. Care Med. 153: 56-64.
  6       Torres, R. A.; Holz, O.; Zachgo, W.;  Timm, P.; Koschyk, S.; Miiller, B.; Grimminger, F.; Seeger, W.; Kelly, F. J.;
  7             Dunster, C.; Frischer, T.; Lubec, G.; Waschewski, M; Niendorf, A.; Magnussen, H. (2000) The effect of
  8             repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
  9             Am. J. Respir. Crit. Care Med. 161: 1855-1861.
10       Kabel, J. R.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung:
11             comparison of nasal and oral quiet breathing. J. Appl. Physiol. 77: 2584-2592.
12       Kagawa, J. (1986) Experimental studies on human health effects of aerosol and gaseous pollutants.  In: Lee, S. D.;
13             Schneider, T.; Grant, L. D.; Verkerk, P. J., eds. Aerosols: research, risk assessment and control strategies -
14             proceedings of the second U.S.-Dutch international symposium; May 1985; Williamsburg, VA. Chelsea, MI:
15             Lewis Publishers, Inc.; pp. 683-697.
16       Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A.
17             (1987) Ozone exposure increases respiratory epithelial permeability in humans. Am. Rev.  Respir. Dis.
18             135:  1124-1128.
19       Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A.
20             (1989) Ozone-induced changes in the pulmonary clearance of 99mTc-DTPA in man. In: Schneider, T.; Lee,
21             S. D.; Wolters, G. J. R.; Grant, L. D.,  eds. Atmospheric ozone research and its policy implications:
22             proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
23             Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 343-351. (Studies in environmental science
24             35).
25       Kehrl, H. R.; Peden, D.  B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased specific  airway reactivity
26             of persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J. Allergy. Clin.
27             Immunol. 104: 1198-1204.
28       Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
29             alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
30       Koenig, J. Q.; Covert, D. S.; Marshall,  S. G.; Van Belle, G.; Pierson, W. E. (1987) The effects of ozone and nitrogen
31             dioxide on pulmonary function in healthy and in asthmatic adolescents. Am. Rev. Respir. Dis.
32             136:  1152-1157.
33       Koenig, J. Q.; Covert, D. S.; Smith, M. S.; Van Belle, G.; Pierson, W. E. (1988) The pulmonary effects  of ozone and
34             nitrogen dioxide alone and combined  in healthy and asthmatic adolescent subjects. Toxicol. Ind. Health
35             4:521-532.
36       Koenig, J. Q.; Covert, D. S.; Hanley, Q. S.; Van Belle, G.; Pierson,  W. E. (1990) Prior exposure to ozone potentiates
37             subsequent response to sulfur dioxide in adolescent asthmatic subjects. Am. Rev. Respir. Dis. 141: 377-380.
38       Koenig, J. Q.; Covert, D. S.; Pierson, W. E.; Hanley, Q. S.; Rebolledo, V; Dumler, K.; McKinney,  S. E. (1994)
39             Oxidant and acid aerosol exposure in  healthy subjects and subjects with asthma. Part I: effects of oxidants,
40             combined with sulfuric or nitric  acid,  on the pulmonary function of adolescents with asthma. Cambridge, MA:
41             Health Effects Institute; pp. 1-36; research report no. 70.
42       Konig, G.; Rommelt, H.; Kienle, H.; Dirnagl, K.; Polke, H.; Fruhmann, G. (1980) Anderung der
43             bronchomotorischen Reagibilitat des Menschen durch Einwirkung von Ozon [Changes in the bronchial
44             reactivity of humans caused by the influence of ozone]. Arbeitsmed. Sozialmed. Praeventivmed.
45             151:261-263.
46       Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.; Kozumbo,  W. J.; Becker, S.;
47             House, D. E.; McDonnell, W. F.; Bromberg, P. A. (1989a) Ozone-induced inflammation in the lower airways
48             of human subjects. Am. Rev. Respir. Dis. 139: 407-415.
49       Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The inflammatory response in
50             human lung exposed to ambient levels of ozone. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D.,
51             eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch international
52             symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science
53             Publishers; pp. 745-753. (Studies in environmental science 35).
54       Koren, H. S.; Hatch, G.  E.; Graham,  D. E. (1990) Nasal lavage as a tool in assessing acute inflammation in response
55             to inhaled pollutants. Toxicology 60:  15-25.
         January 2005                                 AX6-142       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent changes of markers
  2             associated with inflammation in the lungs of humans exposed to ambient levels of ozone. Toxicol. Pathol.
  3             19:406-411.
  4       Kozumbo, W. I; Hanley, N. M; Agarwas, S.; Thomas, M. I; Madden, M. C. (1996) Products of ozonized
  5             arachidonic acid potentiate the formation of DNA single strand breaks in cultured human lung cells. Environ.
  6             Mol. Mutagen. 27: 185-195.
  7       Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. I; D'Arcy, I; Eschenbacher, W. L. (1989) Ozone-induced
  8             changes in pulmonary function and bronchial responsiveness in asthmatics. J. Appl. Physiol. 66: 217-222.
  9       Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G.; Frew, A.; Polak, J.; Holgate, S.
10             (1997a) Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am. J.
11             Respir. Crit. Care Med. 156: 943-950.
12       Krishna, M. T.; Blomberg, A.; Biscione, G. L.; Kelly, F.; Sandstrom, T.; Frew, A.; Holgate, S. (1997b) Short-term
13             ozone exposure upregulates P-selectin in normal human airways. Am. J. Respir. Crit. Care Med.
14             155: 1798-1803.
15       Krishna, M. T.; Madden, J.; Teran, L. M.; Biscione, G. L.; Lau, L.  C. K.; Withers, N. J.; Sandstrom, T.; Mudway, L;
16             Kelly, F. J.; Walls, A.; Frew, A. J.; Holgate, S. T.  (1998) Effects of 0.2 ppm ozone onbiomarkers of
17             inflammation in bronchoalveolar lavage fluid and bronchial mucosa of healthy subjects. Eur. Respir. J.
18             11: 1294-1300
19       Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D. M. (1982) Duration of pulmonary
20             function adaptation to ozone in humans. Am. Ind.  Hyg. Assoc. J. 43: 832-837.
21       Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D.  (1985) Ozone response relationships in healthy
22             nonsmokers. Am. Rev. Respir. Dis. 132: 36-41.
23       Lang, D. S.; Torres, R. A.; Mticke, M.; Siegfried, W.; Magnussen, H. (1998) Interactions between human
24             bronchoepithelial cells and lung fibroblasts after ozone exposure in vitro. Toxicol. Lett. 96-97: 13-24.
25       Larsen, R. I.; McDonnell, W. F.; Horstman, D. H.; Folinsbee, L. J. (1991) An air quality data analysis system for
26             interrelating effects, standards, and needed source reductions: part 11. a lognormal  model relating human lung
27             function decrease to O3 exposure. J. Air Waste Manage. Assoc. 41: 455-459.
28       Lategola, M. T.; Melton, C. E.; Higgins, E. A. (1980) Effects of ozone on symptoms and cardiopulmonary function
29             in a flight attendant surrogate population. Aviat. Space Environ. Med. 51: 237-246.
30       Lategola, M. T.; Melton, C. E.; Higgins, E. A. (1980) Pulmonary and symptom threshold effects of ozone in airline
31             passenger and cockpit crew surrogates. Aviat. Space Environ. Med. 51: 878-884.
32
33       Lefohn, A. S.; Foley, J. K. (1993) Establishing relevant ozone standards to protect vegetation and human health:
34             exposure/dose-response considerations. Air Waste 43: 106-112.
35       Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U. (1988) Die Wirkung von Ozon auf die korperliche
36             Leistungsfa'higkeit [The effect of ozone on physical activity]. Schweiz Z. Sportmed. 36: 5-10.
37       Linn, W. S.; Buckley, R. D.; Spier, C. E.; Blessey, R. L.; Jones, M. P.; Fischer, D. A.; Hackney, J. D. (1978) Health
38             effects of ozone exposure in asthmatics. Am.  Rev. Respir. Dis. 117: 835-843.
39       Linn, W. S.; Jones, M. P.; Bachmayer, E. A.;  Spier,  C. E.; Mazur, S. F.; Avol, E. L.; Hackney, J. D. (1980)
40             Short-term respiratory effects of polluted ambient air: a laboratory study of volunteers in a high-oxidant
41             community. Am. Rev. Respir. Dis. 121: 243-252.
42       Linn, W. S.; Fischer, D. A.; Medway, D. A.; Anzar, U. T.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Hackney,
43             J. D. (1982) Short-term respiratory effects of 0.12 ppm ozone exposure in volunteers with chronic obstructive
44             pulmonary disease. Am. Rev. Respir. Dis. 125: 658-663.
45       Linn, W. S.; Avol, E. L.; Hackney, J. D. (1983) Effects of ambient oxidant pollutants on humans: a movable
46             environmental chamber study. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium
47             on the biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
48             Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 125-137. (Advances in modern environmental
49             toxicology: v. 5).
50       Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Fischer, D. A.; Hackney, J. D.
51             (1986) A dose-response study of healthy, heavily exercising men exposed to ozone at concentrations near the
52             ambient air quality standard. Toxicol. Ind. Health 2: 99-112.
53       Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Peng, R.-C.; Valencia, L. M.; Little, D. E.; Hackney, J. D. (1988) Repeated
54             laboratory ozone exposures of volunteer Los Angeles residents: an apparent seasonal variation in response.
55             Toxicol. Ind. Health 4: 505-520.
         January 2005                                 AX6-143       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D. (1994) Effects of prolonged,
  2             repeated exposure to ozone, sulfuric acid, and their combination in healthy and asthmatic volunteers. Am. J.
  3             Respir. Crit. Care Med. 150: 431-440.
  4       Linn, W. S.; Anderson, K. R.; Shamoo, D. A.; Edwards, S. A.; Webb, T. L.; Hackney, J. D.; Gong, H., Jr. (1995)
  5             Controlled exposures of young asthmatics to mixed oxidant gases and acid aerosol. Am. J. Respir. Crit. Care
  6             Med. 152: 885-891.
  7       Linn, W. S.; Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Avol, E. L. (1997) Chamber exposures of children to
  8             mixed ozone, sulfur dioxide, and sulfuric acid. Arch. Environ. Health 52: 179-187.
  9       Liu, L.; Leech, J. A.; Urch, R. B.; Silverman, F. S. (1997)  In vivo salicylate hyroxylation: a potential biomarker for
10             assessing acute ozone exposure and effects in humans. Am. J. Respir. Crit. Care Med. 156: 1405-1412.
11       Liu, L.; Leech, J. A.; Urch, R. B.; Poon, R.; Zimmerman, B.; Kubay, J. M.; Silverman, F. S. (1999) A comparison of
12             biomarkers  of ozone exposure in human plasma, nasal lavage, and sputum. Inhalation Toxicol. 11: 657-674.
13       MacNee, W.; Morrison, D.; Rahman, I.; Li, X. Y.; Donaldson, K. (1996) Cigarette smoke and  ozone-induced
14             epithelial perturbation in vivo and in vitro: the role of glutathione. Chest 109(suppl.): 39S.
15       Magnussen, H.; Nowak, D.; Jorres, R. (1994) Experimentelle Schadstoffeffekte bei Gesundenund Asthmatikern
16             [Experimental effects of air pollutants in healthy and asthmatic subjects]. Pneumologie 48: 85-88.
17       Marthan, R.; Roux, E.; Savineau, J.-P. (1996) Human bronchial smooth muscle responsiveness after in vitro
18             exposure to oxidizing pollutants. Cell Biol. Toxicol. 12: 245-249.
19       Martin, T. R.; Castile, R. G.; Fredberg, J. J.; Wohl, M. E. B.; Mead, J. (1987)  Airway size is related to sex but not
20             lung size in normal adults. J. Appl. Physiol. 63: 2042-2047.
21       McBride, D. E.;  Koenig, J. Q.; Luchtel, D. L.; Williams, P. V.; Henderson, W. R., Jr. (1994) Inflammatory effects of
22             ozone in the upper airways of subjects with asthma. Am. J. Respir. Crit. Care Med. 149: 1192-1197.
23       McDonnell, W. F.  (1996) Individual variability in human lung function responses to ozone exposure.  Environ.
24             Toxicol. Pharmacol. 2: 171-175.
25       McDonnell, W. F.; Smith, M. V. (1994) Description of acute ozone response as a function of exposure rate and total
26             inhaled dose. J. Appl. Physiol. 76: 2776-2784.
27       McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.;  House, D. E. (1983)
28             Pulmonary effects of ozone exposure during exercise: dose-response characteristics. J. Appl. Physiol.:  Respir.
29             Environ. Exercise Physiol. 54:  1345-1352.
30       McDonnell, W. F., Ill; Horstman, D. H.; Abdul-Salaam, S.; House, D. E. (1985a) Reproducibility of individual
31             responses to ozone exposure. Am. Rev. Respir. Dis. 131: 36-40.
32       McDonnell, W. F., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier,  A. M. (1985b) Respiratory responses
33             of vigorously exercising children to 0.12 ppm ozone exposure. Am. Rev. Respir. Dis. 132: 875-879.
34       McDonnell, W. F.; Horstman, D. H.; Abdul-Salaam, S.; Raggio, L. J.; Green,  J. A. (1987) The respiratory responses
35             of subjects with allergic rhinitis to ozone exposure and their relationship to nonspecific airway reactivity.
36             Toxicol. Ind. Health 3: 507-517.
37       McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.; Folinsbee, L. J.; Devlin, R. B.; O'Neil, J. J.;
38             Horstman, D. H. (1991) Respiratory response of humans exposed to low levels of ozone for 6.6 hours.  Arch.
39             Environ. Health 46: 145-150.
40       McDonnell, W. F.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual differences in acute
41             response to  ozone exposure. Am. Rev. Respir. Dis.  147: 818-825.
42       McDonnell, W. F.; Stewart, P. W.; Andreoni, S.; Seal, E.,  Jr.; Kehrl, H. R.; Horstman, D. H.; Folinsbee, L. J.; Smith,
43             M. V. (1997) Prediction of ozone-induced FEVj changes: effects of concentration, duration, and ventilation.
44             Am.  J. Respir. Crit. Care Med.  156: 715-722.
45       McDonnell, W. F.; Stewart, P. W.; Smith, M. V.; Pan, W.  K.; Pan, J. (1999) Ozone-induced respiratory symptoms:
46             exposure-response models and association with lung function. Eur. Respir. J. 14: 845-853.
47       McGee, M. P.; Devlin, R.; Saluta, G.; Koren, H. (1990) Tissue factor and factor VII messenger RNAs in human
48             alveolar macrophages: effects of breathing ozone. Blood 75: 122-127.
49       McKinnon, K. P.; Madden, M. C.; Noah, T. L.; Devlin, R. B. (1993) In vitro ozone exposure increases release of
50             arachidonic acid products from a human bronchial epithelial cell line. Toxicol. Appl. Pharmacol.
51             118:215-223.
52       Melton, C. E. (1990) Airliner cabin ozone: an updated review. Washington, DC: Federal Aviation Administration,
53             Office of Aviation Medicine; report no. DOT/FAA/AM-89/13. Available from: NTIS, Springfield, VA;
54             AD-A219 264.
55       Messineo, T. D.; Adams, W. C. (1990) Ozone inhalation effects in females varying widely in lung size: comparison
56             with males.  J. Appl. Physiol. 69: 96-103.
         January 2005                                 AX6-144       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Michelson, P. H.; Dailey, L.; Devlin, R. B.; Peden, D. B. (1999) Ozone effects on the immediate-phase response to
  2             allergen in the nasal airways of allergic asthmatic subjects. Otolaryngol. Head Neck Surg. 120: 225-232.
  3       Mohammed, S. P.; Higenbottam, T. W.; Adcock, J. J. (1993) Effects of aerosol-applied capsaicin, histamine and
  4             prostaglandin-E2 on airway sensory receptors of anaesthetized cats. J. Physiol. Lond. 46: 951-966.
  5       Molfino, N. A.; Wright, S. C.; Katz, L; Tarlo, S.; Silverman, F.; McClean, P. A.; Szalai, J. P.; Raizenne, M; Slutsky,
  6             A. S.; Zamel, N. (1991) Effect of low concentrations of ozone on inhaled allergen responses in asthmatic
  7             subjects. Lancet 338(8761): 199-203.
  8       Molfino, N. A.; Slutsky, A. S.; Zamel, N. (1992) The effects of air pollution on allergic bronchial responsiveness.
  9             Clin. Exp. Allergy 22: 667-672.
10       Montuschi, P.; Nightingale, J. A.; Kharitonov, S. A.; Barnes, P. J. (2002) Ozone-induced increase in exhaled
11             8-isoprostane in healthy subjects is resistant to inhaled budesonide. Free Radical Biol. Med. 33: 1403-1408.
12       Mudway, I. S.; Kelly, F. J. (2000) Ozone and the lung: a sensitive issue. Mol. Aspects. Med. 21: 1-48.
13       Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S. T.; Sandstrom, T.; Kelly, F. J. (1999) Antioxidant
14             consumption and repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to
15             ozone. Eur. Respir. J. 13: 1429-1438.
16       Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A. J.; Sandstrom, T.;
17             Kelly, F. J. (2001) Differences in basal airway antioxidant concentrations are not predictive of individual
18             responsiveness to ozone: a comparison of healthy and mild asthmatic subjects. Free Radical Biol. Med.
19             31:962-974.
20       Nagda, N. L.; Fortmann, R. C.; Koontz, M. D.; Baker, S. R.; Ginevan, M. E. (1989) Airliner cabin environment:
21             contaminant measurements, health risks, and mitigation options. Washington, DC: U.S. Department of
22             Transportation,  Office of the Secretary. Available from: NTIS, Springfield, VA; PB91 -1593 84.
23       National Heart, Lung, and Blood Institute (NHLBI). (1997) Guidelines for the diagnosis and management of asthma:
24             expert panel report 2. Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of
25             Health; publication no. 97-4051. Available: http://www.nhlbi.nih.gov/guidelines/asthma/asthgdln.pdf
26             (11 April 2003).
27       National Heart, Lung, and Blood Institute (NHLBI). (2003) Expert panel report: guidelines for the diagnosis and
28             management of asthma. Update on selected topics 2002. Bethesda, MD: U.S. Department of Health and
29             Human Services, National Institutes of Health; NIH publication no. 02-5074. Available:
30             http://www.nhlbi.nih.gov/guidelines/asthma/asthmafullrpt.pdf [26 August, 2004].
31       Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud, D. J.; Speizer, F. E. (1995) The association of ambient air
32             pollution with twice daily peak expiratory flow rate measurements in children. Am. J. Epidemiol.
33             141: 111-122.
34       Newson, E. J.; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000) Effects of
3 5             short-term exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum, exhaled nitric oxide, and
36             lung function in subjects with mild atopic asthma. J. Occup. Environ. Med. 42: 270-277.
37       Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
3 8             factor-KB  activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
39             60:356-362.
40       Nightingale, J. A.; Rogers, D. F.; Barnes, P. J. (1999) Effect of inhaled ozone on exhaled nitric oxide, pulmonary
41             function, and induced sputum in normal and asthmatic subjects. Thorax 54: 1061-1069.
42       Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P.  J. (2000) No effect of inhaled budesonide on the response
43             to inhaled ozone in normal subjects. Am.  J. Respir. Crit. Care Med. 161: 479-486.
44       Niinimaa, V.; Cole, P.; Mintz, S.; Shephard, R. J. (1980) The switching point from nasal to oronasal breathing.
45             Respir. Physiol. 42: 61-71.
46       Nikasinovic, L.; Momas, I.; Seta, N. (2003) Nasal epithelial and inflammatory response to ozone exposure: a review
47             of laboratory-based studies published since 1985. J. Toxicol. Environ. Health B 6: 521-568.
48       Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998) Nociceptive
49             mechanisms modulate ozone-induced human lung function decrements. J. Appl. Physiol. 85: 1863-1870.
50       Peden, D. B. (2001a) Air pollution in asthma: effect of pollutants on airway inflammation. Ann. Allergy Asthma
51             Immunol.  87(suppl. 3): 12-17.
52       Peden, D. B. (200 Ib) Effect of pollutants in rhinitis. Curr. Allergy Asthma Rep. 1: 242-246.
53       Peden, D. B.; Setzer, R. W., Jr.; Devlin, R. B. (1995) Ozone exposure has both a priming effect on allergen-induced
54             responses  and an intrinsic inflammatory action in the nasal airways of perennially allergic asthmatics. Am. J.
55             Respir. Crit. Care Med. 151: 1336-1345.
         January 2005                                 AX6-145      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997) Prolonged acute exposure to 0.16 ppm ozone induces
  2             eosinophilic airway inflammation in asthmatic subjects with allergies. J. Allergy Clin. Immunol.
  3             100:802-808.
  4       Peters, E. A.; Hiltermann, J. T.; Stolk, J. (2001) Effect of apocynin on ozone-induced airway hyperresponsiveness to
  5             methacholine in asthmatics. Free Radical Biol. Med. 31: 1442-1447.
  6       Rayman, R. B. (2002) Cabin air quality: an overview. Aviat. Space Environ. Med. 73: 211-215.
  7       Reisenauer, C. S.; Koenig, J. Q.; McManus, M. S.; Smith, M. S.; Kusic, G.; Pierson, W. E. (1988) Pulmonary
  8             response to ozone exposures in healthy individuals aged 55 years or greater. JAPCA 38: 51-55.
  9       Riediker, M.; Monn, C.; Koller, T.; Stahel, W. A.; Wuthrich, B. (2001) Air pollutants enhance rhinoconjunctivitis
10             symptoms in pollen-allergic individuals. Ann. Allergy Asthma Immunol. 87: 311-318.
11       Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
12             of nitrogen dioxide, sulfur dioxide,  and ozone exposures. Arch. Environ. Health 52: 173-178.
13       Rombout, P. J. A.; Dormans, J. A. M. A.; Marra, M.; VanEsch, G. J. (1986) Influence of exposure regimen on
14             nitrogen dioxide-induced morphological changes in the rat lung. Environ. Res. 41: 466-480.
15       Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del
16             Rio-Navarro, B.; Hernandez-Avila,  M.; London, S. J. (2004) Genetic polymorphism of GSTM1 and
17             antioxidant supplementation influence lung function in relation to ozone exposure in asthmatic children in
18             Mexico City. Thorax 59: 8-10.
19       Samet, J. M.; Hatch, G. E.; Horstman, D.;  Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell,
20             W. F.; Devlin, R. B. (2001) Effect of antioxidant supplementation on ozone-induced lung injury in human
21             subjects. Am. J. Respir.  Crit. Care Med. 164: 819-825.
22       Savin, W. M.; Adams, W. C. (1979) Effects of ozone inhalation on work performance and VO2max. J. Appl. Physiol.:
23             Respir. Environ. Exercise Physiol. 46: 309-314.
24       Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.;  Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996)
25             Greater ozone-induced inflammatory responses in subjects with asthma. Am. J. Respir. Crit. Care Med.
26             154:24-29.
27       Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise time in competitive simulations consequent to  low level
28             ozone exposure. Med. Sci. Sports Exercise 18: 408-414.
29       Schelegle, E. S.; Adams, W. C.; Siefkin, A. D. (1987) Indomethacin pretreatment reduces ozone-induced pulmonary
30             function decrements in human subjects. Am. Rev. Respir. Dis. 136: 1350-1354.
31       Schelegle, E. S.; Adams, W. C.; Giri, S. N.; Siefkin, A. D. (1989) Acute ozone exposure increases plasma
32             prostaglandin F2a in ozone-sensitive human subjects. Am. Rev. Respir. Dis. 140: 211-216.
33       Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced neutrophilia in normal
34             humans. Am. Rev. Respir. Dis. 143: 1353-1358.
35       Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G.; Green, J. F. (1993) Contribution of vagal
36             afferents to respiratory reflexes evoked by acute inhalation of ozone in dogs. J. Appl. Physiol. 74: 2338-2344.
37       Schelegle, E. S.; Eldridge, M. W.; Cross, C. E.; Walby, W. F.; Adams, W. C. (2001) Differential effects of airway
38             anesthesia on ozone-induced pulmonary responses in human subjects. Am. J. Respir. Crit. Care Med.
39             163: 1121-1127.
40       Schierhorn, K.; Zhang, M.; Matthias, C.; Kunkel, G.  (1999) Influence of ozone and nitrogen dioxide on histamine
41             and interleukin formation in a human nasal mucosa culture system. Am. J. Respir. Cell Mol. Biol.
42             20:  1013-1019.
43       Schierhorn, K.; Hanf, G.; Fischer, A.; Umland, B.; Olze, H.; Kunkel, G. (2002) Ozone-induced release of
44             neuropeptides from human nasal mucosa cells. Int. Arch. Allergy Immunol. 129: 145-151.
45       Schonfeld, B. R.; Adams, W. C.;  Schelegle, E. S. (1989) Duration of enhanced responsiveness upon re-exposure to
46             ozone. Arch. Environ. Health 44: 229-236.
47       Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S. (1976) Pulmonary responses of
48             rats to ambient levels of ozone: effects of 7-day intermittent or continuous exposure. Lab. Invest. 34: 565-578.
49       Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.; Green, J.; Raggio, L.
50             (1993) The pulmonary response of white and black adults to six concentrations of ozone. Am. Rev. Respir.
51             Dis. 147: 804-810.
52       Seal, E., Jr.; McDonnell, W. F.; House, D. E. (1996) Effects of age, socioeconomic status, and menstrual cycle on
53             pulmonary response to ozone. Arch. Environ.  Health 51: 132-137.
54       Sekizawa,  S. I.; Tsubone, H. (1994) Nasal receptors responding to noxious chemical irritants. Respir. Physiol.
55             96:37-48.
         January 2005                                AX6-146       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Seltzer, I; Bigby, B. G.; Stulbarg, M; Holtzman, M. I; Nadel, J. A.; Ueki, I. F.; Leikauf, G. D.; Goetzl, E. J.;
  2             Boushey, H. A. (1986) O3-induced change in bronchial reactivity to methacholine and airway inflammation in
  3             humans. J. Appl. Physiol. 60: 1321-1326.
  4       Solway, J.; Leff, A. R. (1991) Sensory neuropeptides and airway function. J. Appl. Physiol. 71: 2077-2087.
  5       Spannhake, E. W.; Reddy, S. P. M.; Jacoby, D. B.; Yu, X.-Y.; Saatian, B.; Tian, J. (2002) Synergismbetween
  6             rhinovirus infection and oxidant pollutant exposure enhances airway epithelial cell cytokine production.
  7             Environ. Health Perspect. 110: 665-670.
  8       Spit, B. J.; Bretschneider, F.; Hendriksen, E. G. J.; Kuper, C. F. (1993) Ultrastructure of free nerve endings in
  9             respiratory and squamous epithelium on the rat nasal septum. Cell Tissue Res. 274: 329-335.
10       Steck-Scott, S.; Arab, L.; Craft, N. E.; Samet, J. M. (2004) Plasma and lung macrophage responsiveness to
11             carotenoid supplementation and ozone exposure in humans. Eur. J. Clin. Nutr. 1-9.
12       Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, L; Helleday, R.; Kelly, F. J.; Frew, A.  J.;
13             Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy  subjects.
14             Respir. Med. 96: 352-358.
15       Sterk, P. J.; Fabbri, L. M.; Quanjer, P. H.; Cockcroft, D. W.; O'Byrne, P. M.; Anderson, S. D.;  Juniper, E.  F.; Malo,
16             J.-L. (1993) Airway responsiveness. Standardized challenge testing with pharmacological, physical and
17             sensitizing stimuli in adults. Report Working Party Standardization of Lung Function Tests, European
18             Community for Steel and Coal.  Eur. Respir. J. 6(suppl. 16): 53-83.
19       Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M.  J.; Ellis, R.; Dahlback, M.; O'Byrne, P. M. (1994) Effect of
20             inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs.
21             J. Appl. Physiol. 77: 2578-2583.
22       Superko, H. R.; Adams, W. C.; Daly, P. W. (1984) Effects of ozone inhalation during exercise  in selected patients
23             with heart disease. Am. J. Med. 77: 463-470.
24       Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. F.; Hatch, G. E. (1989) Unattenuated structural and
25             biochemical alterations in the rat lung during functional adaptation to ozone. Am. Rev. Respir. Dis.
26             140:493-501.
27       Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M.  G. (1990) Measurements of cardiopulmonary response in
28             awake rats during acute exposure to near-ambient concentrations of ozone. Fundam. Appl. Toxicol. 10: 7-15.
29       Torres, A.; Utell, M. J.; Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R.  J.; Speers, D. M.; Tsai, Y.;
30             Frampton, M. W. (1997) Airway inflammation in smokers and nonsmokers with varying responsiveness to
31             ozone. Am. J. Respir. Crit. Care Med. 156: 728-736.
32       Trenga, C. A.; Koenig, J. Q.; Williams, P. V. (2001) Dietary antioxidants and ozone-induced bronchial
33             hyperresponsiveness in adults with asthma. Arch. Environ. Health 56: 242-249.
34       U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other  photochemical oxidants.
3 5             Research Triangle Park, NC:  Office of Health and Environmental Assessment, Environmental Criteria and
36             Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from:  NTIS, Springfield, VA;
37             PB87-142949.
38       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
3 9             Research Triangle Park, NC:  Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
40             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
41             www.epa.gov/ncea/ozone.htm.
42       Ultman, J. S.; Ben-Jebria, A.; Arnold,  S. F. (2004) Uptake distribution of ozone  in human lungs: intersubject
43             variability in physiologic response. Boston, MA: Health Effects Institute.
44       Utell, M. J.; Frampton, M. W.; Morrow, P. E.; Cox, C.; Levy, P. C.; Speers, D. M.; Gibb, F. R. (1994) Oxidant and
45             acid aerosol exposure in healthy subjects and subjects with asthma. Part II: effects of sequential sulfuric acid
46             and ozone exposures on the pulmonary function of healthy subjects and subjects with asthma. Cambridge,
47             MA: Health Effects Institute; pp. 37-93; research report no. 70.
48       Vagaggini, B.; Carnevali, S.; Macchioni, P.; Taccola, M.; Fornai, E.; Bacci, E.; Bartoli, M. L.;  Cianchetti, S.; Dente,
49             F. L.; Di Franco, A.; Giannini, D.; Paggiaro. P. L. (1999) Airway inflammatory  response to ozone in subjects
50             with different asthma severity. Eur. Respir. J. 13: 274-280.
51       Vagaggini, B.; Taccola, M.;  Conti, I.; Carnevali, S.; Cianchetti, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.;
52             Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2001) Budesonide reduces neutrophilic but not functional airway
53             response to ozone in mild asthmatics. Am. J. Respir. Crit. Care Med. 164: 2172-2176.
54       Vagaggini, B.; Taccola, M.;  Cianchetti, S.; Carnevali, S.;  Bartoli, M. L.; Bacci, E.; Dente, F. L.; Di Franco, A.;
55             Giannini, D.; Paggiaro, P. L. (2002) Ozone exposure increases eosinophilic airway response induced by
56             previous allergen challenge. Am. J. Respir. Crit. Care Med. 166: 1073-1077.
         January 2005                                 AX6-147      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Van Bree, L.; Rombout, P. J. A.; Rietjens, I. M. C. M; Dormans, J. A. M. A.; Marra, M. (1989) Pathobiochemical
  2             effects in rat lung related to episodic ozone exposure. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant,
  3             L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
  4             international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
  5             Science Publishers; pp. 723-732. (Studies in environmental science 35).
  6       Vender, R. L.; Horstman, D. H.; Mangione, S. (1994) Red blood cell antioxidants in human volunteers exposed to
  7             ozone. Toxicol. Ind. Health 10: 53-58.
  8       Voter, K. Z.; Whitin, J. C.; Torres, A.; Morrow, P. E.; Cox, C.; Tsai, Y.; Utell, M. J.; Frampton, M. W. (2001)
  9             Ozone exposure and the production of reactive oxygen species by bronchoalveolar cells in humans. Inhalation
10             Toxicol. 13:465-483.
11       Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
12             ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect.
13             110:79-84.
14       Wayne,  W.  S.; Wehrle, P. F.; Carroll, R. E. (1967) Oxidant air pollution and athletic performance. JAMA J. Am.
15             Med. Assoc. 199: 151-154.
16       Weinmann, G. G.; Weidenbach-Gerbase, M.;  Foster, W. M.; Zacur, H.;  Frank, R. (1995) Evidence for
17             ozone-induced small-airway dysfunction: lack of menstrual-cycle and gender effects. Am. J. Respir. Crit.
18             Care Med. 152: 988-996.
19       Weymer, A. R.; Gong, H., Jr.; Lyness, A.; Linn, W. S. (1994) Pre-exposure to ozone does not enhance orproduce
20             exercise-induced asthma. Am. J. Respir. Crit.  Care Med. 149: 1413-1419.
21       Yeadon, M.; Wilkinson, D.; Darley-Usmar, V.; O'Leary, V. J.; Payne, A. N. (1992) Mechanisms contributing to
22             ozone-induced bronchial hyperreactivity in guinea-pigs. Pulm. Pharmacol. 5: 39-50.
23       Ying, R. L.; Gross, K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990) Indomethacin does not inhibit the
24             ozone-induced increase in bronchial responsiveness in human subjects. Am. Rev. Respir. Dis. 142: 817-821.
25       Yu, M.;  Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged  and diluted sidestream cigarette smoke
26             enhances ozone-induced lung injury inB6C3Fl mice. Toxicol. Sci. 65: 99-106.
27       Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R.  (1995) Differential susceptibility to ozone-induced airways
28             hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
29
         January 2005                                AX6-148       DRAFT-DO NOT QUOTE OR CITE

-------
 i                   7.  EPIDEMIOLOGICAL STUDIES OF
 2               HUMAN HEALTH EFFECTS ASSOCIATED
 3                   WITH AMBIENT OZONE EXPOSURE
 4
 5
 6     7.1  INTRODUCTION
 7          This chapter evaluates current epidemiologic literature on health and physiological effects
 8     of ambient O3 exposure. Epidemiologic studies linking community ambient O3 concentrations to
 9     health effects were reported in the 1996 Ozone Air Quality Criteria Document (O3 AQCD; U.S.
10     Environmental Protection Agency, 1996a). Many of those studies reported that pulmonary
11     function decrements, hospital and emergency department admissions, and respiratory symptoms
12     in human populations were associated with ambient levels of O3.  Numerous more recent
13     epidemiologic studies discussed in this chapter evaluate the relationship of ambient O3 to
14     morbidity and mortality, and thereby provide an expanded basis for assessment of health effects
15     associated with exposures to O3 at concentrations currently encountered in the U.S.
16          As discussed elsewhere in this document (Chapters 5 and 6), a substantial amount of
17     experimental evidence links O3 exposure unequivocally with respiratory effects in laboratory
18     animals and humans. These include structural changes in the bronchiolar-alveolar transition
19     (centriacinar) region of the lung, biochemical evidence of acute cellular/tissue injury,
20     inflammation, increased frequency and severity of experimental bacterial infection, and
21     temporary reductions in mechanical lung function. These effects have been observed with
22     exposure to O3 at ambient or near-ambient concentrations. Thus, many of the reported
23     epidemiologic associations of ambient O3 with respiratory health effects have considerable
24     biological credibility. Accordingly, the new epidemiologic studies of ambient O3 assessed here
25     are best considered in combination with information from the other chapters on ambient O3
26     concentration and exposure (Chapter 3), and toxicological effects of O3 in animals and humans
27     (Chapters 5 and 6, respectively). The epidemiologic studies constitute important information on
28     associations between health effects and exposures of human populations to "real-world" O3 and
29     also help to identify susceptible subgroups and associated risk factors.
30
       January 2005                             7-1        DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.1.1   Approach to Identifying O3 Epidemiologic Studies
 2          Numerous O3 epidemiologic papers have been published since completion of the 1996 O3
 3      AQCD. The U.S. Environmental Protection Agency (NCEA-RTP) has implemented a
 4      systematic approach to identify relevant epidemiologic studies for consideration in this chapter.
 5      In general, an ongoing search has been employed in conjunction with other strategies to identify
 6      O3 epidemiology literature pertinent to developing criteria for O3 National Ambient Air Quality
 7      Standards (NAAQS).  A publication base was established using Medline, Pascal, BIOSIS, NTIS,
 8      and Embase, and a set of search terms proven by prior use to identify pertinent literature.  The
 9      search strategy was reexamined and modified to enhance identification of published papers.
10      PubMed was added to the search regime.
11          While the above search regime provided good coverage of the relevant literature,
12      additional approaches augmented the traditional search methods.  First, a Federal Register
13      Notice was issued requesting information and published papers from the public at large. Next,
14      non-EPA chapter authors, expert in this field, identified literature on their own. NCEA-RTP
15      staff also identified publications as an element of their assessment and interpretation of the
16      literature.  Finally, additional potentially relevant publications will be included following
17      external review as a result of comments from both the public and CAS AC.  The combination of
18      these approaches is believed to produce a comprehensive collection of studies appropriate for
19      review and assessment here. The principal objective criteria used for selecting literature for the
20      present assessment is to include all identified studies that evaluated the relationship between
21      measured ambient O3 levels and a human health outcome. All new studies published through
22      October 2004, as identified using the search, have been included in this AQCD and additional
23      efforts have been made to assess more recent studies.
24
25      7.1.2   Approach to Assessing Epidemiologic Evidence
26          Definitions of the various types of epidemiologic studies assessed have been provided in an
27      earlier PM AQCD (U.S.  Environmental Protection Agency, 1996b). Briefly, epidemiologic
28      studies are generally divided into two groups, morbidity studies and mortality studies.  Morbidity
29      studies evaluate O3 effects  on a wide range of health endpoints, including the following:
30      changes in pulmonary function, respiratory symptoms, and self-medication in asthmatics;
31      respiratory-  and cardiovascular-related emergency department visits and hospital admissions;

        January 2005                              7-2        DRAFT-DO NOT QUOTE OR CITE

-------
 1      and changes in cardiovascular physiology/functions and airway inflammation.  Mortality studies
 2      investigate O3 effects on total (nonaccidental) mortality and cause-specific mortality, providing
 3      evidence related to a clearly adverse endpoint.  The epidemiologic strategies most commonly
 4      used in O3 health studies are of four types: (1) ecologic studies; (2) time-series semi-ecologic
 5      studies;  (3) prospective cohort studies; and (4) case-control and crossover studies. All of these
 6      are observational studies rather than experimental studies.
 7           The approach to assessing epidemiologic  evidence has been eloquently stated most
 8      recently in the 2004 PM AQCD (U.S. Environmental Protection Agency, 2004a) and is adapted
 9      here. The critical assessment of epidemiologic evidence presented in this chapter is conceptually
10      based upon consideration of salient aspects of the evidence of associations so as to reach
11      fundamental judgments as to the likely causal significance of the observed associations.  In so
12      doing, it is appropriate to draw from those aspects initially presented in Hill's classic monograph
13      (Hill, 1965) and widely used by the scientific community in conducting such evidence-based
14      reviews. A number of these aspects are judged to be particularly salient in evaluating the body
15      of evidence available in this review, including the aspects described by Hill as strength,
16      experiment, consistency, plausibility, and coherence.  Other aspects identified by Hill, including
17      temporality and biological gradient, are also relevant and considered here (e.g., in characterizing
18      lag structures and concentration-response relationships), but are more directly addressed in the
19      design and analyses of the individual epidemiologic studies included in this assessment.
20      As discussed below, these salient aspects are interrelated and considered throughout the
21      evaluation of the epidemiologic evidence presented in this chapter, and are more generally
22      reflected in the integrative synthesis presented in Chapter 8 of this AQCD.
23           In the following sections, the general evaluation of the strength of the epidemiological
24      evidence reflects consideration not only of the magnitude of reported O3 effect estimates and
25      their statistical significance, but also of the precision of the effect estimates and the robustness of
26      the effects associations.  Consideration of the robustness of the associations takes into account a
27      number of factors, including in particular the impact of alternative models and model
28      specifications and potential confounding by copollutants, as well as issues related to the
29      consequences of measurement error.
30           Consideration of the consistency of the effects associations, as discussed in the following
31      sections, involves looking across the results of multi- and single-city studies conducted by

        January 2005                               7-3         DRAFT-DO NOT  QUOTE OR CITE

-------
 1      different investigators in different places and times. Relevant factors are known to exhibit much
 2      variation across studies, including, for example, the presence and levels of copollutants, the
 3      relationships between central measures of O3 and exposure-related factors, relevant demographic
 4      factors related to sensitive subpopulations, and climatic and meteorological conditions. Thus, in
 5      this case, consideration of consistency and the related heterogeneity of effects are appropriately
 6      understood as an evaluation of the similarity or general concordance of results, rather than an
 7      expectation of finding quantitative results within a very narrow range.
 8           Looking beyond the epidemiological evidence,  evaluation of the biological plausibility of
 9      the O3-health effects associations observed in epidemiologic studies reflects consideration of
10      both exposure-related factors and dosimetric/toxicologic evidence relevant to identification of
11      potential biological mechanisms.  Similarly, coherence of health effects associations reported in
12      the epidemiologic literature reflects consideration of information pertaining to the nature of the
13      various respiratory- and cardiac-related mortality and morbidity effects and biological markers
14      evaluated in toxicologic and human clinical studies. These broader  aspects of the assessment are
15      only touched upon in this chapter but are more fully integrated in the discussion presented in
16      Chapters.
17           In identifying these aspects as being particularly salient in this assessment, it is also
18      important to recognize that no one aspect is either necessary or sufficient for drawing inferences
19      of causality. As Hill (1965) emphasized:
20
21              None of my nine viewpoints can bring indisputable evidence for or against the cause-
22               and-effect hypothesis and none can be required  as a sine qua non.  What they can do,
23              with greater or less strength, is to help us to make up our minds  on the fundamental
24               question — is there any other way of explaining the set of facts before us, is there
25               any other answer equally, or more, likely than cause and effect?
26
27      Thus, while these aspects frame considerations weighed in assessing the epidemiologic evidence,
28      they do not lend themselves to being considered in terms of simple formulas or hard-and-fast
29      rules of evidence leading to answers about causality (Hill, 1965).  One, for example, cannot
30      simply count up the numbers of studies reporting statistically significant results for the various
31      O3 indicators and health endpoints evaluated in this assessment and reach credible conclusions
32      about the relative strength of the evidence and the likelihood of causality. Rather, these
33      important considerations are taken into account and discussed throughout this assessment with

        January 2005                               7-4        DRAFT-DO NOT QUOTE OR CITE

-------
 1      the goal of producing an objective appraisal of the evidence, informed by peer and public
 2      comment and advice, including weighing of alternative views on controversial issues, leading to
 3      conclusions and inferences that reflect the best judgements of the scientists engaged in this
 4      review.
 5           In assessing the relative scientific quality of epidemiologic studies reviewed here and to
 6      assist in interpreting their findings, the following considerations were taken into account:

 7          (1)  To what extent are the aerometric data/exposure metrics used of adequate quality and
                sufficiently representative to serve as credible exposure indicators, well-reflecting
                geographic or temporal differences in study population pollutant exposures in the
                range(s) of pollutant concentrations evaluated?
 8          (2)  Were the study populations well-defined and adequately selected so as to allow for
                meaningful comparisons between study groups or meaningful temporal analyses of
                health effects results?
 9          (3)  Were the health endpoint measurements meaningful and reliable, including clear
                definition of diagnostic criteria utilized and consistency in obtaining dependent
                variable measurements?
10          (4)  Were the statistical analyses used appropriate, and properly performed and interpreted,
                including accurate data handling and transfer during analyses?
11          (5)  Were likely important covariates (e.g., potential confounders or effect modifiers)
                adequately controlled for or taken into account in the study design and statistical
                analyses?
12          (6)  Were the reported findings internally consistent, biologically plausible, and coherent
                in terms of consistency with other known facts?
13           These guidelines provide benchmarks for judging the relative quality of various studies and
14      in assessing the body of epidemiologic evidence. Detailed critical  analysis of all epidemiologic
15      studies on O3 health effects, especially in relation to all of the above questions, is beyond the
16      scope of this document. Of most importance for present purposes are those studies which
17      provide useful qualitative or quantitative information on exposure-response relationships for
18      health effects associated with ambient air levels of O3 likely to be encountered in the U. S. among
19      healthy and susceptible populations.
20
21
        January 2005                               7-5         DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.1.3  Study Designs and Analysis Methods Used to Assess O3 Health Effects
 2           Prior to discussing results from the recent O3 studies, issues and questions arising from the
 3      study designs and analysis methods used in the assessment of O3 effect estimates will be briefly
 4      presented. Air pollution time-series studies in particular have design and analysis aspects that
 5      complicate the interpretation of O3 health effects. Analyses using administrative data (e.g.,
 6      numbers of deaths and emergency hospital admissions) have inherent limitations as well as
 7      strengths (Virnig and McBean, 2001), however in this section we focus mainly on the topics of
 8      exposure assessment and model specification in time-series or longitudinal studies.  Potential
 9      biases that may result from O3 exposure measurement error, and choice of exposure index and
10      lag period are first presented. A discussion of model specification issues and potential
11      confounding by temporal factors, meteorological effects, seasonal trends, and copollutants
12      follow. Integrative discussion of these topics is presented later in Section 7.6.
13
14      7.1.3.1  Exposure Assessment in Epidemiologic Studies
15           In general, the exposure of the participant is not directly observed, and the concentration of
16      O3 and other air pollutants at one or more stationary air monitors is used as a proxy for
17      individual exposure to ambient air pollution.  In an ideal situation, studies of air pollution health
18      effects would be conducted at the individual level, with  information on personal exposure to the
19      various pollutants. However, determining accurate personal exposure information is difficult
20      and often impractical.  In many epidemiologic studies, especially time-series studies with
21      administrative data on mortality and hospitalization outcomes, data from central ambient
22      monitoring sites often are used as the estimate of exposure. Routinely collected ambient data,
23      though readily available and convenient, may not represent true personal exposure.  The use of
24      ambient data tends to underestimate the effect of the air pollutant on health (Krzyzanowski,
25      1997).  As discussed thoroughly in the 2004 PM AQCD (Section 8.4.5), the resulting exposure
26      measurement error and its effect on the estimates of relative risk must be considered. In theory,
27      there are three components to exposure measurement error: (1) the use of average population
28      rather than individual exposure data;  (2) the difference between the average personal ambient
29      exposure and the ambient concentrations; and (3) the difference between the true and measured
30      ambient concentrations. Zeger et al. (2000) indicated that the first and third error components
31      are largely Berksonian errors and would not significantly bias the risk estimate. However, the

        January 2005                              7-6        DRAFT-DO NOT QUOTE OR CITE

-------
 1      second error component resulting from the difference between the average personal ambient
 2      exposure and ambient concentration levels might introduce bias, especially if indoor sources are
 3      associated with ambient levels.
 4           Several studies measured O3 concentrations in a variety of indoor environments, including
 5      homes (Lee et al., 2004), schools (Linn et al., 1996),  and the workplace (Liu et al., 1995).
 6      Indoor O3 concentrations were, in general, approximately one-tenth of the outdoor
 7      concentrations in these studies.  However, the specific contribution of indoor sources to indoor
 8      O3 levels has not been investigated.  Few indoor sources of O3 exist, possible sources being
 9      office equipment (e.g., photocopiers, laser printers) and air cleaners. As described in Section 3.9
10      of this document, indoor O3 exposure primarily results from infiltration of O3 from the outdoors
11      through ventilation and is noted as minimal.
12           The impact of measurement error on O3 effect estimates was demonstrated in a study by
13      Navidi et al. (1999).  In this study, a simulation was conducted using data from the University of
14      Southern California Children's Health Study of the long-term effects of air pollutants on
15      children.  The effect estimate from computed "true" O3 exposure was compared to effect
16      estimates from exposure determined using several  methods:  (1) ambient stationary monitors;
17      (2) the microenvironmental approach (multiply concentrations in various microenvironments by
18      time present in each microenvironment); and (3) personal sampling. Effect estimates based on
19      all three exposure measures were biased  towards the  null.  The bias that results when using the
20      microenvironmental and personal sampling approach is due to nondifferential measurement
21      error.  Use of ambient monitors to determine exposure will tend to overestimate true personal O3
22      exposure (assumes that subjects are outdoors 100% of their time), thus generally their use will
23      result in effect estimates that are biased towards the null.
24
25      7.1.3.2  O3 Exposure Indices Used
26           The results of studies of mortality and morbidity health outcomes from exposure to O3 are
27      usually presented in this document as a relative risk,  or risk rate relative to a baseline mortality
28      or morbidity rate.  These relative risks are based on an incremental change in exposure. To
29      enhance comparability between studies, presenting these relative risks by  a uniform exposure
30      increment is needed. However, determining a standard increment is complicated by the use of
31      different O3 exposure indices in the existing health studies. The three daily O3 exposure indices

        January 2005                               7-7        DRAFT-DO NOT QUOTE OR CITE

-------
 1      that most often appear in the literature are 1-h average maximum (1-h max), 8-h average
 2      maximum (8-h max), and 24-h average (24-h avg). As concentrations are lower and less
 3      variable for the longer averaging times, relative risks of adverse health outcomes for a specific
 4      numeric concentration range are not directly comparable across metrics. Using the nationwide
 5      distributional data for O3 monitors in U.S. Metropolitan Statistical Areas, increments
 6      representative of a low-to-high change in O3 concentrations were developed based on mean
 7      and upper percentile values in the dataset (Langstaff, 2003):
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
                   Daily Exposure Index     Exposure Increment (ppb)
                        1-h max O3
                        8-h max O3
                        24-h avg O3
40
30
20
In the following discussion sections, efforts were made to standardize the O3 excess risks using
these increments, except as noted, so that the risk estimates could be compared across studies.

7.1.3.3  Lag of O3 Exposure Used
     Lags of exposure may reflect the distribution of effects across time in a population and the
potential mechanisms of effects. However, simply choosing the most significant exposure lag
may bias the air pollution risk estimates away from the null, as shown by a simulation by
Lumley and Sheppard (2000) that used PM2 5 as an example. This is especially true when the
choice is made from a large number of lags. Most of the O3 time-series studies examined
relatively small numbers of lagged days, typically 0 through 3 days, and/or cumulative lags
thereof (e.g., cumulative lag of 0 and 1 day).  An examination of the "most significant" lags
suggests that the majority of the single-day associations were immediate (0-day lag), not a
random pattern in which associations could be observed on any of the lags examined with equal
probabilities.  However, the lags may vary by health outcome, as some effects are delayed (e.g.,
airway inflammation) and are captured in longer lag periods. Note that when associations are
found at multiple days, presenting selected risk estimates from single-day lags may result in bias
(i.e., typically toward underestimation of magnitude of overall risk).
        January 2005
                                                     DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.1.3.4  Model Specification Issues
 2           The relationships between daily numbers of deaths and hospital admissions, and levels of
 3      O3 and related environmental factors have been analyzed widely over the past decade, yielding
 4      insights into the possible effects of O3 on acute exacerbations of respiratory and cardiovascular
 5      diseases, and related mortality. These daily time-series studies exploit the high degree of day-to-
 6      day variability in ambient air pollution concentrations to develop quantitative estimates of
 7      impacts on daily health outcomes. The basic analytical approach used to estimate the effects of
 8      O3 in this type of study is multiple regression. Because a given location is followed over time,
 9      many factors that might confound a multicity cross-sectional study do not affect time-series
10      studies. Cross-sectional confounders include cigarette smoking, diet, occupation, and other risk
11      factors that may vary across cities in ways that correlate with variations in air pollution levels.
12      In contrast, these factors are unlikely to vary over time in a way that correlates with day-to-day
13      variations in air pollution, thus confounding by these factors is minimized in a time-series study.
14      Longer-term secular time trends, such as changes in morbidity due to improved clinical
15      management of disease, also generally do not present a confounder problem in time-series
16      studies because these trends are removed analytically. Other advantages of the daily time-series
17      study design include the relatively large sample sizes in terms of person-days and the readily
18      available data, making such studies convenient and economical to conduct in a wide variety of
19      locations.
20           However, several challenges present themselves with respect to designing and interpreting
21      time-series studies. The principal challenge facing the analyst in the daily time-series context is
22      avoiding bias due to confounding by short-term temporal factors operating over time scales from
23      days to seasons. In the current regression models used to estimate short-term effects of air
24      pollution, there are two major potential confounders that need to be considered:  (1) seasonal
25      trend and other "long-wave" temporal trends; and (2) weather effects.  Both of these variables
26      tend to predict a significant fraction of fluctuations in time-series. Unfortunately, both of these
27      terms are also highly correlated with O3, as O3 has strong seasonal cycles and is formed more at
28      higher temperatures.  The correlation of O3 with these confounding terms tends to be higher than
29      that for PM or other gaseous pollutants. In the U.S., the mass concentration of PM25 generally
30      does not have strong seasonal cycles like O3 because PM2 5 tends to reflect both primary
31      emissions (throughout the year, but often higher in winter in most U.S. cities) and secondary

        January 2005                              7-9        DRAFT-DO NOT QUOTE OR CITE

-------
 1      aerosols (higher in summer).  Therefore, PM2 5 and O3 effect estimates from studies primarily
 2      designed to examine PM2 5 health effects may not be comparable as model specifications that
 3      may be appropriate for PM2 5 may not necessarily be adequate for O3.  The following section
 4      reviews the current methodologies used to control for potential confounding by temporal trends
 5      and weather effects.
 6
 7      7.1.3.5  Controlling for Temporal Trends and Meteorologic Effects
 8           An examination of recent time-series studies indicates that several types of fitting
 9      approaches have been used to adjust for temporal trends and weather effects. The use of
10      parametric and nonparametric smoothers with varying degrees of freedom per year has emerged
11      as the prevailing approach. The use of larger degrees of freedom to adjust for potential
12      confounding by time-varying factors may inadvertently result in ascribing more effects to these
13      unmeasured potential confounders and  take away the air pollution effect.  Often smaller
14      pollution effect estimates are observed  when more degrees of freedom are used. Currently, the
15      degrees of freedom used to adjust for temporal trends in time-series studies generally range from
16      4 to 12 degrees of freedom per year using either nonparametric or parametric smoothers.
17      Statistical diagnostics such as Akaike's Information Criteria,  residual autocorrelation, or
18      dispersion of the regression model often are used to choose or evaluate the adequacy of the
19      degrees of freedom for temporal trend,  but these diagnostics do not provide epidemiological
20      justification or interpretation of the fitted model.
21           The issue of model specifications to adjust for temporal trends and weather variables in
22      time-series studies was a consideration of several researchers that conducted sensitivity analyses
23      of PM data (HEI, 2003). The sensitivity of O3 coefficients to model specifications for temporal
24      trend adjustment has not been as well-studied.  Only one recent multicity study examined the
25      sensitivity of O3 coefficients to the extent of smoothing for adjustment of temporal trends and
26      meteorologic factors (Bell et al., 2004). Most, if not all, O3 studies used the same  model
27      specifications to estimate the excess risks for PM and other gaseous pollutants.  The relationship
28      between a pollutant and the temporal trend or weather effect being fitted differs for each
29      pollutant, and interpretation of the excess risk estimates needs to take into consideration this
30      varying concurvity (nonlinear analogue of multiple correlation) across pollutants.  As noted
31      above, O3 is expected to have the strongest correlation with both temporal (seasonal) trend and

        January 2005                              7-10        DRAFT-DO NOT QUOTE OR CITE

-------
 1      weather effects. The strong annual cycle in O3 concentrations presents a unique problem in
 2      time-series analyses where time trends are fitted simultaneously with pollution and other model
 3      terms (i.e., co-adjustment). In this setting, the annual O3 cycle itself may compete with the
 4      smooth function of time to explain some of the annual, cyclic behavior in the health outcome,
 5      which can result in biased effect estimates for O3 when data for all seasons are analyzed
 6      together.
 7           Current weather models used in time-series analyses can be classified into:  (1) quantile
 8      (e.g., quartile, quintile) indicators; (2) parametric functional forms such as V- or U-shape
 9      functions; and (3) parametric (e.g., natural splines) or nonparametric (e.g., locally estimated
10      smoothing splines [LOESS]) smoothing functions. More recent studies tend to use smoothing
11      functions. While these methods provide flexible ways to fit health outcomes as a function of
12      temperature and other weather variables, there are two major issues that need further
13      examination to enable more meaningful interpretation of O3 morbidity and mortality effects.
14           The first issue is the interpretation of weather or temperature effects.  Most researchers
15      agree about the morbidity and mortality effects of extreme temperatures (i.e., heat waves or cold
16      spells). However, as extreme hot or cold temperatures, by definition, happen rarely, much of the
17      health effects occur in the mild or moderate temperature range.  Given the significant correlation
18      between O3 and temperature, ascribing the association between temperature and health outcomes
19      solely to temperature effects may underestimate the effect of O3.
20           The second issue is that in most studies weather model specifications are fitted for year-
21      round data.  Such models will ignore the correlation structure that can change across seasons,
22      resulting  in inefficiency and model mis-specification. This is particularly important for O3,
23      which appears to change its relationship with temperature as well as with other pollutants across
24      seasons.  Ambient O3 levels are typically higher in the summer or warm season, often referred to
25      as the O3  season.  In the winter or colder months, O3 levels tend to be much lower compared to
26      the summer months. During the winter in some urban locations, O3 mainly comes from the free
27      troposphere and can be considered a tracer for relatively clean air (i.e., cold, clear air coming
28      down from the upper atmosphere), as discussed in Chapter 3 of this AQCD. The clean air is
29      associated with the passage of cold fronts and the onset of high-pressure conditions, which occur
30      with colder temperatures.  Thus, sunny clear winter days following a high-pressure system are
31      the days when air pollution levels from primary emissions (e.g., NO2, SO2, and PM from local

        January 2005                              7-11         DRAFT-DO NOT QUOTE OR CITE

-------
 1      sources) tend to be lower and O3 is relatively higher.  This can lead to negative correlations
 2      between O3 and the primary pollutants in the winter. As shown in Figure 3-6 in the Chapter 3
 3      Annex, the relationship between O3 and PM2 5 was U-shaped for the year-round data in Fort
 4      Meade, MD. The negative PM2 5/O3 slope was in the range of O3 concentrations less than 30
 5      ppb, providing supporting evidence of the aforementioned winter phenomenon.
 6           This changing relationship between O3 and temperature, as well as O3 and other pollutants
 7      across seasons, and its potential implications to health effects modeling has not been examined
 8      thoroughly in the time-series literature. Even the flexible smoother-based adjustments for
 9      seasonal and other time-varying variables cannot fully take into account these complex
10      relationships. One obvious way to alleviate or avoid this complication is to analyze data by
11      season. While this practice reduces sample size, its extent would not be as serious as PM (which
12      is collected only every 6th  day in most locations) because O3  is collected daily, though only  in
13      warm seasons in some states. An alternative approach is to include  separate O3 concentration
14      variables for each season (by multiplying O3 concentrations by a season indicator variable).
15           In locations where seasonal variability may be a factor,  O3 effect estimates calculated using
16      year-round data can be misleading, as the changing relationship between O3, temperature, and
17      other pollutants across seasons may have a significant influence on the estimates. Analyses  have
18      indicated that confounding from seasonal variability may be controlled effectively by stratifying
19      the data by season.
20
21      7.1.3.6   Confounding Effects of Copollutants
22           Extensive discussions on the issues related to confounding effects among air pollutants in
23      time-series study design are provided in Section 8.4.3 of the 2004 PM AQCD.  Since the general
24      issues discussed in that document are applicable to all pollutants, such discussions are not
25      repeated here. What was not discussed in the 2004 PM  AQCD was  the issue of changing
26      relationships among air pollutants across seasons.  For O3, the confounder of main interest is
27      PM, especially fine particles or sulfates that are high in  summer, as other copollutants (e.g.,  CO,
28      NO2, SO2) tend to be elevated in the colder season. As mentioned in the previous section, PM
29      indices in some urban locations may be positively correlated with O3 in the  summer and
30      negatively correlated in the winter. Thus the correlation between O3 and PM for year-round data
31      may be misleading. The high reactivity of O3 with certain copollutants further complicates the

        January 2005                              7-12        DRAFT-DO NOT QUOTE OR CITE

-------
 1      analysis.  For example, the reaction between NOX, emitted from motor vehicles, and O3 results in
 2      reduced O3 levels but increased NO2 levels during high traffic periods.
 3           Multipollutant models often are used to assess potential confounding by copollutants. The
 4      limitations of multipollutant regression models, including the potential transfer of "effects" from
 5      causal pollutant to noncausal pollutant in the presence of unequal measurement errors, are
 6      discussed in the 2004 PM AQCD (Sections 8.4.3 and 8.4.5).  In addition, uncertainty remains as
 7      to the use of multipollutant regression models to assess the independent health effects of
 8      pollutants that are correlated. Particularly in the case of O3, there remains concern as to whether
 9      multipollutant regression models for year-round data can adjust for potential confounding
10      adequately due to the changing relationship between O3 and other pollutants.  Despite these
11      limitations, multipollutant models are still the prevailing approach in most, if not all, studies of
12      O3 health effects and serve as an important tool in addressing the issue of confounding by
13      copollutants, especially in season-stratified analyses.
14
15      7.1.3.7   Model Uncertainty and Multiple Testing
16           In the analyses of air pollution health effects, there is often uncertainty as to which model
17      is most appropriate.  In the case of PM, there were concerns that the positive associations found
18      with mortality were the result of multiple testing and selection. Testing many models to identify
19      the best fit can lead to an underestimation of uncertainty, thus there is a need for statistical
20      methods that can identify the best model while properly accounting for model uncertainty.
21      Standard  methods of variable selection include Akaike Information Criterion and Bayes
22      Information Criterion. Bayesian model  averaging is a recent family of methods used to address
23      these issues.  In Bayesian model averaging, predictions and inferences are based on a set of
24      models, rather than a single model, and each model contributes proportionally to the support it
25      receives from the observed data (Clyde, 1999).  While there remains concerns about the large
26      enumeration of models, Bayesian model averaging is a computationally efficient way to
27      incorporate model uncertainty into decision making.
28           Some researchers have used other methods to address the issue of multiple testing.
29      Dominici et al. (2003) used a minimum  number of tests in the U.S. 90 cities study, which
30      minimized the uncertainty associated with multiple testing, but at the cost of possibly not
31      identifying the best model. Another method used by Dominici et al. (2003) to evaluate the

        January 2005                               7-13        DRAFT-DO NOT QUOTE OR CITE

-------
 1      model was sensitivity testing. Lumley and Sheppard (2000) used different control variables to
 2      check the bias in model identification.  They found that the bias was small, but of the same
 3      magnitude as the estimated health impacts. Another approach is to use one set of data for model
 4      identification, and a second set of data for model fitting. Cross validation also sheds light on this
 5      issue.
 6           With the currently available knowledge, multiple testing is unavoidable in air pollution
 7      health effects analyses. To address the issues of model uncertainty and multiple testing, further
 8      research leading to the development of standard methodology may be necessary.
 9
10      7.1.3.8  Impact of GAM Convergence Issue on O3 Risk Estimates
11           Generalized Additive Models (GAM) have been widely utilized for epidemiologic analysis
12      of the health effects attributable to air pollution.  The impact of the GAM convergence issue was
13      thoroughly discussed in Section 8.4.2 of the 2004 PM AQCD. Reports have indicated that using
14      the default convergence criteria in the Splus software package for the GAM function can lead to
15      suboptimal regression estimates for PM and an underestimation of the standard error of that
16      effect estimate (Dominici et al., 2002; Ramsay et al., 2003).  GAM default convergence criterion
17      has a convergence precision of 10~3 and a maximum number of 10 iterations.  The more stringent
18      convergence criterion refers to increased stringency of both the convergence precision and
19      number of iterations. The default convergence criteria was found to be a problem when the
20      estimated relative risks were small, and two or more nonparametric smoothing curves were
21      included in the GAM (Dominici et al., 2002). The magnitude and direction of the bias depend in
22      part on the concurvity of the independent variables in the GAM and the magnitude of the risk
23      estimate.  Most attention has been focused on the influence of the GAM function on effect
24      estimates for PM. However, because O3 covaries more strongly with both weather and time
25      factors than does PM, the issue of GAM convergence criteria for O3 needs to be considered.
26           A recent meta-analysis by Stieb et al. (2003) found a difference in O3-mortality risk
27      estimates between the GAM studies and non-GAM studies.  In the single-pollutant models, the
28      O3-mortality risk estimates for the non-GAM studies and GAM studies were 1.8% (95% CI: 0.5,
29      3.1) and 2.2% (95% CI:  1.4, 2.8), respectively, per 40 ppb daily 1-h max O3.  In the
30      multipollutant models, the pooled risk estimate was 1.0% (95% CI: -0.5, 2.6) for non-GAM
31      studies and 0.5% (95% CI: -1.0, 1.9) for GAM studies.

        January 2005                              7-14        DRAFT-DO NOT QUOTE OR CITE

-------
 1           A few GAM studies reanalyzed the O3 risk estimates using more stringent convergence
 2      criteria or general linear models (GLM). Reanalysis of an asthma hospital admissions study in
 3      Seattle, WA (Sheppard et al., 1999; reanalysis Sheppard, 2003) indicated that there were only
 4      slight changes in the risk estimates when using more stringent convergence precision (10~8) in
 5      GAM. The original GAM analysis indicated an excess risk of 9% (95% CI: 3, 17) whereas the
 6      stringent GAM analysis found an excess risk of 11% (95% CI: 3, 19) per 30 ppb increase in 8-h
 7      max O3.  Similar results were found using GLM with natural splines, 11% (95% CI: 2, 20).
 8      In the reanalysis of Santa Clara County, CA data, Fairley (1999; reanalysis Fairley, 2003) used
 9      the same methods as the original analysis except the convergence precision (e) was increased
10      from 10~4 to 10~12 and the maximum number of iterations (M) were increased from  10 to 107.
11      The O3 nonaccidental mortality risk estimates slightly increased from 2.8% using default GAM
12      parameters to 2.9% using stringent GAM parameters per 30 ppb increase in 8-h max O3. The
13      O3-mortality risk estimates further increased to 3.0% using GLM with natural cubic splines.
14      In the reanalysis of the Netherlands data by Hoek et al.  (2000; reanalysis Hoek, 2003), the O3
15      nonaccidental mortality  risk estimates increased from 1.3% (default GAM) to 1.5% (stringent
16      GAM, e = 10~8, M = 103) and 1.6% (GLM with natural splines) per 30 ppb increase in 8-h avg
17      O3 (12 p.m.-8 p.m.).  In the analysis of the large 90 U.S. cities (Samet et al., 2000; reanalysis
18      Dominici et al., 2003), the year-round combined  estimate of O3 nonaccidental mortality risk
19      changed from a nonsignificant negative value of approximately -0.2% (default GAM) per
20      20 ppb change in 24-h avg O3 to a significantly positive excess risk of 0.8% (stringent GAM,
21      e= 10~15, M=103).
22           Most analyses comparing results using default GAM convergence criteria to results from
23      stringent GAM convergence criteria and GLM have found little difference among the O3 effect
24      estimates.  However, one study by Cifuentes et al. (2000) in Santiago, Chile observed a large
25      difference in the O3-mortality excess risks calculated using default GAM (2.4%) and GLM
26      (0.3%). Therefore, the impact of the GAM convergence problem appears to vary depending on
27      data sets, and likely depends upon the intercorrelation among covariates and the magnitude of
28      the risk estimate. However, in the limited number of studies that have reanalyzed O3 risk
29      estimates, there is little evidence that default GAM analyses resulted in positively biased
30      estimates as observed for PM. Generally it appears that the use of default convergence criteria
31      in GAM tends to bias risk estimates towards the null, in addition to underestimating the standard

        January 2005                              7-15        DRAFT-DO NOT QUOTE OR CITE

-------
 1      errors.  In uniformity with the approach used in the 2004 PM AQCD, the results from studies
 2      that analyzed data using GAM with default convergence criteria and at least two nonparametric
 3      smoothing terms are generally not considered in this chapter, with a few exceptions as noted.
 4
 5      7.1.4   Approach to  Presenting O3 Epidemiologic Evidence
 6           To produce a thorough appraisal of the evidence, we first concisely highlight key points
 7      derived from the 1996 O3 AQCD assessment. Then pivotal information, including
 8      methodological features and results, from important new studies that have become available
 9      since the prior O3 AQCD are presented in summary tables in Chapter 7 of the Annex.  In the
10      main body of the chapter, particular emphasis is focused on those studies and analyses
11      considered to provide information most directly applicable for development of criteria. Not all
12      studies should be accorded equal weight in the overall interpretive assessment of evidence
13      regarding O3-association health effects. Among well-conducted studies with adequate control
14      for confounding, increasing scientific weight should be accorded in proportion to the precision
15      of their effect estimates.  Small-scale studies without a wide range of exposures generally
16      produce less precise estimates compared to larger studies with an adequate exposure gradient.
17      Therefore, the range of exposures, the size of the study as indicated by the length of the study
18      period and total number of events, and the inverse variance of the principal effect estimate are all
19      important indices useful in determining the likely precision of the health effect estimates and in
20      according relative scientific importance to the findings of a given study. In any case, emphasis
21      should  be accorded to estimates from studies with narrow confidence bands.
22           Emphasis is placed on text discussion of (1) new multicity studies that employ
23      standardized methodological analyses for evaluating O3 effects across several or numerous cities
24      and often provide overall effect estimates based on combined analyses of information pooled
25      across multiple cities; (2) studies that consider O3 as a component of a complex mixture of air
26      pollutants, including in particular the gaseous criteria pollutants (CO, NO2, SO2) and PM; and
27      (3) North American studies conducted in the U.S. or Canada. Multicity studies are of particular
28      interest and value due to their evaluation of a wider range of O3 exposures and large numbers of
29      observations, thus possibly providing more precise effect estimates than most smaller scale
30      independent studies of single cities. Another potential advantage of the multicity studies,
31      compared to meta-analyses of multiple "independent" studies, is consistency in data handling

        January 2005                             7-16         DRAFT-DO NOT QUOTE OR CITE

-------
 1      and model specifications that eliminates variation due to study design. Also, unlike regular
 2      meta-analyses, they do not suffer from potential omission of nonsignificant results due to
 3      "publication bias." Furthermore, geographic patterns of air pollution effects have the potential
 4      to provide especially valuable evidence regarding relative homogeneity and/or heterogeneity of
 5      O3 health effects relationships across geographic locations.  In accordance to the emphasis
 6      placed on the O3 epidemiology studies in this chapter, the tables in the Chapter 7 Annex were
 7      organized by region with multicity studies in each region presented first.
 8           In the coming sections, field/panel studies and studies of emergency department visits and
 9      hospital admissions, which contributed to the establishment of the revised 1997 NAAQS for O3,
10      are presented first. This is followed by a discussion of O3-related mortality and chronic effects.
11      The chapter ends with an integrative discussion providing a summary and conclusions.
12
13
14      7.2   FIELD STUDIES ADDRESSING ACUTE EFFECTS OF OZONE
15      7.2.1   Summary of Key Findings on Field Studies of Acute Effects
16              From the 1996 O3 AQCD
17           In the 1996 O3 AQCD, individual-level camp  and exercise studies provided useful
18      quantitative information on the exposure-response relationships linking human lung function
19      declines with O3 exposure occurring in ambient air. The available body of evidence supported a
20      dominant role of O3 in the observed lung function decrements. Extensive epidemiologic
21      evidence of pulmonary function responses to ambient O3 came from camp studies.  Six studies
22      from  three separate research groups provided a combined database on individual exposure-
23      response relationships for 616 children (mostly healthy, nonasthmatic) ranging in age from 7 to
24      17 years, each with at least six sequential measurements of FEVj (forced expiratory volume in
25      1 second) while attending summer camps (Avol et al., 1990; Higgins et al., 1990; Raizenne
26      et al., 1987, 1989; Spektor et al., 1988a,  1991). When analyzed together using consistent
27      analytical methods, these data yielded an average relationship between afternoon FEVj  and
28      concurrent-hour O3 concentration of -0.50 mL/ppb  (p < 0.0001), with study-specific slopes
29      ranging from -0.19 to -1.29 mL/ppb. Exposure in  camp studies usually extended for multiple
30      hours. Although the regression results noted above were based on one-hour O3 levels, single-
31      and multiple-hour averages were observed to be highly correlated, thus these results might

        January 2005                             7-17        DRAFT-DO NOT QUOTE  OR CITE

-------
 1      represent, to some extent, the influence of multihour exposures.  In addition to the camp study
 2      results, two studies involving lung function measurements before and after well-defined exercise
 3      events in adults yielded exposure-response slopes of -0.4 mL/ppb (Selwyn et al., 1985) and
 4      -1.35 mL/ppb (Spektor et al., 1988b).  Ozone concentrations during exercise events of
 5      approximately !/2-hour duration ranged from 4 to 135  ppb in these studies.
 6           Results from other field panel studies also supported a consistent relationship between
 7      ambient O3/oxidant exposure and acute respiratory morbidity in the population.  Respiratory
 8      symptoms (or exacerbation of asthma) and decrements in peak expiratory flow (PEF) occurred
 9      with increased ambient O3 concentrations, especially  in asthmatic children (Lebowitz et al.,
10      1991; Krzyzanowski et al.,  1992). The aggregate results showed greater responses in asthmatic
11      individuals than in nonasthmatics (Lebowitz et al., 1991; Krzyzanowski et al., 1992), indicating
12      that asthmatics might constitute a sensitive group in epidemiologic studies of oxidant air
13      pollution. Since the  1996 O3 AQCD, new research has examined a broad scope  of field studies
14      which are presented next.
15
16      7.2.2  Introduction to Recent Field Studies of Acute O3 Effects
17           Numerous field studies carried out over the past decade have tested for and, in many cases,
18      observed acute associations between measures of respiratory ill-health and O3 concentrations in
19      groups of subjects (Table AX7-1 in Chapter 7 Annex).  Acute field studies are distinguished
20      from other acute epidemiologic study designs in  that they recruit and collect data from
21      individual human subjects instead of utilizing administrative data on aggregate health outcomes
22      such as daily mortality, hospital admissions, or emergency department visits.  Because of the
23      logistical burden associated with direct data collection from individual subjects,  field/panel
24      studies tend to be small in both numbers of subjects and in duration of follow-up.  While this
25      may limit the statistical power  of field  studies, it is compensated for by the ability to determine
26      individual-level information on health  outcomes, exposure levels, and other potentially
27      confounding factors.
28           The most common outcomes measured in acute field studies on the effects of air pollution
29      exposure are lung function and various respiratory symptoms. Other respiratory outcomes
30      examined on a limited basis include inflammation and generation of hydroxyl radicals in the
31      upper airways, and school absences.  Several studies examined cardiovascular outcomes

        January 2005                              7-18        DRAFT-DO NOT QUOTE OR CITE

-------
 1      including heart rate variability and risk of myocardial infractions.  The first group of studies
 2      provides varying degrees of evidence supporting the conclusion that elevated O3 levels can have
 3      negative impacts on lung function and symptoms, confirming and adding to the body of
 4      knowledge that is presented in the 1996 O3 AQCD. Some emphasis has been placed in
 5      examining the independent role of O3 in the presence of PM and other pollutants.  The other new
 6      studies contribute information on cardiopulmonary outcomes which had not been as well-
 7      documented previously.
 8
 9      7.2.3    Acute O3 Exposure and Lung Function
10          As discussed in the 1996 O3 AQCD and in the earlier chapter of this document on
11      controlled human exposure studies (Chapter 6), a large body of literature from clinical and field
12      studies has clearly and consistently demonstrated reversible decrements in pulmonary function
13      following acute O3 exposure. Significant O3-induced spirometric and symptom responses have
14      been observed in clinical studies of exercising healthy young adults (see Section 6.2) and in
15      some potentially susceptible subpopulations, namely asthmatics and children (see Sections 6.3.2
16      and 6.5.1). Field studies of acute O3 exposure that examine pulmonary function fall into two
17      distinct groupings, those that conduct spirometry (FEVj and FVC  [forced vital capacity]) and
18      those that measure PEF. Results from the previous O3 AQCD and Chapter 6 of this document
19      support the conclusion that the spirometric parameter FEVj is the  stronger and more consistent
20      measure  of lung function.  PEF is a useful clinical measure that is  more feasible to perform in
21      field studies, however its measurements are more variable and possibly less reliable than FEVj
22      (Fuhlbriggeetal., 2001).
23          Studies of FEVj will be presented first, followed by a discussion of PEF studies.  Other
24      dividing  aspects between these two major types of lung function studies include health status of
25      subjects  (e.g., healthy, mildly asthmatic, severely asthmatic), time spent outdoors, and exertion
26      levels. Several studies brought these factors together to produce informative data.  Some FEVj
27      studies involved both increased outdoor O3 exposure and higher exertion levels. The results
28      from this group of subjects are comparable to those from the exercising subjects in the clinical
29      studies discussed in Chapter 6.
30
31

        January 2005                              7-19        DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
7.2.3.1   Acute O3 Studies with Spirometry (FEVt)
      Studies published over the past decade have provided some new insights on the acute
effects of O3 on FEVj. Tables 7-la, 7-lb, and 7-lc summarize the results of all studies that
investigated quantitative O3-related effects on FEVj. Four studies of spirometry were not
included in the tables; three studies did not provide quantitative O3 data (Cuijpers et al., 1994;
Delfino et al., 2004; Frischer et al., 1997) and one measured FEV075 (forced expiratory volume
in 0.75 seconds) (Scarlett et al.,  1996).  With few exceptions, the O3 effect estimates showed
decrements for FEVj across studies and several were statistically significant.  These studies are
discussed in further detail, starting with the O3 effect on individuals with elevated exertion
levels.

Exercise and outdoor worker panels
      The current 8-hour NAAQS for O3 has its original basis in the controlled human exposure
studies, as discussed in Chapter 6. These field studies with subjects at elevated exertion levels
          Table 7-la.  Field Studies that Investigated the Association between Acute Ambient O3
                                      Exposure and Changes in FEVt
         Reference
                       Study Location
Study Period
n"
         Linn etal. (1996)
         Korricketal. (1998)
         Braueretal. (1996)

         Hoppeetal. (1995a)
         Ulmeretal. (1997)
         Castillejos etal. (1995)
         Romieuetal. (1998)
         Romieu et al. (2002)
         Chen etal. (1999)
                      Rubidoux, Upland, and Torrance, CA
                      Mount Washington, NH
                      Fraser Valley, British Columbia,
                      Canada
                      Munich, Germany
                      Freudenstadt and Villingen, Germany
                      SW Mexico City
                      Mexico City
                      Mexico City
                      Sanchun, Taihsi, and Linyuan, Taiwan
Fall-spring 1992-1993, 1993-1994    269
Summers 1991, 1992              530
Jun-Aug 1993                     58

Apr-Sep 1992-1994                208
Mar-Oct 1994                    135
Aug 1990-Oct 1991                 40
Mar-May 1996; Jun-Aug 1996        47
Octl998-Apr2000                158
May 1995-Jan 1996                895
         aThe number of the total study population is presented. Some of the effect estimates presented in Tables 7-lb and
          7-lc were based on a subset of the total population.
        January 2005
                                            7-20
DRAFT-DO NOT QUOTE OR CITE

-------
   Table 7-lb. Changes in FEVt (95% CI) Associated with Acute Ambient O3 Exposures,
                           Ordered by Size of the Estimate

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Reference
Braueretal. (1996)
Braueretal. (1996)
Romieuetal. (1998)
Ulmeretal. (1997)
Hoppeetal. (1995a)
Romieuetal. (1998)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Hoppeetal. (1995a)
Ulmeretal. (1997)
Romieuetal. (1998)
Chen etal. (1999)
Chen etal. (1999)
Romieu et al. (2002)
Romieu et al. (2002)
Romieuetal. (1998)
Chen etal. (1999)
Study Population/Analysis
Berry pickers, next morning
Berry pickers, afternoon
Street workers on placebo
(1st phase, lag 0-1)
School children in
Freudenstadt
Juvenile asthmatics
Street workers on placebo
(1st phase, lagO)
Clerks
School boys in Freudenstadt
and Villingen
Athletes
School children in
Freudenstadt and Villingen
Forestry workers
School girls in Freudenstadt
and Villingen
Street workers on
supplement (1st phase,
lag 0-1)
Children, with NO2 in model
(lag 1)
Children (lag 1)
Moderate to severe asthmatic
children on placebo (lag 1)
Moderate to severe asthmatic
children on placebo, with
NO2 and PM10 in model
(lag 1)
Street workers on
supplement
(1st phase, lagO)
Children (lag 2)
Mean O3
Level (ppb)
40.3
40.3
123
50.6
74 b
123
68 b
41.4
71b
41.4
64 b
41.4
123
19.7-110.3°
19.7-110.3°
102
102
123
N/A
Exposure
Index
1-hmax
1-hmax
1-hmax
!/2-h max
!/2-h max
1-hmax
i/2-h max
!/2-h max
!/2-h max
i/2-h max
i/2-h max
i/i-h max
1-hmax
1-h max
1-h max
1-hmax
1-hmax
1-hmax
24-h avg
Change in FEV/
(mL)
-180.0
-152.0
-117.2
-87.5
-84.0
-71.6
-63.2
-61.4
-60.8
-56.6
-56.0
-44.2
-41.2
-34.0
-25.6
-18.8
-18.4
-17.6
-17.4
(-227.0
(-183.4
(-207.4
(-143.2
(-196.4
(-113.9
(-108.8
(-122.9
(-115.2
(-101.3
(-118.4
(-105.0
(-143.9
(-60.7,
(-49.1,
(-34.2,
(-35.5,
(-68.6,
(-41.7,
,-133.0)
, - 120.6)
,-27.0)
,-31.7)
, 28.4)
,-29.3)
,-17.6)
,0.0)
,-6.4)
,-12.0)
,6.4)
, 16.7)
,61.5)
-7.3)
-2.1)
-3.4)
-1.3)
33.4)
6.9)
January 2005
7-21
DRAFT-DO NOT QUOTE OR CITE

-------
      Table 7-lb (cont'd).  Changes in FEVt (95% CI) Associated with Acute Ambient O3
                         Exposures, Ordered by Size of the Estimate
Reference
20
21
22
23
24
25
26
27

28
29
30
31
Chen et al.
(1999)
Ulmeretal. (1997)
Chen et al.
Romieu
et
Chen et al.
Romieu
Chen et
et
al.
Linn et al.


Linn et al.
Romieu
Romieu
Romieu
et
et
et
(1999)
al. (1998)
(1999)
al. (1998)
(1999)
(1996)

(1996)
al. (2002)
al. (2002)
al. (2002)
Study Population/Analysis
Children (lag 2)
School children in Villingen
Children (lag 1)
Street workers on placebo
(2nd phase, lag 0)
Children (lag 7)
Street workers on placebo
(2nd phase, lag 0-1)
Children (lag 7)
School children, next
morning
School children, afternoon
All asthmatic children on
placebo (lag 1)
Moderate to severe asthmatic
on supplement (lag 1)
Moderate to severe asthmatic
Mean O3
Level (ppb)
19.7-110.3°
32.1
N/A
123
19.7-110.3°
123
N/A
23

23
102
102
102
Exposure
Index
1-h max
Vi-h max
24-h avg
1-h max
1-h max
1-h max
24-h avg
24-h avg

24-h avg
1-h max
1-h max
1-h max
Change in FEVj3
(mL)
-16.0
-15.0
-13.6
-13.2
-12.4
-12.0
-6
-5

-3
-3
-0
-0
.0
.2

.6
.6
.7
.2
(-44.2,
(-74.
(-33.
(-64.
(-31.
6,
2,
2,
2,
(-96.7,
(-20.
(-15.

(-13.
(-13.
(-15.
(-15.
9,
o,

8,
5,
1,
6,
12.2)
44.5)
6.0)
37.8)
6.4)
72.7)
8.9)
4.6)

6.6)
6.3)
13.7)
15.1)
 32   Romieu et al. (2002)
 33   Romieu etal. (1998)
on supplement, with NO2
and PM10 in model (lag 1)

All asthmatic children on
supplement (lag 1)

Street workers on
supplement
(2nd phase, lag 0)
102       1-h max     0.8   (-9.8,11.3)


123       1-h max     6.0   (-23.8,35.8)
34
35
Hoppeetal. (1995a)
Romieu etal. (1998)
Seniors
Street workers on
supplement
(2nd phase, lag 0-1)
70 b
123
!/2-h max
1-h max
13.6
27.2
(-26.
(-25.
8, 54.0)
3,79.7)
 a Change in FEVl is per standard unit ppb O3 (40 ppb for i/2-h max O3 and 1-h max O3, 30 ppb for 8-h max O3, and
  20 ppb for 24-hr avg O3).
 bMean O3 concentration on high O3 days.
 0 Range of O3 concentrations.
January 2005
                   7-22
  DRAFT-DO NOT QUOTE OR CITE

-------
  Table 7-lc. Cross-day Changes in FEVt Associated with Acute Ambient O3 Exposures,
                               Ordered by Size of the Estimate

1

2

3

4

5

6

7

8


9

10

11

12

13

14

15

16

17

Reference
Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)


Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Korrick et al. (1998)

Linnetal. (1996)

Castillejos et al. (1995)

Braueretal. (1996)

Study Population/
Analysis
Hikers with wheeze or
asthma (post-pre-hike)
Hikers who hiked 8-12
hours (post-pre-hike)
Hikers age 28-37 years
(post-pre-hike)
Hikers who never smoked
(post-pre-hike)
Hikers male
(post-pre-hike)
Hikers age 38-47 years
(post-pre-hike)
All hikers
(post-pre-hike)
All hikers, with PM2 5
and acidity in model
(post-pre-hike)
Hikers age 18-27 years
(post-pre-hike)
Hikers female
(post-pre-hike)
Hikers age 48-64 years
(post-pre-hike)
Hikers without wheeze or
asthma (post-pre-hike)
Hikers who hiked 2-8 hours
(post-pre-hike)
Hikers who formerly
smoked (post-pre-hike)
School children
(p.m. -a.m.)
Private primary school
(post- pre -exercise)
Berry pickers
(post- pre-workshift)
Mean O3
Level (ppb)
40

40

40

40

40

40

40

40


40

40

40

40

40

40

23

112.3

40.3

Exposure
Index
8-h avg

8-h avg

8-h avg

8-h avg

8-h avg

8-h avg

8-h avg

8-h avg


8-h avg

8-h avg

8-h avg

8-h avg

8-h avg

8-h avg

24-h avg

1-hmax

1-hmax

Cross-day Change
in FEVj a (mL)
-182.5b

-84.5b

-82.1b

-72. 3 b

-67.4b

-64.9b

-62.5b

-58.8b


-52.7b

-47.8b

-46.5b

-44.1b

-40.4b

-29.4b

-11.6

-9.1C

0

(-312.2, -52.9)

(-154.1, -14.9)

(-139.7, -24.4)

(-132.3, -12.2)

(-127.4, -7.3)

(-127.3, -2.5)

(-115.3, -9.7)

(-135.6, 18.0)


(-117.5, 12.2)

(-141.4,45.9)

(-125.8,32.7)

(-101.7, 13.5)

(-110.0,29.2)

(-125.4,66.6)

(-20.6, -2.6)

(-13.6, -4.7)

(-47.0,47.0)

 aCross-day change in FEVj is per standard unit ppb O3 (40 ppb for 1-h max O3, 30 ppb for 8-h avg O3, and 20 ppb
  for 24-h avg O3).
 b Korrick et al. presented % change in FEVj. The data was transformed to FEVj units of mL by multiplying by the
  total population average FEVj of 4,083 mL.
 0 Castillejos et al. presented % change in FEVj. The data was transformed to FEVj units of mL by multiplying by
  the total population average FEVj of 1,900 mL.
January 2005
7-23
DRAFT-DO NOT QUOTE OR CITE

-------
 1      are of particular interest due to their similarities to the human chamber studies. The majority of
 2      human chamber studies have examined the effects of O3 exposure in subjects performing
 3      continuous or intermittent exercise for variable periods  of time (see Chapter 6 of this O3 AQCD).
 4           A study by Brauer and colleagues (1996) reported unusually large O3 effects on lung
 5      function among outdoor workers.  This study presented O3 effects during an extended outdoor
 6      exposure period combined with elevated levels of exertion.  The investigators repeatedly
 7      measured spirometric lung function before and after outdoor summer work shifts over 59 days
 8      on a group of 58 berry pickers in Fraser Valley, British  Columbia, Canada.  Subjects, both male
 9      and female, ranged from 10 to 69 years old, with a mean age of 44 years.  Outdoor work shifts
10      averaged 11 hours in duration. The mean 1-h max O3 concentration was 40.3  ppb.  Exertion
11      levels were estimated using portable heart rate monitors carried over a period  of four or more
12      hours by a representative subset of subjects during 16 work shifts.  Heart rates were essentially
13      constant over the work shift, averaging 36% higher than resting levels. The authors estimated
14      that minute ventilations may have averaged roughly 30  L/min during work.  Post-shift FEVj and
15      FVC showed large decreases as a function of O3 concentration and those effects remained
16      significant when PM2 5 was included in the analysis.  Significant declines in lung function also
17      were observed on the morning following high O3 exposure.  The effects seen in this study are
18      larger than have been reported previously.  For example, afternoon FEVj was  3.8 mL lower per
19      1 ppb increase in O3 concentrations, compared to the decline of 0.4 mL/ppb and 1.35 mL/ppb
20      observed in the earlier adult exercise studies (Spektor et al., 1988b;  Selwyn et al.,  1985).
21      Further, when data were restricted to days with 1-h max O3 concentrations under 40 ppb, the O3
22      effects on afternoon FEVj did not change in magnitude  and remained  significant.
23           In a Mexico City study of 47 outdoor street workers (Romieu et al., 1998), spirometry was
24      performed repeatedly at the end of the workshift over a  two month period. Subjects were
25      exposed to outdoor ambient O3 levels for a mean of 7.4  hours during the workday.  Among those
26      who had never taken an antioxidant supplement (subjects who received a placebo during the 1st
27      phase of the study), same day O3 concentrations were significantly associated  with decreases in
28      FEVj. A mean decline of 71.6 mL (95% CI: 29.3, 113.9) was observed per 40 ppb increase in
29      1-h max O3.  The results from this study, in addition to those from the Canadian study of berry
30      pickers (Brauer et al., 1996), indicate  that outdoor workers are a potentially susceptible
31      population that may need protection from O3 exposures.

        January 2005                              7-24        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Hoppe et al. (1995a) examined forestry workers (n = 41) for changes in pulmonary
 2      function attributable to O3 exposure in Munich, Germany. In addition, athletes (n = 43) were
 3      monitored in the afternoon after a two-hour outdoor training period. Pulmonary function tests
 4      were conducted on days of both "high" (mean lA>-h max O3 of 64 to 74 ppb) and "low" (mean
 5      !/2-h max O3 of 32 to 34 ppb) ambient O3 concentrations. From the average activity levels,
 6      ventilation rates were estimated.  Athletes, who had a fairly high ventilation rate of 80 L/min,
 7      experienced a significant decrease of 60.8 mL (95% CI: 6.4, 115.2) in FEVj per 40 ppb increase
 8      in !/2-h max O3. Among the forestry workers, an O3-related decline in FEVj also was observed
 9      (- 56.0 mL), however the change was not statistically significant.
10           For the above studies that examine outdoor workers and athletes who train outdoors,
11      Table 7-2 presents the estimated O3 external doses and compares them to changes in FEVj
12      associated with acute ambient O3 exposures. The use of estimated O3 external doses offers
13      another potential for insight into studies that examine subjects with elevated ventilation rates and
14      prolonged outdoor exposures at varying ambient O3 concentrations.  No consistent relationship
15      between estimated O3 external  doses and changes in FEVj can be derived from the limited
16      evidence.
17           One FEVj study appears to mirror the sort of outcome typically seen  in clinical studies. In
18      a study by Korrick et al. (1998), adult hikers (n =  530) of Mount Washington, NH performed
19      spirometry before and after hiking for a mean of 8 hours (range: 2 - 12). The mean hourly O3
20      concentration  ranged from 21 to 74 ppb. After their hike, subjects experienced a mean decline
21      of 1.5%  (95% CI: 0.2, 28) in FEVi and 1.3% (95% CI: 0.5, 2.1) in FVC per 30 ppb increase in
22      the mean of the hourly  O3 concentration during the hike. In addition, Korrick et al. (1998)
23      compared hikers who hiked 8 to 12 hours to those who hiked 2 to 8 hours.  Among those who
24      hiked longer, the % change in FEVj was more than two-fold greater per ppb exposure compared
25      to those who hiked only for 2 to 8 hours. Each hour hiked, which may reflect dose, was
26      associated with a decline of 0.3% (p = 0.05) in FEVj, after adjusting for O3.
27           In a Mexico City  study, the O3 effect attributable to exercise was determined using a group
28      of school children (n = 40) chronically exposed to moderate to high levels  of O3 (Castillejos
29      et al., 1995). Children  were tested up to 8 times between August 1990 and October 1991.
30      Spirometry was performed by the children before and after a one-hour intermittent exercise
31      session outdoors. Outdoor O3 levels ranged up to 365 ppb, with a mean of 112.3 ppb. Linear

        January 2005                              7-25        DRAFT-DO NOT QUOTE OR CITE

-------
           Table 7-2. Estimated O3 External Doses and Changes in FEVt Associated with Acute
                         Ambient O3 Exposures in Outdoor Workers and Athletes

1
2
3
4
Reference
Brauer et al.
Hoppe et al.
Hoppe et al.
Romieu et al
Study Location
( 1 996) Fraser Valley, Canada
(1995a) Munich,
(1995a) Munich,
. (1998) Mexico
Germany
Germany
City
Study Period
Jun-Aug 1993
Apr-Sep 1994
Apr-Sep 1993
Mar-May 1996
N Study
58 Berry
Population
pickers
43 Athletes
41 Forestry Workers
13 Street
Workers3
Age
(years)
10-69
13-38
20-60
18-58

Mean O3 Exposure
Level (ppb) Duration (h)
1
2
3
4
26.0
71.0C
64.0 c
67.3
11
2
7
9
Ventilation
Rate (L/min)
30
80
40
28 d
Exposure
(mg/h)
91
660
298
219
Dose
(mg)
997
1320
2083
1971
Change in FEV/
(mL)
-152.0 (-183.4
-60.8 (-115.2
-56.0 (-118.4
-71.6 (-113.9
,-120.6)
, -6.4)
,6.4)
,-29.3)
         aFor the street workers in Romieu et al. (1998), only results from subjects who had never taken the antioxidant
          supplement (on placebo during 1st phase of study) are presented here.
         b Change in FEV{ is per 40 ppb increase in 1-h max O3 or equivalent.
         °Mean O3 concentration during exposure period was not presented in Hoppe et al. (1995). The !/2-h max
          O3 concentrations are shown here.
         d Ventilation rate was not presented. The ventilation rate of 28 L/min was calculated for a male performing
          a heavy workload for 1/8 of the workday in Table B. 17 of the International Commission on Radiological
          Protection (ICRP) Publication 66 (ICRP, 1994).
 1      trend analyses of the relationship between quintiles of O3 and % change in lung function were
 2      significant. However, stratified analyses indicated that statistically significant changes were
 3      observed only with higher quintiles of O3 exposure (72-125 ppb and 183-365 ppb).  Therefore,
 4      when exercising at higher O3 levels, children experienced significant declines in pulmonary
 5      function despite the repeated daily exposure to moderate and high levels of O3 in Mexico City.
 6           Collectively, the above studies confirm and extend clinical observations that prolonged
 7      exposure periods, combined with elevated levels of exertion or exercise, may magnify the effect
 8      of O3 on lung function.  The most representative data comes from the Korrick et al. (1998) hiker
 9      study.  This U.S. study provided outcome measures stratified by several factors (e.g., gender,
10      age, smoking status, presence of asthma) within a population capable of more than normal
11      exertion.
        January 2005
7-26
DRAFT-DO NOT QUOTE OR CITE

-------
 1      Panels of other risk groups
 2           Hoppe et al. (1995a,b) examined several potentially susceptible populations for changes in
 3      pulmonary function attributable to O3 exposure in Munich, Germany. The forestry workers and
 4      athletes were discussed in the previous section.  Senior citizens (n = 41) and juvenile asthmatics
 5      (n = 43) were also monitored on "low" O3 and "high" O3 days. Subjects were requested to stay
 6      outdoors for at least 2 hours just before the afternoon pulmonary function test.  Clerks (n = 40)
 7      were considered the nonrisk control group.  Although clerks spent the majority of their time
 8      indoors, their outdoor exposures on the "high" O3 days were similar to that of the four other risk
 9      groups. The results showed no significant O3 effects on the senior citizens, who had the lowest
10      ventilation rate.  Asthmatics and clerks experienced slight reductions in FEVj on high O3 days.
11      Among all risk groups, juvenile asthmatics experienced the largest O3-related decline in FEVl3
12      though not statistically significant.
13           Several other panel studies performed spirometry in children, another potentially
14      susceptible group (Chen et al., 1999; Cuijpers et al., 1994; Frischer et al.,  1997; Linn et al., 1996;
15      Romieu et al., 2002; Scarlett et al., 1996; Ulmer et al., 1997). All studies, with the exception of
16      Scarlett et al. (1996), observed a statistically significant decrease in FEVj associated with O3
17      exposure. One large study measured spirometric lung function in 895 school children in three
18      towns in Taiwan (Chen et al., 1999). Lung function was measured only once for each subject.
19      The authors reported statistically significant associations between diminished FEVj and FVC
20      with a 1-day lag of O3 concentrations.  Effect sizes were typical of those observed in past studies,
21      i.e., 0.5 to 1.0 mL decline in FEVj per ppb increase in O3 concentration. Ozone was the only
22      significant air pollutant in multipollutant models including SO2, CO, PM10, and NO2. The O3
23      associations became nonsignificant when days with O3 above 60 ppb were excluded from the
24      analysis, implying a practical threshold of around 60 ppb in this individual study.
25           Linn et al.  (1996) repeatedly measured spirometric lung function among 269 school
26      children in three southern California communities (Rubidoux, Upland, and Torrance). Lung
27      function was measured over five consecutive days, once in each of three seasons over two school
28      years. Between-week variability was effectively removed from the analysis by seasonal terms in
29      the model. Statistical power was limited by the narrow range of exposures that were
30      experienced within each week.  In addition, the study was restricted to the school year,
31      eliminating most of the "high" O3 season from consideration. During the study period, 24-h

        January 2005                             7-27        DRAFT-DO NOT QUOTE OR CITE

-------
 1      avg O3 levels at the central monitoring site ranged up to 53 ppb (mean 23 ppb) while personal
 2      measurements ranged up to 16 ppb (mean 5 ppb).  The difference between morning (tested near
 3      the beginning of the school day) and afternoon (tested following lunch) FEVj was significantly
 4      associated with same-day O3 concentrations. Other associations (involving individual morning
 5      or afternoon FVC and FEVj measurements) went in the plausible direction but were not
 6      statistically significant.
 7           Ulmer et al. (1997) examined 135 children aged 8 to 11 years in two towns in Germany
 8      from March to October 1994 for O3 effects on pulmonary function at four time periods.  The
 9      cross-sectional results at each of the four time points showed limited FVC and no FEVj
10      associations. However,  the longitudinal analysis, which combined data from all four periods,
11      obtained a statistically significant negative association between O3 exposure and both FVC and
12      FEVj for the town with the higher O3 levels (median lA>-h max of 50.6 ppb versus 32.1 ppb).
13      In the cross-sectional analysis, between-person variability could not be distinguished from
14      within-person variability, limiting the statistical power. The longitudinal study design, in which
15      subjects provided multiple days of measurements, had greater power as it provided information
16      about both between- and within-subject responses.
17
18      7.2.3.2   Acute O3 Studies of PEF
19           Many studies of the acute effect of O3 on PEF examined PEF levels daily, both in the
20      morning and afternoon.  PEF follows a circadian rhythm with the highest values found during
21      the late afternoon and lowest values during the night and early morning. Due to the diurnal
22      variation in PEF, most studies analyzed their data after stratifying by time of day. The peak flow
23      studies examined both asthmatic panels and healthy individuals. The asthma panels are
24      discussed first.
25
26      Asthma panels
27           Asthmatics were examined in several panel studies.  Several aspects of these studies affect
28      the outcomes. For example, large panels have a greater opportunity to test the hypothesis of an
29      O3 effect on PEF.  In addition, the severity of asthma in the panel subjects and the medications
30      that they take may affect the results of the study. Figures 7-la and 7-lb present % changes in
31      morning and evening PEF outcomes from six panel studies of mostly asthmatic children. The

        January 2005                              7-28        DRAFT-DO NOT QUOTE OR CITE

-------
10

U- 5
LJJ
Q_
C
d> n
OJ u
C
re
O
-10
-


-
-
-
-
-



1
Asthmatic
n
X





1

no
yes


1
t

2

—



E


i i l i I I
„ t t I t H I ' -
-
-
3 4 5 6 7 8 9 10 11 12
Code
Figure 7-la.  Percent change (95% CI) in morning PEF in children per 40 ppb increase in
              1-h max O3 or equivalent, arranged by size of the effect estimate. Study
              codes are explained in the tables below. Information on study location and
              period, study population, and O3 exposure is presented.
  Code   Reference
Study Location
Study Period
N
1
2
3
4
5
6
7
8
9
10
11
12
Code
1
2
3
4
5
6
7
8
9
10
11
12
Ross etal. (2002) a
Gold etal. (1999)
Neasetal. (1999)
Neasetal. (1999)
Romieu etal. (1996)
Gielen etal. (1997)
Romieu etal. (1996)
Romieu etal. (1997)
Romieu etal. (1997)
Gielen etal. (1997)
Romieu etal. (1997)
Romieu etal. (1996)
Population
asthmatic
nonasthmatic
nonasthmatic
nonasthmatic
mildly asthmatic
asthmatic
mildly asthmatic
mildly asthmatic
mildly asthmatic
asthmatic
mildly asthmatic
mildly asthmatic
East Mo line, IL
SW Mexico City
Philadelphia, PA
Philadelphia, PA
N Mexico City
May-Oct 1994




Amsterdam, the Netherlands
N Mexico City
SW Mexico City
SW Mexico City



Amsterdam, the Netherlands
SW Mexico City
N Mexico City


Age (years) Mean O3 Level
5-49
8-11
6-11 57.5
6-11 57.5
5-13
7-13
5-13
5-13
5-13
7-13
5-13
5-13
41.5
52
(SW); 55.9
(SW); 55.9
190
35
190
196
196
35
196
190
Jan-Feb
Jul-Sep
Jul-Sep
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
Apr-Jul
(Ppb)


(ME)
(ME)








, Apr-May, Oct-Nov
1993
1993
1991
1995
1991
1991
1991
1995
1991
1991


,Novl991-Feb

,Novl991-Feb
,Novl991-Feb
,Novl991-Feb

,Novl991-Feb
,Novl991-Feb
O3 Exposure Index












8-h max
24-h avg
12-havgb
12-havg"
1-h max
8-h max
1-h max
1-h max
1-h max
8-h max
1-h max
1-h max
1991


1992

1992
1992
1992

1992
1992
40
40
156
156
71
61
71
65
65
61
65
71
Exposure Lag Days
0-1
1-10
1-5
0
0
2
2
0
2
1
1
1












 a Study population also includes adults.
 b Percent PEF change is presented per 25 ppb increase in 12-h avg O3. The standard units are used for other O3 indices.
January 2005
                7-29
     DRAFT-DO NOT QUOTE OR CITE

-------
10

Uj 5
Q.
_C
0) 0
re
6
SS -5
-10
	
. Asthmatic
n no
x yes
.
I ill"
- , | t I ! " f H -
- f I -
1 2 3 4 5 6 7 8 9 10 11 12
Code
Figure 7-lb.  Percent change (95% CI) in afternoon PEF in children per 40 ppb increase
              in 1-h max O3 or equivalent, arranged by size of the effect estimate.  Study
              codes are explained in the tables below.  Information on study location and
              period, study population, and O3 exposure is presented.
  Code   Reference           Study Location           Study Period                        N
1
2
3
4
5
6
1
8
9
10
11
12
Code
1
2
3
4
5
6
1
8
9
10
11
12
Ross etal. (2002) "
Gold etal. (1999)
Romieuetal. (1997)
Romieu etal. (1996)
Romieuetal. (1997)
Neas etal. (1999)
Gielen etal. (1997)
Romieuetal. (1996)
Neas etal. (1999)
Romieuetal. (1996)
Gielen etal. (1997)
Romieuetal. (1997)
Population
asthmatic
nonasthmatic
mildly asthmatic
mildly asthmatic
mildly asthmatic
nonasthmatic
asthmatic
mildly asthmatic
nonasthmatic
mildly asthmatic
asthmatic
mildly asthmatic
East Moline, IL
SW Mexico City
SW Mexico City
N Mexico City
SW Mexico City
Philadelphia, PA
May-Oct 1994
Jan-Feb, Apr-May, Oct-Nov




Amsterdam, the Netherlands
N Mexico City
Philadelphia, PA
N Mexico City



Amsterdam, the Netherlands
SW Mexico City

Age (years) Mean O3 Level
5-49
8-11
5-13
5-13
5-13
6-11 57.
7-13
5-13
6-11 57.
5-13
7-13
5-13
41.5
52
196
190
196
,5 (SW); 55.9
35
190
,5 (SW); 55.9
190
35
196
Apr-Jul
Apr-Jul
Apr-Jul
Jul-Sep
Apr-Jul
Apr-Jul
Jul-Sep
Apr-Jul
Apr-Jul
Apr-Jul
(Ppb)





(ME)


(ME)



1991
1991
1991
1993
1995
1991
1993
1991
1995
1991
,Novl991-Feb
,Novl991-Feb
,Novl991-Feb


,Novl991-Feb

,Novl991-Feb

,Novl991-Feb
O3 Exposure Index












8-h max
24-h avg
1-h max
1-h max
1-h max
12-havgb
8-h max
1-h max
12-havgb
1-h max
8-h max
1-h max
1991
1992
1992
1992


1992

1992

1992
40
40
65
71
65
156
61
6
156
71
61
65
Exposure Lag Days
0
0-9
1
2
0
0
2
1
1-5
0
1
2












 1 Study population also includes adults.
 b Percent PEF change is presented per 25 ppb increase in 12-h avg O3. The standard units are used for other O3 indices.
January 2005
7-30
DRAFT-DO NOT QUOTE OR CITE

-------
 1      individual effect estimates are identified by study codes which are linked to the associated tables
 2      that provide study details. The tables present general information on the study location and
 3      period as well as specifics on the study population and O3 exposure.  Only single city studies that
 4      performed analyses stratified by time of day are included in the figure.  Studies that examined
 5      cross-day changes and daily variability in PEF were excluded from the figure (e.g., Just et al.,
 6      2002; Thurston et al., 1997).  Collectively, all of the studies indicated decrements of peak flow
 7      but most of the individual estimates were not statistically significant.
 8           Mortimer et al. (2000, 2002) examined 846 asthmatic children from the National
 9      Cooperative Inner-City Asthma Study (NCICAS) for O3-related changes in PEF. Children from
10      eight urban areas in the U.S. (St. Louis, MO; Chicago, IL; Detroit, MI; Cleveland, OH;
11      Washington, DC; Baltimore, MD; East Harlem, NY; and Bronx NY) were monitored from June
12      through August 1993.  This multicities study provides representative data for the U.S. Asthmatic
13      children from urban areas are an important subgroup of asthmatics.  Study children either had
14      physician-diagnosed asthma and symptoms in the past 12 months or respiratory symptoms
15      consistent with asthma that lasted > 6 weeks during the previous year.  In a focused analysis,
16      Mortimer et al. (2000) observed that the subpopulation of low birth weight and premature
17      asthmatic children had significantly greater O3-associated declines in PEF than normal birth
18      weight children.
19           In the main study, Mortimer et al. (2002) further investigated changes in morning PEF
20      associated with O3 concentrations in the eight urban areas.  The reductions in morning PEF
21      were not significant in each individual  city, however when the data from all eight cities were
22      combined, a statistically significant change of-1.18% (95% CI: -2.10, -0.26) per 30 ppb
23      increase in 8-h avg O3 (10 a.m.-6 p.m.) was observed with a cumulative lag of 1 to 5 days.
24      Figure 7-2 illustrates the probability density curves (or density curves) of the results from the
25      city-stratified analysis and that from the pooled analysis of all eight cities. Summary density
26      curves serve as a descriptive aid to the understanding of multiple effect estimates. These curves
27      can be viewed as smoothed histograms. However, unlike a histogram,  summary density curves
28      account for varying standard errors of the individual mean effect estimates. Normal  distribution
29      functions can be calculated for each effect estimate and standard error.  The density curve for the
30      all cities analysis was calculated by taking the derivative of the normal distribution function
31      from the analysis that pooled data from all eight cities. This density curve is a graphical

        January 2005                               7-31         DRAFT-DO NOT QUOTE OR CITE

-------
     1
    0.8
~  0.6
W
Q  °-4
    0.2
     0
                        City-Stratified Analysis
                        All City Analysis
                                 -3
                                    -2            -1
                                   % Change in PEF
       Figure 7-2.  Density curves of the % change in PEF per 30 ppb increase in 8-h avg O3
                    with a cumulative lag of 1 to 5 days for the individual eight NCICAS cities and
                    the pooled average of all cities.  Note that 99% and 78% of the areas under
                    the curves are less than zero for the pooled cities analysis and individual
                    cities analysis, respectively.
       Source: Derived from Mortimer et al. (2002).
 1     presentation of the all cities regression analysis presented by Mortimer et al. (2002) and only
 2     represents one effect estimate and corresponding standard error.  The summary density curve for
 3     the city-stratified analysis was calculated by summing together the normal distribution functions
 4     for each of the eight cities, then taking the derivative of the summed function. The individual
 5     city estimates presented by Mortimer et al. (2002) were used to calculate the summary density
 6     curve. Both density curves graphically depict the mean and distribution of the % change in PEF
 7     per 30 ppb increase in 8-h avg O3.  The area under the density curve and to the left of a value on
 8     the x-axis is an estimate of the probability that the effect  estimate will be less than or equal to
 9     that value. For example, the area under the density  curve to the left of 0% change in PEF is  99%
10     in the all cities analysis.  As a statistically significant decline in PEF was observed in the all
11     cities regression analysis, it is expected that more than 97.5% (p < 0.05) of this area would be
12     less than zero. A wider distribution was observed in the city-stratified analysis, with only 78%
13     of the area less than zero. The all cities analysis likely had a smaller standard error compared to
14     the city-specific analysis as it was based upon more subjects and considered differences between
15     cities to vary about the same mean effect. The regression analysis by Mortimer et al. (2002)
       January 2005
                                       7-32
DRAFT-DO NOT QUOTE OR CITE

-------
 1      suggested a lack of heterogeneity by city, as indicated by the nonsignificant interaction term
 2      between O3 effect and city. As shown in Figure 7-2, the summary density curve of the city-
 3      stratified analysis has a peak at about the same value as the curve of the all cities analysis,
 4      suggesting a common O3 effect for all eight cities and small variation among them.  The
 5      unimodal shape of the density curve of the city-stratified analysis also indicates the absence of
 6      outlying cities.
 7           Mortimer et al. (2002) further noted that small declines in PEF may be of uncertain clinical
 8      significance, thus they calculated the incidence of > 10% declines in PEF.  A 5 to 15% change in
 9      FEVj has been expressed as having clinical importance to asthma morbidity (American Thoracic
10      Society, 1991; Lebowitz et al., 1987; Lippmann, 1988). Although greater variability is expected
11      in PEF measurements, a > 10% change in PEF also likely has clinical significance.  In Mortimer
12      et al. (2002), the incidence of > 10% declines in PEF was statistically significant, indicating that
13      O3 exposure may be associated with clinically significant changes in PEF in asthmatic children.
14      This study also observed that excluding days when  8-h avg O3 levels were less than 80 ppb
15      provided effect estimates which were similar to those when all days were included in the
16      analysis.
17           In Mexico City, two studies of asthmatic school children were carried out simultaneously
18      in the northern (Romieu et al., 1996) and southwestern sections of the city (Romieu et al., 1997).
19      In the northern study, 71 mildly asthmatic school children aged 5 to 13 years old, were followed
20      over time for daily morning (before breakfast) and afternoon (bedtime) PEF. In  single-pollutant
21      models, O3 at  0-,  1-, and 2-day lags was associated with diminished morning and afternoon PEF,
22      but only the 0-day lag morning effect was statistically significant. The O3 effect became
23      nonsignificant when PM2 5 was added to the model. In the southwestern study, 65 mildly
24      asthmatic children aged 5  to 13 years old were followed during the summer and winter for daily
25      morning and afternoon PEF.  Significant effects at 0- and  1-day lag O3 were observed on
26      afternoon PEF, with effects larger with a 1-day lag. Associations involving O3 were stronger
27      than those involving PM10. Several additional studies, both in the U.S. and in  other countries,
28      reported statistically significant associations between O3 exposure and decrements in PEF among
29      asthmatics (Gielen et al., 1997; Jalaludin et al., 2000; Just et al., 2002; Ross et al., 2002;
30      Thurston et al., 1997).
        January 2005                              7-33         DRAFT-DO NOT QUOTE OR CITE

-------
 1           Other epidemiologic studies did not find a significant O3 effect on the lung function of
 2      asthmatics. Delfino et al. (1997a) examined morning and evening PEF among 22 asthmatics
 3      ranging in age from 9 to 46 years, living in Alpine, CA. Daily ambient 12-h avg O3 (8 a.m.-
 4      8 p.m.) concentrations ranged from 34 to 103 ppb, with a mean value of 64 ppb. Unique to this
 5      study, personal O3 exposures were measured using 12-h passive O3 samplers that were worn by
 6      the subjects.  The personal 12-h avg O3 (8 a.m.-8 p.m.) concentrations, which had a mean value
 7      of 18 ppb, were much lower than the fixed-site ambient levels. Quantitative O3 results were not
 8      reported but researchers stated that no significant O3 effects were observed on morning and
 9      evening PEF.  In Hiltermann et al.  (1998), 60 nonsmoking adults aged 18 to 55 years in
10      Bilthoven, the Netherlands, were followed between July and October 1995 with morning and
11      afternoon PEF measurements. Ozone was associated with declines in PEF, but statistical
12      significance was not achieved.
13           Results from the multicities study by Mortimer et al.  (2002), as well as those from several
14      regional studies provide evidence of a significant relationship between O3 concentrations and
15      PEF among asthmatics.  Collectively, these  studies indicate that O3 may be associated with
16      declines in lung function in this potentially susceptible population.
17
18      Panels of healthy subjects
19           The effect of O3 on PEF in healthy subjects also was investigated in several studies.
20      During the summer of 1990, Neas et al. (1995) examined 83  children in Uniontown, PA and
21      reported twice daily PEF measurements. Researchers found that evening PEF was associated
22      with O3 levels weighted by hours spent outdoors. Using a  similar repeated measures design,
23      Neas et al. (1999) saw evidence for effects due to ambient  O3 exposure among 156 children
24      attending two summer day camps in the Philadelphia, PA area. Associations were found
25      between afternoon PEF (recorded before leaving camp) and same-day O3 concentrations, and
26      between morning PEF (recorded upon arrival at camp)  and previous-day O3 concentrations.
27      However, the relationship between PEF and O3 was statistically significant only when  a
28      cumulative lag period of 1 to 5 days was considered. Similarly, Naeher et al. (1999), in a sample
29      of 473 nonsmoking women (age 19 to 43 years) living in Vinton, VA, also showed the largest,
30      significant O3-related decrease in evening PEF with a 5-day cumulative lag exposure.
        January 2005                              7-34       DRAFT-DO NOT QUOTE OR CITE

-------
 1           Another study in southwestern Mexico City analyzed morning and afternoon PEF data
 2      collected from 40 school children aged 8 to 11 years (Gold et al., 1999). Subjects provided
 3      measurements upon arriving and before departing from school each day. Diminished PEF was
 4      associated with 1-day lag O3 concentrations, but the only statistically significant findings were
 5      obtained for PEF regressed on O3 concentrations with a cumulative 10-day lag period. This may
 6      imply a cumulative effect of O3, or it may reflect confounding by other time-varying factors.
 7      These results, however,  are in accord with controlled human exposure studies that have shown
 8      an attenuated decline in  pulmonary function with repeated days of O3 exposure (see Section 6.6,
 9      Repeated Exposure to O3), and with epidemiologic studies that have assessed lung function over
10      the course of the O3 season (Brauer et al., 1996; Kinney and Lippmann 2000).
11
12      7.2.4   Respiratory Symptoms
13           Studies published  over the past decade represent an improved new body of data on the
14      symptom effects of O3.  Respiratory symptoms are usually measured in the context of acute air
15      pollution field studies using questionnaire forms, or "daily diaries," that are filled out by study
16      subjects, usually without the direct supervision of research staff.  Questions address the daily
17      experience of coughing, wheezing, shortness of breath (or difficulty breathing), production of
18      phlegm, and others. While convenient and potentially useful in identifying acute episodes of
19      morbidity, measurements of daily symptoms are prone to a variety of errors. These include
20      misunderstanding of the meaning of symptoms, variability in individual interpretation of
21      symptoms, inability to remember symptoms if not recorded soon after their occurrence, reporting
22      bias if days  of high air pollution levels are identifiable by subjects, and the possibility of falsified
23      data. In spite of these potential problems, the ease of data collection has made daily symptom
24      assessment a common feature of field studies.  Many of the studies reviewed above for lung
25      function results also included measurements of daily symptoms.  Pearce et al. (1998) reports that
26      one advantage in the case of asthma panels is that the population is usually already familiar with
27      the symptom terms such as wheezing and cough. Delfino et al. (1998a) further states that the use
28      of repeated daily symptom diaries has additional advantages  of reducing recall bias given the
29      proximity of events and allowing health effects to be modeled with each subject serving as their
30      own control over time.  Also, study design can blind the participants from  the air pollution
        January 2005                              7-3 5        DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
aspect of the study.  Careful efforts by study staff can help ensure that the symptom diaries will
provide information that is less affected by the potential problems noted.
      Similar to studies of lung function, respiratory symptom studies can be divided into two
groups,  asthma panels or healthy subjects. Asthma panel studies are presented first.

Asthma panels
      Most studies examining respiratory symptoms related to O3 exposure focused on asthmatic
children. Among the health outcomes, of particular interest were those associated with asthma,
including cough, wheeze, shortness of breath, and increased medication use. Figures 7-3 and 7-4
present the probability density curves for O3-related cough and medication use among asthmatic
children from six studies (Gielen et al., 1997; Jalaludin et al., 2004; Just et al., 2002; Ostro et al.,
2001; Romieu et al., 1996, 1997).  Only single city/region studies that present odds ratios are
included in the figure for consistency.  Studies that present change in severity  of symptoms,
another informative health outcome,  are excluded from the figure since this expression differs
from indicating simple presence of symptoms. The study by Gent et al. (2003) also is excluded
from this figure as odds ratios for cough and mediation use were analyzed for  quintiles of O3
concentrations using the lowest quintile as the reference.
                             0.8
                                0.9        1         1.1       1.2
                                    Odds Ratio for Cough
                  1.3
1.4
       Figure 7-3.  Density curves of the odds ratios for the prevalence of cough among asthmatic
                    children.  Fifteen odds ratios from six studies (Gielen et al., 1997; Jalaludin
                    et al., 2004; Just et al., 2002; Ostro et al., 2001; Romieu et al., 1996,1997)
                    are standardized per 40 ppb increase in 1-h max O3 or equivalent.
       January 2005
                                          7-36
DRAFT-DO NOT QUOTE OR CITE

-------
               40 F^
                               0.9
      1          1.1          1.2
Odds Ratio for Bronchodilator Use
                1.3
1.4
        Figure 7-4.  Density curves of the odds ratios for the prevalence of extra bronchodilator
                    use among asthmatic children.  Nine odds ratios from six studies (Gielen
                    et al., 1997; Jalaludin et al., 2004; Just et al., 2002; Ostro et al., 2001; Romieu
                    et al., 1996,1997) are standardized per 40 ppb increase in 1-h max O3 or
                    equivalent.
 1           The various effect estimates for the association between O3 concentrations and cough are
 2     depicted as probability density curves (or density curves) in Figure 7-3. Each density curve
 3     represents an individual effect estimate and corresponding standard error. The use of density
 4     curves allows one to visualize better the distribution of O3-related effects on cough. Despite the
 5     variability in the individual effect estimates, there is consistency in the O3 effects as indicated by
 6     the considerable overlap in distributions. In general, the majority of the area under the density
 7     curves appears to be greater than an odds ratio of one, suggesting a positive association between
 8     O3 concentration and cough among asthmatic children. Figure 7-4 presents the density curves
 9     for the various effect estimates for O3-associated bronchodilator use.  Similar to cough, the
10     majority of the area under the density curves in this figure is greater than an odds ratio of one,
11     indicating that O3 may be associated with increased bronchodilator use in children  with mild to
12     severe asthma.
13           Among the studies reporting results for daily symptoms and asthma medication use,
14     several observed associations with O3 concentrations that appeared fairly robust (Delfino et al.,
15     2003; Desqueyroux et al.,  2002ab; Gent et al., 2003; Hilterman et al.,  1998; Just et al., 2002;
16     Mortimer et al., 2000, 2002; Newhouse et al., 2004; Romieu et al., 1996, 1997; Ross et al., 2002;
17     Thurston et al., 1997).
       January 2005
            7-37
DRAFT-DO NOT QUOTE OR CITE

-------
 1          Mortimer et al. (2002) reported morning symptoms in 846 asthmatic children from eight
 2     urban areas of the U.S. to be most strongly associated with a cumulative 4-day lag of O3
 3     concentrations in the NCICAS. The NCICAS used standard protocols which included
 4     instructing caretakers of the subjects to record symptoms in the daily diary by observing or
 5     asking the child (Mitchell et al., 1997). Symptoms reported included cough, chest tightness, and
 6     wheeze.  In the analysis pooling data from all eight cities, the odds ratio for the incidence of
 7     symptoms was 1.35 (95% CI: 1.04, 1.69) per 30 ppb increase in 8-h avg O3.  Excluding days
 8     when 8-h avg O3 (10 a.m.-6 p.m.) was greater than 80 ppb, the odds ratio was 1.37 (95% CI:
 9     1.02, 1.82) for incidence of morning symptoms. Figure 7-5 presents the density curves of the
10     odds ratios for the incidence of symptoms from the city-stratified analysis and that from the all
11     cities analysis. This figure confirms the regression results that there is a significant increase in
12     odds for incidence of symptoms, as the area under the density curve with an odds ratio greater
13     than one  is 99%. The unimodal distribution of the city-stratified summary density curve
14     indicates a lack of significant heterogeneity among the  eight cities.
15
                                                                  City-Stratified Analysis
                                                                  All City Analysis
                                          1                     1.5
                                 Odds Ratio for Incidence of Symptoms
       Figure 7-5.  Density curves of the odds ratios for the incidence of symptoms per 30 ppb
                   increase in 8-h avg O3 with a cumulative lag of 1 to 4 days for the individual
                   eight cities and the pooled average of all cities. Note that 99% and 76% of the
                   areas under the curves are greater than one for the pooled cities and
                   individual cities analyses, respectively.
       Source: Derived from Mortimer et al. (2002).
       January 2005
7-38
DRAFT-DO NOT QUOTE OR CITE

-------
 1           Another one of the larger studies was that of Gent and colleagues (2003), where
 2      271 asthmatic children under age 12 and living in southern New England were followed over
 3      6 months (April through September) for daily symptoms.  The data were analyzed for two
 4      separate groups of subjects, 130 who used maintenance asthma medications during the follow-up
 5      period and 141 who did not.  The need for regular medication was considered to be a proxy for
 6      more severe asthma. Not taking any medication on a regular basis and not needing to use a
 7      bronchodilator would  suggest the presence of very mild asthma.  Significant effects of 1-day lag
 8      O3 were observed on a variety of respiratory symptoms only in the medication user group. Both
 9      daily 1-h max and 8-h max O3 concentrations  were similarly related to symptoms such as chest
10      tightness and shortness of breath. Effects of O3, but not PM25, remained significant and even
11      increased in magnitude in two-pollutant models.  Some of the significant associations were noted
12      at 1-h max O3 levels below 60 ppb.  In contrast, no significant effects were observed among
13      asthmatics not using maintenance medication.  In terms of person-days of follow-up, this is one
14      of the larger studies currently available that address symptom outcomes in relation to O3, and
15      provides supportive evidence for effects of O3 independent of PM2 5.
16           Some international studies have reported significant symptoms associations with O3. The
17      incidence of asthma attacks was significantly  associated with  O3 concentrations in a group of
18      60 severe asthmatics (mean age 55 years) followed over a 13-month period in Paris
19      (Desqueyroux et al., 2002a).  In a similar study, Desqueyroux et al. (2002b) observed significant
20      O3-associated exacerbation of symptoms in 39 adult patients (mean age 67 years) with chronic
21      obstructive pulmonary disease (COPD). Interestingly, in contrast to the controlled human
22      studies (see Section 6.3.1, Subjects with COPD), the O3 effect appeared larger among subjects
23      who smoked and those with more severe COPD.  However, the low O3 concentrations
24      experienced during this study (summer mean 8-h max O3 of 41 |ig/m3 or approximately 21 ppb)
25      raise plausibility questions.  In  a study of 60 nonsmoking asthmatic adults (aged 18 to 55 years)
26      in Bilthoven, the Netherlands, Hilterman and colleagues (1998) reported significant associations
27      between O3 and daily symptoms of shortness of breath and pain upon deep inspiration.  The O3
28      associations were stronger than those of PM10, NO2, SO2, and  black smoke (BS). No differences
29      in response were evident between subgroups of subjects defined on the basis of steroid use or
30      airway hyperresponsiveness. Daily use of bronchodilators or  steroid inhalers was not found to
31      be associated with O3 in this study.

        January 2005                              7-39        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Other studies showed only limited or a lack of evidence for symptom increases associated
 2      with O3 exposure (Avol et al., 1998; Chen et al., 1998; Delfmo et al., 1996, 1997a, 1998a;
 3      Gielen et al., 1997; Jalaludin et al., 2004; Ostro et al., 2001; Taggart et al., 1996). Ostro et al.
 4      (2001) reported no associations between daily symptoms and ambient O3 concentrations in a
 5      cohort of 138 African-American children with asthma followed over 3 months (August to
 6      October) in Central Los Angeles and Pasadena, CA. However, the use of extra asthma
 7      medication was associated with 1-h max O3 concentrations at a 1-day lag. Delfmo and
 8      colleagues (1996) followed 12 asthmatic teens living in San Diego, CA for respiratory symptoms
 9      over a two-month period and saw no relationship with central site ambient O3. Personal O3
10      exposures measured with passive diffusion monitors were associated with the composite
11      symptom score and p2-agonist inhaler use, but the relationship with symptom score disappeared
12      when weekday/weekend differences were controlled in the statistical analysis. Study power was
13      likely compromised by the small sample size. This observation of stronger effects based on
14      personal monitoring is intriguing; it suggests that substantial gains in power may be achieved if
15      exposure misclassification is reduced through the use of personal exposure measurements rather
16      than central site O3 concentrations.  A similar study of 22 asthmatics in Alpine, CA observed no
17      effects of O3 on symptoms when personal O3 exposure was used as the exposure metric (Delfmo
18      et al., 1997a). However, a later study in the same location involving 24 subjects (Delfmo et al.,
19      1998a) did find an association between respiratory symptoms and ambient O3 exposure, with
20      stronger O3 effects experienced by asthmatics not on anti-inflammatory medication. In this
21      study, a binary symptom score was used, whereas the earlier study used a linear symptom score
22      of 0 through  6.
23           In conclusion, the various studies seem to indicate a positive association between O3
24      concentrations, and respiratory symptoms and increased medication use in asthmatics.  The
25      multicities study by Mortimer et al. (2002) provides an asthmatic population most representative
26      of the U.S., but several single city studies also add to the knowledge base.
27
28      Panels of healthy subjects
29           Fewer studies examined the effect of O3 on respiratory symptoms in healthy individuals.
30      Neas et al.  (1995) reported evening cough was  associated with O3 levels weighted by hours spent
31      outdoors in school children.  The study by Linn and colleagues (1996) of 269 school children in

        January 2005                            7-40        DRAFT-DO NOT QUOTE OR CITE

-------
 1      southern California reported no associations between respiratory symptoms and O3, but subjects
 2      were exposed to fairly low O3 concentrations as determined using personal monitors.  Gold et al.
 3      (1999) examined symptoms in 40 healthy children in southwest Mexico City. Pollutant
 4      exposures were associated with increased production of phlegm in the morning, although the
 5      effects of the air pollutants (PM2 5, PM10, and O3) could not be separated in multipollutant
 6      models.  Hoek and Brunekreef (1995) did not find a consistent association between ambient O3
 7      levels, and prevalence and incidence of respiratory symptoms in children living in two rural
 8      towns in the Netherlands. Collectively, these studies indicate that there is no consistent evidence
 9      of an association between O3 and respiratory symptoms among healthy children.
10
11      7.2.5    Acute Airway Inflammation
12          Acute airway inflammation has been shown to occur among adults exposed to 80 ppb O3
13      over 6.6 hours with exercise in controlled chamber studies (Devlin et al., 1991). Kopp and
14      colleagues (1999) attempted to document inflammation of the upper airways in response to
15      summer season O3 exposures by following a group of 170  school children in two towns in the
16      German Black Forest from March to October of 1994. To assess inflammation, the investigators
17      collected nasal lavage samples at 11 time points spanning the follow-up period. The nasal
18      lavage samples were analyzed for markers of inflammation, including eosinophil  cationic
19      protein, albumin, and leukocyte counts.  Subjects who were sensitized to inhaled allergens were
20      excluded. When analyzed across the entire follow-up period, no association was detected
21      between upper airway inflammation and O3 concentrations. More detailed analysis showed that
22      the first significant O3 episode of the summer was followed by a rise in eosinophil cationic
23      protein levels, however, subsequent and even higher O3 episodes had no effect. These findings
24      suggest an adaptive response of inflammation in the nasal airways that is consistent with
25      controlled human studies (see Section 6.9, Effects  of Inflammation and Host Defense).
26          Frischer and colleagues (1993) collected nasal lavage samples from 44 school children in
27      Umkirch, Germany the morning after "low" O3 days (< 140 |ig/m3 or approximately 72 ppb) and
28      "high" O3 days (> 180 |ig/m3 or approximately  93 ppb) to measure levels of biochemical markers
29      of inflammation. The researchers found that higher O3 levels were significantly associated with
30      increased polymorphonuclear leukocyte counts in all children, and increases in
31      myeloperoxydases and eosinophilic cation proteins among children without symptoms of rhinitis

        January 2005                             7-41       DRAFT-DO NOT QUOTE OR CITE

-------
 1      (n = 30).  These results indicated that O3 was associated with inflammation in the upper airways.
 2      Frischer et al. (1997) further investigated whether hydroxyl radical attacks played a role in
 3      mediating the O3-associated inflammatory response of the airways.  Ortho- and/>ara-tyrosine
 4      levels were measured in the nasal lavage samples and the ortholpara radical ratio was used to
 5      determine the generation of hydroxyl radicals. Significant increases in the ortholpara ratio were
 6      observed  on days following high ambient O3 levels. However, the ortholpara ratio was not
 7      related to polymorphonuclear leukocyte counts, suggesting that there was no detectable
 8      relationship between hydroxyl radical attacks and the inflammatory response seen in these
 9      children.  Similar to the study by Kopp et al. (1999), the ortholpara ratio decreased at the end of
10      the summer although O3 concentrations were still high, providing additional evidence for a
11      possible adaptive response. These findings, however, do not preclude the possibility that other
12      unmeasured effects, including cell damage or lower airway responses, may have occurred with
13      ongoing summer season exposures. In fact, a study of joggers repeatedly exposed to O3 while
14      exercising over the summer in New York City suggested that cell damage may occur in the
15      absence of ongoing inflammation (Kinney et al., 1996).
16           In two Mexico City studies by Romieu et al. (1998, 2002), the effect of antioxidant
17      supplements  on the association between O3 and lung function in outdoor workers and asthmatic
18      children was investigated.  Romieu and colleagues (1998) observed significant inverse
19      associations between O3 and lung function parameters, including FVC, FEVl3 and FEF25.75
20      (forced expiratory flow at 25 to 75% of FVC), among outdoor workers who were on the placebo,
21      but not among those taking the antioxidant supplement during the 1st phase of testing.  Likewise,
22      O3 concentrations were associated with declines in lung function among children with moderate
23      to severe  asthma who were on the placebo, but no associations were found among those who
24      were taking the vitamin C and E supplement (Romieu et al., 2002).  These results indicate that
25      supplementation with antioxidants may modulate the impact of O3 exposure on the small airways
26      of two potentially susceptible populations, outdoor workers and children with moderate to severe
27      asthma. In a further analysis, genetic factors were found to contribute to the variability between
28      individuals in the effects of O3 on lung function (Romieu et al., 2004). Individuals with
29      polymorphism of the glutathione S-transferase gene (GSTM1 null genotype) lack glutathione
30      transferase enzyme activity, which plays an important role in protecting  cells against oxidative
31      damage.  Results from this analysis indicate that asthmatic children with GSTM1 null genotype

        January 2005                             7-42        DRAFT-DO NOT QUOTE OR CITE

-------
 1      were found to be more susceptible to the impact of O3 exposure on small airways function.
 2      Romieu et al. (2004) noted that supplementation with the antioxidant vitamins C and E above the
 3      minimum daily requirement might compensate for the genetic susceptibility.
 4
 5      7.2.6  Acute O3 Exposure and School Absences
 6           The association between school absenteeism and ambient air pollution was assessed in two
 7      studies (Chen et al., 2000; Gilliland et al., 2001).  In the study by Chen and colleagues (2000),
 8      daily school absenteeism was examined in 27,793 students (kindergarten to 6th grade) from
 9      57 elementary school students in Washoe County, NV over a two-year period. One major
10      limitation of this study was that the percent of total daily absences was the outcome of interest,
11      not illness-related absences, as reasons for absences were not noted in all schools.  In models
12      adjusting for PM10 and CO concentrations, ambient O3 levels were significantly associated with
13      school absenteeism.  With a distributed lag of 1 to 14 days,  O3  concentrations were associated
14      with a 10.41% (95% CI: 2.73, 18.09) excess rate of school absences per 40 ppb increase in 1-h
15      max O3. PM10 and CO concentrations also were significantly associated with school
16      absenteeism, however, the effect estimate for PM10 was negative.  The inverse relationship
17      between O3 and PM10 may have partially attributed to the negative association observed between
18      PM10 and school absenteeism.
19           Ozone-related school absences also were examined in a study of 1,933 4th grade students
20      from 12 southern California communities participating in the Children's Health Study (Gilliland
21      et al., 2001).  Due to its size and comprehensive characterization of health outcomes, this study
22      is especially valuable in assessing the effect of O3 on illness-related school absenteeism in
23      children.  The study spanned a period, January through June 1996, that captured a wide range of
24      exposures while staying mostly below the highest levels observed in the summer season.
25      All school absences that occurred during this period were followed up with phone calls to
26      determine whether they were  illness-related.  For illness-related absences, further questions
27      assessed whether the illness was respiratory or gastrointestinal, with respiratory symptoms
28      including runny nose/sneeze,  sore throat, cough, earache, wheezing, or asthma attacks.  Multiple
29      pollutants were measured at a central site in each of the 12 communities. The statistical analysis
30      controlled for temporal cycles, day of week, and temperature, and expressed exposure as a
31      distributed lag out to 30 days.  Some concern exists regarding the possibility of residual seasonal

        January 2005                               7-43         DRAFT-DO NOT QUOTE OR CITE

-------
 1      confounding given the six-month time span of the monitoring period.  Significant associations
 2      were found between the 30-day distributed lag of 8-h avg O3 (10 a.m.-6 p.m.) and all absence
 3      categories. Larger O3 effects were seen for respiratory causes (147% increase per 30 ppb
 4      increase in 8-h avg O3) than for nonrespiratory causes (61% increase). Among the respiratory
 5      absences, larger effects were seen for lower respiratory diseases with wet cough than for upper
 6      respiratory diseases. PM10 was only associated with upper respiratory disease absences.
 7           Results from Chen et al. (2000) and, more notably, Gilliland et al. (2001) indicate that
 8      ambient O3 concentrations, lagged over two to four weeks, are significantly associated with
 9      school absenteeism,  particularly respiratory illness-related absences. These two studies on
10      school absenteeism were conducted in communities in Nevada and southern California,
11      however the  results are most likely representative of U.S. populations.
12
13      7.2.7   Cardiac Physiologic Endpoints
14           Limited pathophysiologic air pollution studies have examined cardiac physiologic
15      endpoints (Table AX7-2 in Chapter 7 Annex).  Several studies examined associations between
16      PM exposures and gaseous pollutants, and various measures of heart beat rhythms in panels of
17      elderly subjects as discussed in the 2004 PM AQCD (Section 8.3.1). Decreased heart rate
18      variability has been identified as a predictor of increased cardiovascular morbidity and mortality.
19      One study examined the increased risk of myocardial infarction and pollutants (Peters et al.,
20      2001).  Lack of consistency  in the limited studies argued for caution in regards to drawing
21      conclusions on the relationship between cardiovascular outcomes and PM exposure. Among
22      these studies, Gold et al. (2000; reanalysis Gold et al., 2003) and Peters et al. (2000a, 2001)
23      discussed limited evaluation of a potential role for O3 exposure.  In addition, two recent studies
24      provided limited evidence for an association between O3 concentrations  and heart rate variability
25      in primarily elderly populations (Holguin et al.,  2003; Park et al., 2004). Two related studies by
26      Rich et al. (2004)  and Vedal et al. (2004) examined the relationship between various air
27      pollutants, including O3, and cardiac arrhythmias using two different study designs, but both did
28      not find any consistent evidence that exposure to air pollution affected the risk of arrhythmias.
29      Two limited  controlled human exposure studies with cardiovascular outcomes (Gong et al.,
30      1998a;  Superko et al., 1984), described in Chapter 6, Section 6.3.4, provide no supporting data.
        January 2005                              7-44        DRAFT-DO NOT QUOTE OR CITE

-------
 1           The above panel studies with small numbers of subjects had limited ability to adequately
 2      test the hypothesis of heart rate variability. A recent large population-based study, the first in
 3      this field, examined PM10, O3, and other gaseous air pollutants, and their potentially adverse
 4      effects on cardiac autonomic control (Liao et al., 2004).  Liao et al. investigated short-term
 5      associations between ambient pollutants and cardiac autonomic control from the 4th cohort
 6      examination (1996-1998) of the population-based Atherosclerosis Risk in Communities Study.
 7      PM10 (24-h avg) and O3 exposure (8-h avg, 10 a.m.-6 p.m.) one day prior to the randomly
 8      allocated examination date were used.  They calculated 5-minute heart rate variability indices
 9      between 8:30 a.m. and 12:30 p.m. and used logarithmically-transformed data on high-frequency
10      (0.15 to 0.40 Hz) and low-frequency (0.04 to 0.15 Hz) power, standard deviation of normal
11      R-R intervals, and mean heart rate.  The effective sample sizes for PM10 and O3 were 4,899 and
12      5,431, respectively, from three U.S. study centers in North Carolina, Minnesota, and Mississippi.
13      PM10 concentrations measured one day prior to the heart rate variability measurements were
14      inversely associated with both frequency and time domain heart rate variability indices.
15      Ambient O3 concentrations were inversely associated with high-frequency power among whites.
16      Consistently more pronounced associations were suggested between PM10 and heart rate
17      variability among persons with a history of hypertension. These findings were cross-sectionally
18      derived from  a population-based sample and reflect only the short-term effects of air pollution
19      on heart rate variability.  When the regression coefficients for each individual pollutant model
20      were compared, the effects for PM10 was considerably larger than the effects for gaseous
21      pollutants such as O3.  While these data are supportive of the hypothesized  air pollution-heart
22      rate variability-cardiovascular disease pathway at the population level, replication of these
23      interactions in other studies is needed before any conclusions can be made.
24
25      7.2.8    Summary of Field Studies Assessing Acute O3 Effects
26         •   Results from recent field/panel studies support the evidence from clinical studies that
              acute O3 exposure is associated with a significant effect on lung function, as indicated
              by decrements in FEVj, FVC, and PEF. The declines in lung function were noted
              particularly in children and asthmatics.
27         •   Limited evidence suggests that more time spent outdoors, higher levels of exertion,
              and the related increase in O3 exposure may potentiate the risk of respiratory effects.
              In addition to children and asthmatics, adults who work or exercise outdoors may be
              particularly susceptible to O3-associated health effects.

        January 2005                               7-45        DRAFT-DO NOT QUOTE OR CITE

-------
 4
 5
             Many new studies have examined the association between O3 concentrations and a
             wide variety of respiratory symptoms (e.g., cough, wheeze, production of phlegm,
             and shortness of breath). Collectively, the results indicate that acute exposure to O3
             is associated with increased respiratory symptoms and increased as-needed medication
             use in children and asthmatics.
             Additional panel studies investigated the effect of O3 on other health outcomes,
             including school absences, and markers of inflammation and oxidative damage.
             Ozone exposure was associated with significant increases in respiratory-related school
             absences, as well as increased inflammation and generation of hydroxyl radicals in the
             upper airways. Use of antioxidant supplements was found to  diminish the O3 effect on
             lung function.
             Few field studies have examined the association between O3 and cardiac physiologic
             outcomes. The current evidence is rather limited but supportive of a potential effect on
             heart rate variability.  Additional studies need to be performed before any conclusions
             can be made regarding an O3 effect on cardiovascular outcomes.
 6     7.3    EFFECTS OF OZONE ON DAILY EMERGENCY DEPARTMENT
 7            VISITS AND HOSPITAL ADMISSIONS
 8     7.3.1   Summary of Key Findings on Studies of Emergency Department Visits
 9             and Hospital Admissions from the 1996 O3 AQCD
10          In the 1996 O3 AQCD, aggregate population time-series studies of O3-related health effects
11     provided relevant evidence of acute responses, even below a 1-h max O3 of 0.12 ppm.
12     Emergency room visits and hospital admissions were examined as possible outcomes following
13     exposure to O3. In the case of emergency room visits, the evidence was limited (Bates et al.,
14     1990; Cody et al., 1992; Weisel et al., 1995; White et al., 1994), but results generally indicated
15     an O3 effect on morbidity. The strongest and most consistent evidence of O3 effects, at levels
16     both above and below 0.12 ppm 1-h max O3, was provided by the multiple studies that had been
17     conducted on summertime daily hospital admissions for respiratory causes in various locales in
18     eastern North America (Bates and Sizto, 1983, 1987, 1989; Burnett et al., 1994; Lipfert and
19     Hammerstrom, 1992; Thurston et al., 1992, 1994).  These studies consistently demonstrated that
20     O3 air pollution was associated with increased hospital admissions, accounting for roughly one to
21     three excess respiratory hospital admissions per million persons with  each 100 ppb increase in
22     1-h max O3.  This association had been shown to remain even after statistically controlling for
23     the possible confounding effects of temperature and copollutants (e.g., FT, SO4 2, PM10), as well

       January 2005                             7-46       DRAFT-DO NOT QUOTE OR CITE

-------
 1      as when considering only days with 1-h max O3 concentrations below 0.12 ppm. Furthermore,
 2      these results implied that O3 air pollution could account for a substantial portion of summertime
 3      hospital admissions for respiratory causes on the most polluted days.  Overall, the aggregate
 4      population time-series studies considered in the 1996 O3 AQCD provided strong evidence that
 5      ambient exposures to O3 can cause significant exacerbations of preexisting respiratory disease in
 6      the general public at concentrations below 0.12 ppm.
 7
 8      7.3.2  Review of Recent Studies of Emergency Department Visits for
 9             Respiratory Diseases
10           Emergency department visits represent an important acute outcome that may be affected by
11      O3 exposures. Morbidities that result in emergency department visits are closely related to, but
12      are generally less severe than, those that result in unscheduled hospital admissions.  In many
13      cases, acute health problems are successfully treated in the emergency department; a subset of
14      more severe cases that present initially to the emergency department may require admission to
15      the hospital.
16           Several studies have been published in the past decade examining the temporal
17      associations between O3 exposures and emergency department visits for respiratory diseases
18      (Table AX7-3 in Chapter 7 Annex).  Total respiratory causes for emergency room visits may
19      include asthma, pneumonia, bronchitis, emphysema, other upper and lower respiratory infections
20      such as influenza, and a few other minor categories.  Asthma visits typically dominate the daily
21      incidence counts. Chronic bronchitis and emphysema often are combined to define COPD,
22      which is a prominent diagnosis among older adults with lung disease. Figure 7-6 presents %
23      changes in emergency department visits for asthma, with results expressed in standardized
24      increments. Results from all lags presented are included in the figure. Weisel et al. (2002) was
25      excluded as relative risks were not presented and could not be  estimated. Among the U.S.
26      studies, there was one multicity study which examined three cities in Ohio (Jaffe et  al., 2003).
27      Several presented Atlanta, GA data.  In general, O3 effect estimates from summer only analyses
28      tended to be positive and larger compared to results from cool  season or all year analyses.
29           Among studies with adequate controls for seasonal patterns, many reported at least one
30      significant positive association involving O3.  These studies examined emergency department
31      visits for total respiratory complaints (Delfmo et al., 1997b, 1998b; Hernandez-Garduno et al.,

        January 2005                              7-47        DRAFT-DO NOT QUOTE OR CITE

-------
                                             % Change in Emergency Department Visits for Asthma
to
o
o
oo


Jaffe etal. (2003):
Cleveland OH (aqe 5-34)
Columbus OH (aqe 5-34)


Cincinnati, Cleveland, —
and Columbus, OH (age 5-34)
Friedman etal. (2001):
Atlanta, GA (age 1-1 6)
Peel etal. (2004):
Atlanta, GA (all ages)
Tolbert etal. (2000):
Atlanta, GA (age 0-1 6)
Stieb etal. (1996):
St. John, Canada (all ages)
Atkinson et al. (1999a): 	
London, England (all ages)
Atkinson etal. (1999a): 	
London, England (age 0-14)
Atkinson etal. (1999a):
London, England (agel 5-64)
Thompson etal. (2001):
Belfast, N. Ireland (children)




Castellsagueetal. (1995):
Barcelona, Spain (age 15-65)
Tobias etal. (1999): 	
Barcelona, Spain (age >14)
Teniasetal. (1998):
Valencia, Spain (age > 14)


-20 0 20 40 60 80 100 120 140 160 180 20C
i i i i i i i i i i i i i i i i i i i i i i i
| U.S. and Canada
X All year
£ bg ° 9 Warm
• I -in "3
lag J O Cool

• Hn n
0 lag 0-1

lag u-z
-X- lag 0-2
• lag 0-2 (SE not given)
— •— lag 1
• Inn °

| Europe |
	 X — lagO
—X— Iag1
^ laa 0
x" Ian D-1
X bg 0 °

X lay 0-3
« lin 0 1

lag 0-1
A lin n
v |uy u
— X 	 lagO
V inn 1

©hn 1

     Figure 7-6.  Ozone-associated % change (95% CI) in emergency department visits for asthma per 40 ppb increase in 1-h max

                O3 or equivalent.

-------
 1      1997; Ilabaca et al., 1999; Jones et al., 1995; Lin et al., 1999), asthma (Friedman et al., 2001;
 2      Jaffe et al., 2003; Stieb et al., 1996; Tenias et al., 1998; Tobias et al., 1999; Tolbert et al., 2000;
 3      Weisel et al., 2002), and COPD (Tenias et al., 2002).
 4          One recent study examined emergency department visits for total and cause-specific
 5      respiratory diseases in Atlanta, GA over an 8-year period (Peel et al., 2004). A distributed lag of
 6      0 to 2 days was specified a priori. Ozone concentrations were significantly associated with
 7      emergency department visits for total respiratory diseases and upper respiratory infections in all
 8      ages. A marginally significant association was observed with asthma visits (2.6% excess risk
 9      per 30 ppb increase in 8-h max O3), which became stronger when analysis was restricted to the
10      warm months (3.1% excess risk).  In multipollutant models adjusting for PM10, NO2 and CO, O3
11      was the only pollutant that remained significantly associated with upper respiratory infections.
12      Another large asthma emergency department study was carried out during the months of May
13      through September from  1984 to 1992 in St. John, New Brunswick, Canada (Stieb et al., 1996).
14      Effects were examined separately among children aged less than 15 years and in persons aged
15      15 years and older. A significant effect of O3 on emergency department visits was reported
16      among persons 15 years and older. There was evidence of a threshold somewhere in the range
17      below a 1-h max O3 of 75 ppb. A study in Valencia, Spain from 1994 to 1995 observed that
18      emergency room visits for asthma among persons over 14 years old were robustly associated
19      with relatively low O3 levels (median 1-h max O3 of 62.8 |ig/m3 or approximately 32.4 ppb)
20      (Tenias et al., 1998).  The excess risk of asthma emergency room visits was larger in the warm
21      season (May to October), 85% excess risk per 40 ppb increase in 1-h max O3, compared to the
22      cool season (November-April), 31% excess risk (Tenias et al., 1998).
23          Among the studies that observed a statistically significant association between O3 and
24      emergency department visits for respiratory outcomes, O3 effects were found to be robust to
25      adjustment for PM10, NO2, SO2, and black smoke (Lin et al., 1999; Peel et al., 2004; Tenias et al.,
26      1998). One study by Tolbert and colleagues (2000) observed that the significant univariate
27      effects of both O3 and PM10 on pediatric asthma emergency department visits in Atlanta, GA
28      became non-significant in two-pollutant regressions, reflecting the high correlation between the
29      two pollutants (r = 0.75).
30          For several other "positive"  studies with total respiratory and asthma outcomes,
31      inconsistencies confound an interpretation of likely causal effects. For example, in a Montreal,

        January 2005                             7-49        DRAFT-DO NOT QUOTE OR CITE

-------
 1      Canada study, O3 effects on total respiratory emergency department visits were seen in a short
 2      data series from the summer of 1993 but not in a similar data series from the summer of 1992
 3      (Delfino et al.,  1997b). The  significant 1993 results were seen only for persons older than
 4      64 years, in spite of greater asthma prevalence among children. A very similar analysis of two
 5      additional  summers (1989 and 1990) revealed an O3 association only for 1989 and again only in
 6      persons over 64 years old (Delfino et al., 1998b).  An analysis of data on respiratory emergency
 7      department visits from June to August of 1990 in Baton Rouge, LA reported O3 effects in adults,
 8      but not in children or among the elderly (Jones et al., 1995).
 9           Tobias and colleagues  (1999) showed that regression results for asthma emergency
10      department visits could be quite sensitive to methods used to control for asthma epidemics.
11      Ozone was associated with the outcome variable in only one of eight models tested. An Atlanta,
12      GA study by Zhu et al. (2003) examined asthma emergency department visits in children during
13      three summers using Bayesian hierarchical modeling to address model variability. Data was
14      analyzed at the zip code level to account for spatially misaligned longitudinal data. Results
15      indicated a positive, but nonsignificant relationship between O3 and emergency room visits
16      for asthma.
17           Other studies also reported nonsignificant findings for O3 (Atkinson et al., 1999a;
18      Castellsague et al., 1995; Chew et al., 1999). One study by Thompson and colleagues (2001)
19      in Belfast,  Northern Ireland observed a significant 21% decrease in risk of childhood asthma
20      admissions per 20 ppb increase in 24-h avg O3 in the cold season (November-April). After
21      adjusting for benzene  levels, O3 was no longer associated with asthma emergency department
22      visits. The inverse relationship of O3 with benzene concentrations (r = -0.65), and perhaps with
23      other pollutants, might have  produced the apparent protective effect of O3. No significant O3
24      effect was  found in the warm season (May-October). The O3 levels were low in both seasons,
25      with a mean 24-h avg  O3 concentration of 18.7 ppb in the warm season and  17.1 ppb in the cold
26      season. Atkinson et al. (1999a) in London, England also did not find an association between O3
27      and emergency department visits at a mean 8-h max O3 concentration of 17.5 ppb. Several other
28      emergency department studies looking at O3 are more difficult to interpret due to inadequate
29      control for seasonal patterns, very low O3 levels, or because no quantitative results were shown
30      for O3 (Buchdahl et al., 1996, 2000; Garty et al., 1998; Holmen et al., 1997; Lierl and Hornung,
31      2003; Lipsett et al., 1997; Nutman et al., 1998).

        January 2005                             7-50        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Although several studies found a significant association between O3 concentrations and
 2      emergency department visits for respiratory causes, some inconsistencies were observed. The
 3      inconsistencies may be attributable, at least partially, to differences in study design and model
 4      specifications among the various studies. For example, ambient O3 concentrations, length of the
 5      study period, and statistical methods used to control confounding by seasonal patterns and
 6      copollutants appear to affect the observed O3 effect on emergency department visits. In general,
 7      an excess risk of emergency department visits was observed during the summer season when O3
 8      concentrations were higher.
 9
10      7.3.3   Studies of Hospital Admissions for Respiratory Diseases
11           Hospital admissions represent a medical response to a serious degree of morbidity for a
12      particular disease. Scheduled hospitalizations are planned in advance when a particular clinical
13      treatment is needed. However, unscheduled admissions are ones that occur in response to
14      unanticipated disease exacerbations and are more likely to be affected by environmental factors,
15      such as air pollution. As such, the hospital admissions studies reviewed here focused
16      specifically on unscheduled  admissions. Study details and results from hospital admissions
17      studies published over the past decade are summarized in Table AX7-4 (in the Chapter 7
18      Annex). As a group, these hospitalization studies tend to be larger in terms of geographic and
19      temporal coverage, and indicate results that are generally more consistent than those reviewed
20      above for emergency department visits.  The following aspects of these studies should be
21      considered in comparing results: (1) difference in type of respiratory diseases for hospital
22      admission; (2) analysis by season versus all year; (3) O3 only versus multipollutant models;
23      (4) age of study population;  (5) number of exposure lag days; (6) single-city versus multicity
24      studies; (7) mean level of O3 during study; (8) length of study (e.g., < 5 years versus > 5 years);
25      and (9) type of study (e.g., case-crossover versus time-series).
26           Figures 7-7 through 7-9 present risk estimates from all total respiratory hospital admission
27      studies.  Burnett et al. (1995), which did not present quantitative results for O3, and Yang et al.
28      (2003), which only presented odd ratios, were excluded from the figure. In cases where multiple
29      lags were presented, the multiday lag was selected to represent the cumulative effect from all
30      days examined.  For Luginaah et al. (2004), cumulative lags are not analyzed, thus the effect
31      estimates from a 1-day lag are included in this figure. For studies that presented risk estimates

        January 2005                               7-51         DRAFT-DO NOT QUOTE OR CITE

-------
                  Linn et al. (2000): Los Angeles,CA (age 30+)
                    Gwynn et al. (2000): Buffalo, NY (all ages)
                Burnett et al. (2001): Toronto, Canada (age <2)
     Luginaah et al. (2004): Windsor, Canada (all ages, males only)
   Luginaahetal. (2004): Windsor, Canada (all ages,females only)
             Atkinson et al. (1999b): London, England (all ages)
            Atkinson et al. (1999b): London, England (age 0-14)
           Atkinson etal.(1999b):London,England (age 15-64)
            Atkinson etal.(1999b): London, England (age 65+)
          Poncede Leon etal.(1996):London,England (all ages)
         Ponce de Leon et al. (1996): London, England (age 0-14)
        Ponce de Leon et al. (1996): London, England (age 15-64)
         Ponce de Leon et al. (1996): London, England (age 65+)
           Prescott et al. (1998): Edinburgh, Scotland (age <65)
           Prescott et al. (1998): Edinburgh, Scotland (age 65+)
   Schouten etal. (1996): Amsterdam, the Netherlands (age 15-64)
    Schouten et al. (1996): Amsterdam, the Netherlands (age 65+)
   Schouten etal. (1996): Rotterdam, the Netherlands (age 15-64)
               Hagen et al. (2000): Drammen, Norway (all ages)
              Oftedal et al. (2003): Drammen, Norway (all ages)


         Gouveia and Fletcher (2000a):Sao Paulo, Brazil (age<5)  —


        Petroeschevskyetal.(2001):Brisbane,Austrailia (all ages)
        Petroeschevsky et al. (2001): Brisbane, Austrailia (age 0-4)
       Petroeschevsky et al. (2001): Brisbane, Austrailia (age 5-14)
      Petroeschevsky etal. (2001): Brisbane, Austrailia (age 15-64)
       Petroeschevsky et al. (2001): Brisbane, Austrailia (age 65+)
                   Wongetal.(1999a): Hong Kong (all ages)
                   Wong etal.(1999a): Hong Kong (age 0-4)
                  Wong etal.(1999a): Hong Kong (age 5-64)
                   Wong et al. (1999a): Hong Kong (age 65+)
                                                         %  Change in Respiratory Hospitalization
                                                          -20         -10         0          10         20         30
                                                           I      I      I      I      I      I      I      I      I       I      I
                                                       U.S. and Canada
                                                       Europe
                  lagO
                                                       Asia
                               lagO
                              9- Iag1
                                                        lag 0-4
                                                        Iag1
                                            Iag1
                                    lagO
                              -•	  lag 0-2
                              	 lag 0-3
                                                       lag 0-3
                                     lag 0-2
                                        lag 0-1
                                                    lag 0-5
                              - lagO
                               lagO
                                   lagO
                                           lag 2
                                                   lagO
                                                          lag 2
                          lag 3-
                                      —•	  lag 0-3
                                      -•	  lag 0-3
                                      	•	lag 0-3
                               lag 0-3
Figure 7-7.     Ozone-associated % change (95% CI) in total respiratory hospitalizations
                   for all year analyses per 40 ppb increase in 1-h max O3 or equivalent.
January 2005
7-52
DRAFT-DO NOT QUOTE OR CITE

-------
                                                                            % Change in Respiratory Hospitalizations
to
o
o
H
6
o

o
H
O

O
H
W
O

O
HH
H
W
                                    Linn etal. (2000):
                             Los Angeles, CA (age 30+)

                                Schwartz etal. (1996):
                              Cleveland, OH (age 65+)

                                 Burnett etal.(1997a):
                            16 Canadian Cities (all ages)

                                Burnett etal.(1997b):
                            Toronto, Canada (all ages)

                                 Burnett etal. (2001):
                             Toronto, Canada (age < 2)
                           Ponce de Leon etal.(1996):
                            London, England (all ages)

                           Ponce de Leon etal.(1996):
                           London, England (age 0-14)

                           Ponce de Leon etal.(1996):
                          London, England (age 15-64)

                           Ponce de Leon etal.(1996):
                           London, England (age 65+)

                                Schouten etal.(1996):
                 Amsterdam,the Netherlands (age 15-64)

                                Schouten etal.(1996):
                  Amsterdam, the Netherlands (age 65+)

                                Schouten etal.(1996):
                  Rotterdam, the Netherlands (age 15-64)

                                Schouten etal.(1996):
       Amsterdam and Rotterdam, the Netherlands (all ages)

                                 Wongetal.(1999a):
                                Hong Kong (all ages)
0
|u.S. and Canada
— €
— €


PI

| Europe]















-e
|Asia



0 20 40 60 80 10C
i i i i i i i i i i

-K-
£«- lagO
• lin 1 °
g 	 * lag 1
• I--, H o




*• lag 0-2
A Ian 0-2

0 lag 0 °

• inn n °


K Inn n °

vv
• Ian 0 1

3
• Ian O-^

X All year
h -•" lag 2
h • Warm
x O Cool

n lag 0-3

      Figure 7-8.   Ozone-associated % change (95% CI) in total respiratory hospitalizations by season per 40 ppb increase in 1-h
                      max O3 or equivalent.

-------
                             -20
% Change in Respiratory Hospitalization
               0         10       20
              Burnett etal.(1997b):
                 Toronto, Canada
          (all ages, warm season only) —
                     with PM2 5
                     withPM10
                     with PM10_25

               Burnett etal. (2001):
                 Toronto, Canada
          (age <2, warm season only) -
                     with PM2 5
                     With PM10_25-

          Ponce de Leon etal. (1996):
           London, England (all ages) -
                        with BS-

                Hagen etal.(2000):
          Dram men, Norway (all ages)-
                       withPM10-

         Gouveia and Fletcher (2000a):
            Sao Paulo, Brazil (age <5) -
                      withPM10-

          Petroeschevsky etal. (2001):
          Brisbane, Australia (all ages)-
                with nephelometer -

          Petroeschevsky etal. (2001):
        Brisbane, Australia (ages 15-64)-
                with nephelometer -

                Wongetal.(1999a):
               Hong Kong (all ages)-
                      withPM10-
        -10
          I
 10
_l	
30
 I
| Europe
             -e-
 Latin America
 Australia
 Asia
                                -e-
                                         lag 1-3
                                                -e-
                          Iag1
                            lagO
                           lagO
                         -e-
                                   Iag2
                                -e-
                                               Iag2
                               -e-
                                       lag 0-3
 40
	I	
50
       Figure 7-9.   Ozone-associated % change (95% CI) in total respiratory hospitalizations
                     with adjustment for PM indices per 40 ppb increase in 1-h max O3 or
                     equivalent.  Analyses performed using all year data unless noted otherwise.
1      from all four seasons, only the summer and winter estimates are presented. Figure 7-7 plots the
2      relative risk estimates and 95% CIs from 13 studies that analyzed all year data. The

3      preponderance of positive risk estimates, with some that are statistically significant, is readily
4      apparent.  The impact of seasonal and multipollutant analyses were further examined in Figures
5      7-8 and 7-9. In Figure 7-8, it appears that the warm season estimates, collectively, tend to be
       January 2005
                      7-54
       DRAFT-DO NOT QUOTE OR CITE

-------
 1      larger, positive values compared to all year and cool season estimates. Most of the negative
 2      estimates were from analyses using cool season data only, which might reflect the inverse
 3      correlation between O3 and copollutants, namely PM, during that season.  Figure 7-9 compares
 4      the risk estimates from models with and without adjustment for PM indices.  This figure
 5      indicates that O3 risk estimates are fairly robust to PM adjustment in the all year and warm
 6      season only data. None of the studies examined PM-adjusted O3 risk estimates in cool season
 7      only data.
 8           The most robust and informative results on the effects of O3 on respiratory hospital
 9      admissions are those from studies carried out using a consistent analytical methodology across a
10      broad geographic area (Anderson et al., 1997; Burnett et al., 1995, 1997a). These  studies have
11      all reported a significant O3 effect on respiratory hospital admissions.  The largest such study
12      to-date was carried out using data on all-age  respiratory hospital admissions from 16 Canadian
13      cities with populations exceeding 100,000 covering the period 1981 to 1991 (Burnett et al.,
14      1997a).  In addition to O3, the authors evaluated health effects of SO2, NO2, CO, and coefficient
15      of haze (a surrogate for black carbon particle concentrations). Pooling the 16 cities, a significant
16      positive association was observed between respiratory hospital admissions and the 1-day lag O3
17      concentration in the spring (5.6% excess risk per 40 ppb increase in 1-h max O3) and summer
18      (6.7%).  The results for fall were also positive, though of smaller magnitude (3.8%).  There was
19      no evidence for an O3 effect in the winter season (-0.8%).  Control outcomes related to blood,
20      nervous system, digestive system, and genitourinary system disorders were not associated with
21      O3.  In a previous study focused mainly on evaluating health impacts of sulfate particles, Burnett
22      and colleagues (1995) reported results from a time-series analysis of all-age respiratory hospital
23      admissions to 168 hospitals in Ontario, Canada over the 6-year period 1983 to 1988. The
24      outcome data were  prefiltered to remove seasonal variations using a weighted 19-day moving
25      average. The authors reported that O3 was associated with respiratory hospital  admissions;
26      however no quantitative results for O3 were presented.
27           Results from an analysis of five European cities indicated strong and consistent O3 effects
28      on unscheduled hospital admissions for COPD (Anderson et al., 1997). The five cities examined
29      - London, Paris, Amsterdam, Rotterdam, and Barcelona - were among those included in the
30      multicity APHEA (Air Pollution on Health:  European Approach) study. The number of years of
31      available data varied from 5 to 13 years among the cities.  In addition to O3, the study considered

        January  2005                              7-55         DRAFT-DO NOT QUOTE OR CITE

-------
 1      health impacts of BS, TSP, NO2, and SO2. City-specific effect estimates were pooled across
 2      cities using weighted means.  Significant effects were seen for O3, BS, TSP, and NO2.  Ozone
 3      effects were statistically significant in full year analyses and were larger in the warm season
 4      (April-September), 4.6% excess risk per 40 ppb increase in 1-h max O3, compared to the cool
 5      season (October-March), 1.5% excess risk.  The authors reported that among all pollutants
 6      examined, the most consistent and significant findings were for O3.  In addition, there was no
 7      significant heterogeneity in O3 effects among the cities.  No two-pollutant model results were
 8      reported.
 9           Several additional studies carried out in one or two cities over a span of five or more years
10      provided substantial additional evidence regarding O3 effects on  respiratory hospital admissions
11      (Anderson et al., 1998; Burnett et al., 1999, 2001; Moolgavkar et al.,1997; Petroeschevsky et al.,
12      2001;  Ponce de Leon et al., 1996; Sheppard et al., 1999  [reanalysis Sheppard, 2003]; Yang et al.,
13      2003). Two separate analyses of a large dataset from Toronto, Canada spanning the years 1980
14      to 1994 reported significant O3 effects on respiratory hospitalizations for all ages (Burnett et al.,
15      1999) and for persons under the age of 2 years (Burnett  et al., 2001). Analysis was performed
16      using Poisson GAM (default convergence criteria) with a nonparametric LOESS prefilter applied
17      to the  pollution and hospitalization data.  Both studies demonstrated that O3 effects were robust
18      when PM measures were added to the regression, whereas PM effects from univariate
19      regressions were markedly attenuated when O3 was added to the  regression.  These results imply
20      more robust associations with respiratory hospitalizations for O3  than PM.
21           Moolgavkar and colleagues (1997) reported significant and robust O3 effects on respiratory
22      hospital admissions in adults 65 years and older in Minneapolis and St. Paul, MN, but not in
23      Birmingham, AL.  The absence of effects in the southern city may reflect less penetration of O3
24      into the indoor environment due to greater use of air conditioning, and thus less correlation
25      between central  site O3 monitoring and actual exposures of the urban populace. In Brisbane,
26      Australia during 1987 to  1994, significant O3 effects on all-age and age-stratified asthma and
27      total respiratory hospital admissions were observed (Petroeschevsky et al., 2001).  The O3 effects
28      were robust to inclusion of PM (based on light scattering) and SO2 in copollutant regression
29      models.  Effect sizes appeared consistent in the warm and cool seasons, possibly reflecting the
30      relatively small  degree of seasonal variation in O3 levels observed in Brisbane.
        January 2005                              7-56        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Less consistent effects of O3 were seen in other respiratory hospitalization studies
 2      (Schouten et al., 1996; Lin et al., 2003; Lin et al., 2004; Morgan et al., 1998a; Oftedal et al.,
 3      2003).  In a study conducted in Amsterdam and Rotterdam, the Netherlands, significant
 4      associations were observed, however results were difficult to interpret due to the large number of
 5      statistical tests performed (Schouten et al., 1996).  Using a different analytical approach
 6      (case-crossover analysis), Lin and colleagues (2003) found no evidence for O3 effects on asthma
 7      admissions in 6- to 12-year-olds over the period 1981 to 1993 in Toronto, Canada.  In a
 8      California study by Niedell (2004),  a negative association was observed between hospitalizations
 9      for asthma and naturally occurring seasonal variations in O3 within zip codes in children aged
10      0 to 18 years. However, the O3 effect was found to be influenced by socioeconomic status.
11      Among children of low socioeconomic status, O3 generally was associated with increased
12      hospitalizations, with statistical significance reached in certain age groups.  Niedell further stated
13      that avoidance behavior on high O3  days may have attributed to the negative relationship
14      observed in children of higher socioeconomic status.
15           Another set of studies have examined associations between O3 and respiratory
16      hospitalizations in single cities over shorter (< 5 years) time spans. Positive and significant O3
17      effects were reported in Cleveland,  OH (Schwartz et al., 1996); Northern New Jersey (Weisel
18      et al., 2002); Toronto, Canada (Burnett et al., 1997b); Helsinki, Finland (Ponka and Virtanen,
19      1996);  Sao Paulo, Brazil (Gouveia and Fletcher, 2000a); and Hong Kong (Wong et al., 1999a).
20      The Helsinki study reported significant effects of O3 on both asthma and on digestive disorders
21      in a setting of very low O3 concentrations (Ponka and Virtanen, 1996), which raises questions
22      of plausibility.
23           No significant association with O3 was seen in studies from Los Angeles, CA (Linn et al.,
24      2000; Mann et al., 2002; Nauenberg and Basu, 1999); Vancouver, Canada (Lin et al., 2004);
25      London, England (Atkinson et al., 1999b); Edinburgh, Scotland (Prescott et al.,  1998); and
26      Drammen, Norway (Hagen et al., 2000).  Several of the studies reporting non-significant O3
27      effects were carried out in locations with low O3 levels, suggestive of a nonlinear exposure-
28      response relationship (Lin et al., 2004; Prescott et al., 1998). The non-significant findings in the
29      South Coast air basin, CA area are surprising given the elevated O3 concentrations observed
30      there (Mann et al., 2002). Inadequate control of seasonal confounding may underlie some of the
        January 2005                              7-57        DRAFT-DO NOT QUOTE OR CITE

-------
 1      non-significant and negative findings.  An additional factor likely contributing to the variability
 2      of results is the relatively small sample sizes included in some of these studies.
 3           In conclusion, while some inconsistencies are noted across studies, the evidence supports
 4      the findings of significant and robust effects of O3 on various respiratory disease hospitalization
 5      outcomes.  Large multicity studies, as well as many studies from individual cities have reported
 6      significant O3 associations with total respiratory hospitalizations, asthma, and COPD, especially
 7      in studies analyzing the O3 effect during the summer or warm season.
 8
 9      7.3.4   Association of O3 with Hospital Admissions for Cardiovascular Disease
10           Among the subset of hospital admissions studies that have examined associations of O3
11      with cardiovascular outcomes, most have found no consistent positive associations (Ballester
12      et al., 2001; Burnett et al., 1995, 1999; Linn et al., 2000; Mann et al., 2002; Petroeschevsky
13      et al., 2001; Prescott et al., 1998).  The exceptions are one study in Toronto, Canada, which
14      reported robust associations with both  total respiratory and cardiovascular hospital admissions
15      (Burnett et al., 1997b), and one in Hong Kong, in which circulatory, ischemic heart, and heart
16      failure were all significantly associated with O3 in the cool but not the warm season (Wong et al.,
17      1999b).  In the Hong Kong study,  O3 concentrations were similar in both seasons, with warm
18      season levels slightly lower,  mean 31.2 |ig/m3, compared to the cool season, mean 34.8 |ig/m3.
19      The authors speculated that differing activity patterns and home ventilation factors may have
20      contributed to the seasonal differences in O3 effects.  Weather in Hong Kong is mild throughout
21      the year, but less humid and  cloudy in  the cool season.  Thus, during the cool season people are
22      more likely to open windows or stay outdoors, resulting in higher personal exposures even with
23      similar ambient concentrations.  Based on this small set of studies, current evidence does not
24      support a conclusion that O3 has independent effects on cardiovascular hospitalizations.
25
26      7.3.5   Summary of Acute O3 Effects on Daily Emergency Department Visits
27              and Hospital  Admissions

28          • The vast majority  of hospitalization studies conducted over the past decade have
              looked at effects of O3 on either total respiratory  diseases and/or asthma. Significant
              associations with O3 were observed with both outcomes in many cases. Studies of
              emergency department visits for respiratory conditions also reported significant
              O3 effects, but the results tend to be less consistent across studies.

        January 2005                              7-58        DRAFT-DO NOT QUOTE OR CITE

-------
 1          • Many of the daily emergency department visits and hospitalization studies analyzed
              O3 risk estimates using year-round data.  Given the strong seasonal variations in
              O3 concentrations and the changing relationship between O3 and other copollutants
              by seasons, inadequate adjustment for seasonal effects might have masked or
              underestimated the association between O3 and the respiratory disease outcomes.
              Season stratified analyses typically yield more reliable O3 effect estimates.
 2          • Numerous studies have reported O3 effects while controlling for copollutants, including
              PM, in the analytical model. The evidence is supportive of independent O3 effects on
              respiratory admissions and emergency department visits. In most studies, O3 effects
              have been reported to be at least as robust as PM, and in some cases more so.
 3          • A subset of hospital admission studies examined the effect of O3 on cardiovascular
              outcomes. The limited evidence is inconclusive regarding the association between
              O3 exposure and cardiovascular hospitalizations.
 4
 5
 6      7.4   ACUTE EFFECTS OF OZONE ON MORTALITY
 7      7.4.1   Summary of Key Findings on Acute Effects of O3 on Mortality From
 8              the 1996 O3 AQCD
 9           A limited number of studies examined O3-mortality associations at the time of the previous
10      O3 AQCD, most of which were from the 1950s and 1960s. The 1996 O3 AQCD considered these
11      historical studies to be flawed because of either inadequate adjustment for seasonal trend or
12      temperature, or because of the use of questionable exposure indices. There were only a few
13      time-series studies that examined O3-mortality associations between the 1980s and mid-1990s.
14      These studies used more sophisticated approaches in addressing seasonal confounding and
15      weather models. One of these studies (Shumway et al., 1988) focused on the associations with
16      long-term fluctuations in Los Angeles, CA but did not examine short-term associations. A study
17      that reanalyzed the Los Angeles, CA data with a focus on the short-term associations (Kinney
18      and Ozkaynak, 1991) did find that, of the PM and gaseous criteria pollutants, O3 (reported as
19      total oxidants) was most strongly associated with total nonaccidental mortality. Then two
20      studies, one using Detroit, MI data (Schwartz, 1991) and the other using St. Louis, MO and
21      Kingston-Harriman, TN data (Dockery et al., 1992), reported that PM but not O3 was
22      significantly associated with mortality. However, the 1996 O3 AQCD discussed that, without
23      sufficient presentation of model diagnosis regarding the relationship between O3 and the weather
        January 2005                              7-59        DRAFT-DO NOT QUOTE OR CITE

-------
 1      models used, it was difficult to evaluate whether the lack of O3-mortality associations was
 2      possibly due to overspecification of the weather model. In summary, due to the insufficient
 3      number of studies that examined O3-mortality associations and the uncertainties regarding
 4      weather model specifications, the 1996 O3 AQCD was unable to quantitatively assess O3-
 5      mortality excess risk estimates, or even provide qualitative assessment of the likelihood of
 6      O3-mortality associations.
 7
 8      7.4.2   Introduction to Assessment of Current O3-Mortality Studies
 9           Introductory discussions of the PM mortality effects often cite historical air pollution
10      incidents such as the 1952 London, England smog episode in which thousands of deaths were
11      attributed to the air pollution from coal burning.  There is no counterpart "historical episode" for
12      O3-mortality effects. Instead, the early recognition of the adverse health effects of summer
13      oxidant air pollution, mainly from Los Angeles and other major cities with a high density of
14      automobiles, were based on symptoms such as eye and throat irritations. Thus, the focus of PM
15      epidemiology and that of O3 epidemiology have been historically different.
16           As shown in Table AX7-5 in the Chapter 7 Annex, the number of short-term mortality
17      studies that analyzed O3 has increased markedly since the last publication of the O3 AQCD in
18      1996. The increased attention to PM-mortality associations in the early  1990s lead to the
19      increase in studies that also examined O3, most often as a potential confounder for PM.
20      Although many of these PM studies also reported O3 estimates, they often lacked specific
21      hypotheses regarding mortality effects of O3 as the focus of these studies was to examine the
22      PM-mortality  effect. This is in contrast to the O3-morbidity studies, most of which were
23      specifically designed to  examine effects of "summer haze" and O3 (or oxidants) on respiratory
24      and other symptoms, lung  functions, and emergency department visits, etc.  However, new
25      studies with hypotheses  developed specifically for O3 effects on  mortality have become
26      available, such as the large U.S. 95 communities study by Bell et al. (2004), the U.S. 14 cities
27      study by Schwartz (2004), and the 23 European cities study by Gryparis et al.  (2004) discussed
28      in the next section.
29
        January 2005                              7-60        DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.4.3   Single-Pollutant Model O3-Mortality Risk Estimates
 2           To facilitate a quantitative overview of the O3-mortality effect estimates and their
 3      corresponding uncertainties, the percent excess risks of total nonaccidental mortality calculated
 4      using all year data are plotted in Figure 7-10.  Studies that only conducted seasonal analyses will
 5      be presented in the next section. This figure does not include studies that only examined cause-
 6      specific mortality. In studies where multiple lags were presented, the multiday lag was selected
 7      to represent the cumulative effect from all days examined.  If cumulative lags were not analyzed,
 8      the effect estimate from the 0- or  1-day lag was selected for presentation. All effect estimates
 9      are from single-pollutant models and include all age groups unless noted otherwise.  The
10      majority of the estimates are positive with a few exceptions.  Four multicity studies showed
11      positive and significant O3 effect estimates for all cause (nonaccidental) mortality. An excess
12      mortality risk of 4.5% per 40 ppb increase in 1-h max O3 was estimated from the four European
13      cities of the APHEA project (Touloumi et al., 1997).  The European effect estimate was larger
14      than those from the large U.S. National Morbidity, Mortality and Air Pollution Study
15      (NMMAPS). An excess risk of 1.0% and 0.8% per 20 ppb increase in 24-h avg O3 was observed
16      from the U.S. 95  communities study (Bell et al., 2004) and U.S. 90 cities study (Samet et al.,
17      2000; reanalysis Dominici et al., 2003), respectively.  Similarly, the U.S. 14 cities study by
18      Schwartz (2004)  observed an 0.8% excess risk per 40 ppb increase in 1-h max O3.
19           Only one multicity study did not observe a statistically  significant O3 effect on mortality.
20      As an extension of the four European cities study, researchers of the APHEA project
21      investigated the effect of O3 on total, cardiovascular, and respiratory mortality in 23 cities
22      throughout Europe (Gryparis et al., 2004). A cumulative lag of 0 to 1 days was  hypothesized
23      a priori.  A two-stage hierarchical model, which accounted for statistical variance and
24      heterogeneity among cities, was used to estimate the pooled regression coefficients.  Due to
25      substantial heterogeneity among cities, random effects regression models were applied. The
26      pooled  effect estimate for the 23 European cities was a positive, but nonsignificant value of
27      0.23% (95% CI: -0.85, 1.95) per 40 ppb increase in 1-h max O3 for all  seasons.  The researchers
28      noted that there was a considerable seasonal difference in the O3 effect on mortality, thus the
29      nonsignificant effect for the all year data might be attributable to inadequate adjustment for
30      confounding by season. This seasonal effect will be discussed further in the next section.
31

        January 2005                              7-61         DRAFT-DO NOT QUOTE OR CITE

-------
                        Bell et al. (2004): U.S.95 communities -
     Sametetal.(2000;reanalysis Dominic! etal.,2003): U.S.90cities -
                               Schwartz(2004): U.S.14dties-
            Kinney and Ozkaynak (1991): Los Angeles County, CA —
                   Kinneyetal.(1995): Los Angeles County, CA -
                        Fairley (2003): Santa Clara County,CA -
                                  Gamble (1998): Dallas,TX-
                           Dockeryetal.(1992): St. Louis, MO -
                      ItoandThurston (1996): Cook County, IL -
           Lippmannetal. (2000; reanalysis Ito, 2004): Detroit, Ml -
                        Lipfertetal.(2000a): Philadelphia,PA -
                        Lipfert et al.(2000a): 4 Counties in PA -
                   Lipfert et al. (2000a): 7 Counties in PA and NJ -
          Lipfert et al. (2000a): 7 Counties in PA and NJ (age 0-65) -
           Lipfert et al. (2000a): 7 Counties in PA and NJ (age 65+) -
                   Chock etal. (2000):  Pittsburgh, PA (age 0-74) -
                   Chocketal.(2000): Pittsburgh,PA (age 75+) -
                      Dockery et al. (1992): Eastern Tennessee —
                        Klemm and Mason (2000): Atlanta, GA -
                     Klemm etal. (2004): Atlanta, G A (age 65+) -
                      Gryparis etal.(2004): 23 European cities —
     Touloumi etal. (1997):  4 European cities (Athens,Barcelona,London, Paris) —
                      Anderson etal  (1996):  London,England —
                       Bremneretal.(1999):  London,England —
               Bremner etal. (1999): London, England (age 0-64) —
               Bremner etal.(1999):  London, England (age 65+) —
              Bremner etal. (1999): London, England (age 65-74) -
               Bremner etal.(1999):  London, England (age 75+) —
                     Prescottetal.(1998): Edinburgh,Scotland -
                            Zmirouetal.(1996): Lyon,France —
        Hoek etal. (2000; reanalysis Hoek,2003): The Netherlands -
      RoemerandvanWijnen (2001): Amsterdam, the Netherlands —
              Verhoeffetal.(1995): Amsterdam,the Netherlands -
               Peters et al. (2000b): Coal basin in Czech Republic -
                     Peters et al.(2000b):  NE Bavaria,Germany -
                 Garcia-Aymerichetal.(2000): Barcelona,Spain —
                         Sunyeretal.(1996): Barcelona,Spain —
                 Sunyeretal.(1996): Barcelona,Spain (age70+) —
                        Borja-Aburto et al. (1997): Mexico City —
                 Borja-Aburto et al (1997): Mexico City (age <5) -
                Borja-Aburto et al. (1997): Mexico City (age 65+) -
                     Borja-Aburto et al. (1998): SW Mexico City -
             Borja-Aburto et al. (1998): SW Mexico City (age 65+) -
                             O'Neill etal. (2004): Mexico City-
                     O'Neill et al. (2004): Mexico City (age 65+) -
                 Gouveia and Fletcher (2000b): Sao Paulo, Brazil —
         Gouveia and Fletcher (2000b): Sao Paulo, Brazil (age 65+) -
                 Saldiva etal.(1995): Sao Paulo,Brazil (age65+) -
                           Ostroetal.(1996): Santiago,Chile -
                       Morgan etal. (1998b): Sydney, Australia —
                     Simpson etal.(1997): Brisbane,Australia —
             Simpson etal.(1997): Brisbane,Australia (age 0-64) -
              Simpson etal. (1997): Brisbane, Australia (age 65+) —
                               Kim etal. (2004): Seoul, Korea -
                               Lee eta I. (1999): Seoul, Korea -
                        Lee and Schwartz (1999): Seoul, Korea —
                               Lee etal. (1999): Ulsan, Korea -
                                                                           % Change in  Mortality
                                                                   -10
                                                  10
                                                   I
                                            20
                             lag 0-1  •
                             lag 0-1  •
                             lag 0-1  •
                             lag 0-1  •
                             lag 0-1  •
Latin America
                                  > lag 0-6
                                   lagO
                                   lagO
                                   Iag1
                                   Iag1
                      © Multicity combined
                      •  Single city
                                            lag 0-1
I                 Confidence bands not provided but
                 noted as significant at p=0.055
                                    •  lagO
                                           - Iag1
                                                       lag 0-1
                                    lag 0-1
                                                            lagO
                                         lag 0-6
                                    - lagO
                                    -•	
                                            - lags
                                            - lagO
                                             Iag1
                                lagO
                                         lag 0-1
                                          • lagO
                                          	•-
                                                         • lagO
                                   • lagO
                                                                                                                     lagO
Figure 7-10.   All  cause (nonaccidental) O3 excess  mortality  risk estimates (95% CI) for  all
                     year analyses per 40 ppb increase in 1-h max  O3 or equivalent.  Analyses
                     include all  ages unless otherwise noted.
January 2005
       7-62
DRAFT-DO NOT QUOTE OR CITE

-------
 1           The U.S. 95 communities study by Bell et al. (2004) extended an earlier analysis (Samet
 2      et al., 2000; reanalysis Dominici et al., 2003) on short-term effects of O3 on total and
 3      cardiopulmonary mortality using the NMMAPS data base from 1987 to 2002. The results of this
 4      study are discussed in detail here because of the study's emphasis on U.S. data and the inclusion
 5      of 95 large communities across the country, making this mortality study most representative of
 6      the U.S. population. In addition, this study is one of few that have focused specifically on O3
 7      hypotheses testing. Within-community results first were calculated using single-day lags of 0, 1,
 8      and 2 days, and a 7-day distributed lag in O3 exposure.  A two-stage hierarchical model was used
 9      to determine a national average effect estimate. Figure 7-11 presents community-specific and
10      national average O3 risk estimates  for total mortality per 10 ppb increase in 24-h avg O3 from a
11      constrained 7-day distributed lag model. For total mortality in the single-day lag models, the
12      national estimates of the O3 effect  on mortality were statistically significant for 0-, 1-, and 2-day
13      lags, with the largest effect observed from O3 concentrations at a 0-day lag.  The 7-day
14      distributed lag model estimated the cumulative risk of mortality associated with O3
15      concentrations on the same day and six previous days.  Results from the constrained distributed
16      lag model  indicated that a 10 ppb increase in 24-h avg O3 in the previous week was associated
17      with a statistically significant increase  of 0.52% (95% CI: 0.27, 0.77%) in daily mortality. The
18      mean 24-h avg O3 concentration was approximately 26 ppb for the 95 communities. Figure 7-11
19      illustrates the preponderance of cities with a positive yet nonsignificant relationship between O3
20      and mortality.  However, the national average O3 mortality estimate, which includes results from
21      all 95 U.S. communities, is positive and statistically significant.  These results further support
22      the data presented in Figure 7-10 from nearly 40 studies in locations both in the U.S. and in other
23      countries.
24           Several studies conducted meta-analyses of  O3-mortality associations (Stieb et al., 2002,
25      2003; Thurston and Ito, 2001; World Health Organization, 2004). Most of these studies included
26      GAM studies using default convergence criteria except Stieb et al. (2003), which compared
27      effect estimates from GAM-affected studies to non-GAM studies. All of these meta-analyses
28      reported fairly consistent and positive combined estimates, approximately 2% excess total
29      non-accidental mortality per 40 ppb increase in 1-h max O3. However, most of these studies
30      were not analytical in design in that they did not attempt to examine the source of heterogeneity,
        January 2005                              7-63        DRAFT-DO NOT QUOTE OR CITE

-------
                                      % Change in Mortality

Orlando
_
_ , . .
Las Veaas


Birmingham
— npino
«Lt P^tpr«hnrn








„ P .


^anta Ana 1 Anahipm


1 nkr Phnrlr-
Tucson






P Phri"Ti





^T 1 nni"


Mn^lfrnnn





!__.._,, pjttf


— j'l .
...
Baltimore
-
.. _ i
Jackson
Portland OR
_ '.


1 h "t
Austin
San Rprnarlino
P 'rl
Cleveland
flrlinntnn
' M ri t




1 inrnln
Washington





° UmMad!son
,, .
Phrirniv
Knoxville
1 nni«:villp
Hnnnlnln
Rochester
....
.,, P
Colorado Sennas
„ P ~
Phiririn
Shreveoort





NATIONAL —
-1 C
I


*





























































































1123
I I I


—


5


5

™



^


^


^


5









*
5
*
*
~
^
2

5
*
5
^

























~


*


~


~


^


5

"



' _


' ^
— . —
Figure 7-11.  Bayesian city-specific and national average estimates for the % change (95%
             CI) in daily mortality per 10 ppb increase in 24-h avg O3 in the previous
             week using a constrained distributed lag model for 95 U.S. communities
             (NMMAPS), arranged by size of the effect estimate.

Source: Derived from Bell et al. (2004).
January 2005
7-64
DRAFT-DO NOT QUOTE OR CITE

-------
 1      although some suggested an influence of weather model specification (Thurston and Ito, 2001)
 2      and another reported evidence of publication bias (World Health Organization, 2004) in the past
 3      literature. None of these studies addressed the issue of season-specific estimates, and therefore,
 4      interpreting these combined estimates requires caution. The estimates from these meta-analyses
 5      appear to be larger than the national average estimate of 1.0% excess risk per 20 ppb increase in
 6      24-h avg O3 from the largest U.S. 95 communities study (Bell et al., 2004).  There are a few new
 7      meta-analyses and multicity studies currently being conducted specifically to address such issues
 8      as season-specific analyses,  publication bias, weather model specification, potential confounding
 9      by fine particles, distributed lag effects, and the potential influence of air conditioning.  These
10      studies are expected to provide new information that will shed light on the outstanding questions.
11           Collectively, the above studies suggest an excess risk of total nonaccidental mortality
12      associated with acute O3 exposure. Despite the different analytical approaches and alternative
13      model specifications used in the various studies, overall, the range of estimates were relatively
14      narrow, with the positive estimates from 0 to 7% per 40 ppb increase in 1-h max O3 or
15      equivalent.
16
17      7.4.4   Seasonal Variation in O3-Mortality Risk Estimates
18           Since the seasonal cycle  of O3 follows the seasonal cycle of temperature (which is
19      inversely related to the mortality seasonal cycle), inadequate adjustment of temporal trends in
20      the regression model may lead to negative O3-mortality risk estimates. In addition, as discussed
21      in Section 7.1.3.5, in some cities low-level O3 during winter may be negatively correlated with
22      PM and  other primary pollutants, resulting in negative correlations between O3 and mortality
23      even in short-term relationships.  The confounding effect by season could be substantially
24      reduced  by conducting season-stratified analyses.
25           A fewer number of O3 mortality studies performed seasonal analyses.  Figure 7-12 presents
26      the studies that reported O3 risk estimates for all cause mortality by season.  For those studies
27      that obtained O3 risk estimates for each of the four seasons, only summer and winter results are
28      shown.  The estimates for year-round data analyses, when available, also are shown for
29      comparisons.  In all the studies, the O3 risk estimates are larger during the warm season than the
30      cool season, with the all year estimates generally in between the two seasonal estimates.
31

        January  2005                               7-65        DRAFT-DO NOT QUOTE OR CITE

-------
                                                         % Change in Mortality
          Bell etal.(2004): U.S.95 communities -

                 Samet et al.(2000; reanalysis
              Dominici et al.2003):U.S.90 cities"

                 Schwartz (2004): U.S. 14 cities -

        Ostro et al. (1995): San Bernadino County
                     and Riverside County, CA

                     Gamble (1998):Dallas,TX-


        Moolgavkaretal.(1995):Philadelphia,PA -


      Chock etal.(2000): Pittsburgh, PA (age 0-74) -


      Chock etal. (2000): Pittsburgh, PA (age 75+) -


           Vedaletal.(2003):Vancouver,Canada -
         Gryparis et al. (2004): 21 European cities -


         Anderson etal.(1996): London,England -

                  Hoeketal.(2000;reanalysis
                 Hoek,2003):The Netherlands"


            Sunyer et al. (1995): Barcelona, Spain -


    Sunyeretal.(1996):Barcelona,Spain (age 70+) -



           Borja-Aburto et al. (1997): Mexico City -


           Cifuentes et al.(2000):Santiago,Chile -


              Ostro etal. (1996): Santiago,Chile -


         Simpson etal.(1997):Brisbane,Australia -



                 Kim etal.(2004):Seoul,Korea -
                                                 -10
                                                   I
                                   10
                                    I
                                    20
30
 I
                                         |u.S. and Canada]
                   -e-
          *
          • lag 0-6
          <
          •  Iag1

          <*r
                     -0-
lag 1 -2
                                                 lagO
Latin America
Australia
                  -e-
               -e-
                -e-
              lagO


              lagO
                            •  (SE not given; significant at 0.05 level)
                          O   (SE not given; not significant at 0.05 level)
               ~ Iag1

                lagO
                              lagO
                                                                 -G-
                     -B-
                             lag 0-1
                                lagO
                              ,ag1
                     -e-
                        -9-
                      lagO


                      Iag1
                          lagO


                         lag 1-2

                             Iag1
                          -e-
                                 Iag1
                                              lagO
Figure 7-12.   All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
                  by season per 40 ppb increase in 1-h max O3 or equivalent.  Analyses include
                  all ages unless otherwise noted.
January 2005
            7-66
               DRAFT-DO NOT QUOTE OR CITE

-------
 1           In three U.S. and European multicity studies (Gryparis et al., 2004; Samet et al., 2000
 2      [reanalysis Dominici et al., 2003]; Schwartz, 2004), season-stratified analyses indicated that the
 3      O3-mortality effect estimates were significant and positive in the warm season, with larger
 4      effects observed compared to the year-round analyses.  The effect estimates from the cool season
 5      were notably smaller and less or nonsignificant. In the case of the U.S. 90 cities study, the cool
 6      season mortality estimate was negative, which was most likely attributable to the inverse
 7      relationship between O3 and PM in the winter.
 8           In the U.S. 95 communities study by Bell et al. (2004), the warm season (April-October)
 9      effect estimate was 0.78% (95% CI: 0.26, 1.30) excess risk per 20 ppb increase in 24-h avg O3,
10      compared to 1.04% (95% CI: 0.54, 1.55) calculated using all available data. The small
11      difference in the size of the effect estimates between the analysis using all available data and that
12      using warm season only data might be attributable, at least partially, to the varying peak O3
13      seasons and the difference in O3 concentrations by community or region. In addition, the
14      varying seasonal relationship between O3 and PM by community also might have contributed to
15      the difference, as these mortality effect estimates were not adjusted for confounding by PM.
16           Studies that conducted analysis by season indicate that O3 mortality risk estimates are often
17      larger in the warm season compared to the colder season. The larger effects observed in the
18      warm season when O3 levels are higher are consistent with causal association. The seasonal
19      dependence of O3-mortality effects complicates interpretation of O3 risk estimates calculated
20      from year-round data without adequate adjustment of temporal trends.
21
22      7.4.5   O3-Mortality Risk Estimates Adjusting for PM Exposure
23           As previously mentioned, the confounding between "winter type" pollution (e.g., CO, SO2,
24      and NO2) and O3 is not of great concern because the peaks of these pollutants do not strongly
25      coincide. The main confounders of interest for O3, especially for the northeast U.S., are
26      "summer haze" type pollutants such as acid aerosols and sulfates.  Since  very few studies had
27      these chemical measurements, PM (especially PM2 5), may serve as surrogates. However, due to
28      the expected high correlation among the constituents of the "summer haze mix," multipollutant
29      models including these pollutants may result in unstable coefficients, and therefore, an
30      interpretation of such results requires some caution.
31

        January 2005                             7-67        DRAFT-DO  NOT QUOTE OR CITE

-------
 1           Figure 7-13 shows the O3 risk estimates with and without adjustment for PM indices using
 2      all year data in studies that conducted two-pollutant analyses.  Approximately half of the O3 risk
 3      estimates slightly increased while the other half slightly decreased in value with the inclusion of
 4      PM indices in the model.  In general, the O3-mortality risk estimates were robust to adjustment
 5      for PM in the models, with the exception of Los Angeles, CA data with PM10 (Kinney et al.,
 6      1995) and Mexico City  data with TSP (Borja-Aburto et al., 1997).
 7           The U.S. 95 communities study by Bell et al. (2004) examined the sensitivity of acute
 8      O3-mortality effects to potential confounding by PM10. Restricting analysis to days when both
 9      O3 and PM10 data were available, the community-specific O3-mortality effect estimates as well as
10      the national average results indicated that O3 was robust to adjustment for PM10 (Bell et al.,
11      2004). One study (Lipfert et al., 2000a) reported O3 risk estimates with and without  sulfate
12      adjustment. Lipfert et al. (2000a) calculated O3 risk estimates based on mean (45 ppb) less
13      background (not stated) levels of 1-h max O3 in seven counties in Pennsylvania and New Jersey.
14      The O3 risk estimate was not substantially affected by the addition of sulfate in the model (3.2%
15      versus 3.0% with sulfate) and remained statistically significant.
16           Several O3-mortality studies examined the effect of confounding by PM indices in different
17      seasons (Figure 7-14). In analyses using all year data and warm season only data, O3 risk
18      estimates were once again fairly  robust to adjustment for PM indices, with values showing both
19      slight increases and decreases with the inclusion of PM in the model. In contrast, in  the analyses
20      using cool season data only, the O3 risk estimates all increased with the adjustment of PM
21      indices, although none reached statistical significance. For example, in the European study of
22      21 cities (two cities that did not have 8-h max O3 data were excluded from the analysis), the
23      summer O3-mortality  estimate was relatively robust to adjustment for PM10, slightly decreasing
24      from 1.82% (95% CI: 0.99, 3.06) to 1.58% (95% CI: 0.47, 2.88) excess risk per 30 ppb increase
25      in 8-h max O3 (Gryparis et al., 2004). In contrast, the winter effect estimate increased from
26      0.70% (95% CI: -0.70,  2.17) to  1.29% (95% CI: -0.46, 3.00) per 30 ppb increase in 8-h max O3
27      after adjusting for PM10. These results indicate that the confounding effect by PM may vary by
28      season. Although PM does not appear to significantly confound the association between O3 and
29      mortality in the analysis of warm season data, during the cool season, the inverse relationship
30      between O3 and PM10 may influence the effect estimate for O3-related mortality.
31

        January 2005                              7-68         DRAFT-DO NOT QUOTE OR CITE

-------
                                           % Change in Mortality
Samet et al, (2000, reanalysis
Dominid et al., 2003): U.S. 90 cities -
with PM10 -
Schwartz (2004): U.S. 14 cities -
with PM10 -
Kinney and Ozkaynak (1991): Los Angeles County, CA -
with KM -
Kinney etal. (1995): Los Angeles County CA -
with PM10 -



with PM10

Lipfertetal. (2000a): Philadelphia, PA -
withPM25 -
with PM10 -
Lipfert et al. (2000a): 4 Counties in PA -
withPM25 -
with PM10 -
Lipfert et al. (2000a): 7 Counties in PA and NJ -
withPM26 -
with PM10 -
Lipfert et al. (2000a): 7 Counties in PA and NJ (age 0-65) -
withPM25 -
with PM10 ~
Lipfertetal. (2000a): 7 Counties in PA and NJ (age 65+) -
withPM25 -
with PM10 -
with PM10 -

Touloumietal. (1997): 4 European cities

with BS
Anderson et al.(1996): London, England -
with BS -
Hoek(2000, reanalysis Hoek, 2003): The Netherlands -
with PM10 -

Borja-Aburto etal. (1997): Mexico City -
with TSP -
Borja-Aburto et al. (1997): Mexico City (age <5) -
Borja-Aburto et al. (1997): Mexico City (age 65+) -
with TSP



Ostroetal. (1996): Santiago, Chile -


are 7-13. All cause (nonaccid
with adjustment foi
10 -5 0 5 10 15
I ill
I 	 1
\¥±] ^_ lag 0 o O3 only
I§I lag 0 • Q3 with PM
"^ lagi
	 ^~ Iag1
f~\
lag 0 ° -

9 laa 1-2

9 laa 1-2
lag 0-1 0° A
Q / Confidence bands not
lag °-1 • /*• provided but noted as
* \ significant at p=0.055
O )
lag 0-1 • _J Noted as nonsignificant
^ at p=0.055
0 ~N
lag 0-1 • 1
9 / Confidence bands not
>• provided but noted as
laq 0 1 2 \ significant at p=0.055
0° lag 0

•° laaO

| Europe |
ff laa varied by city
^ lagO
~^" Iag1
^ laa 2

Latin America ^ ^ lag 0
§ laqO
- ~e~ laq 0

^ lag 1-2

t lag 1-7
2 laq 1

1 	 1 c~\
|Asia| ~ laa 1
	 W
ental) O3 excess mortality risk estimates (95% CI)
r PM indices for all year analyses per 40 ppb increas
             1-h max O3 or equivalent. Analyses include all ages unless otherwise noted.
January 2005
7-69
DRAFT-DO NOT QUOTE OR CITE

-------
                                                     % Change in Mortality
                                  -15
Ostro et al. (1995): San Bernadino County
           and Riverside County, CA —
                      with PM2.5 -


Moolgavkar et al. (1995): Philadelphia, PA -
                       with TSP -

                       with TSP -
  Chock et al. (2000): Pittsburgh, PA (age 0-74) -
                          with PM10 —

                          with PM10 -
  Chock et al. (2000): Pittsburgh, PA (age 75+) -
                          with PM10 •

                          with PM10 -

                          with PM,0 -
     Gryparisetal.(2004):21 European cities —
                          with PM,n -
                                10 "
                          with PM1(
     Anderson etal.( 1996):London, England -
                            with BS -

                            with BS -

                            with BS •
         Ostro et al. (1996): Santiago, Chile -
                          with PM10 •

                          with PM,n •
            Kim etal.(2004): Seoul, Korea -
                          with PM10 -

                          with PM10 -
                                    U.S.
                                    Europe
                                          -10
                                          J	
                                    Latin America
-5
 I
5
I
                                                                -e-
                                                                   -e-
                                                             -x-
                                                            -e-
                                                            -e-
                                                          -e-
                                                          -e-
                                                            -x-
                                     X All year  • Warm  O Cool
                                                                              lagO
                                                                                    lag 1
                                                                              lagO
                                                                                 lag 0
                                                                                 lag 0-1
                                                                     -e-
                                                                                   lag 0
                                                                     -e-
                                                                       -e-
                                                                                   Iag1
                                                                              -X-
10
                                                                                      Iag1
Figure 7-14.  All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
               with adjustment for PM indices by season per 40 ppb increase in 1-h max
               O3 or equivalent.  Analyses include all ages unless otherwise noted.
January 2005
                                            7-70
  DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.4.6   O3 Risk Estimates for Specific Causes of Mortality
 2           Many of the time-series mortality studies examined broad underlying causes of mortality,
 3      such as cardiovascular and respiratory causes.  The U.S. 95 communities study (Bell et al., 2004)
 4      analyzed O3 effect estimates from cardiovascular and respiratory mortality.  Significant effects
 5      were seen at 0- and 2-day lags with results similar to total mortality.  The national average
 6      estimate from the constrained distributed lag model was slightly greater for cardiopulmonary
 7      deaths, with an excess risk of 1.28% (95% CI: 0.62, 1.97) per 20 ppb increase in 24-h avg O3 in
 8      the preceding week. In a related study, Huang et al. (2004) examined O3 effects on
 9      cardiopulmonary mortality during the summers of 1987 to 1994 in 19 large U.S. cities from the
10      NMMAPS database.  In the 7-day distributed lag model, the O3 effect estimate was 2.52% (95%
11      CI: 0.94, 4.10) excess risk in cardiopulmonary  mortality per 20 ppb increase in 24-h avg O3
12      (Huang et al., 2004).  Several studies observed that the risk estimates for the respiratory category
13      were larger than the cardiovascular and/or total nonaccidental categories (e.g., Anderson et al.,
14      1996; Gouveia and Fletcher, 2000b; Gryparis et al., 2004; Zmirou et al., 1998). In the European
15      21 multicities study (Gryparis  et al., 2004), the warm season effect estimate for respiratory
16      mortality was 6.75% excess risk per 30 ppb increase in 8-h max O3, compared to 2.70% for
17      cardiovascular mortality  and 1.82% for total mortality. In contrast, other studies have found that
18      the risk estimates for the respiratory category were essentially zero or even negative while the
19      risk estimates for total or cardiovascular categories were positive (e.g., Borja-Aburto et al., 1998;
20      Bremner et al., 1999;  Lipfert et al., 2000a; Morgan et al., 1998b).  These apparent
21      inconsistencies across studies may be due in part to the difference in model specifications, but
22      they may also reflect the lower statistical power associated with the smaller daily counts of the
23      respiratory category (usually accounting for less than 10% of total deaths) compared to the larger
24      daily counts for the cardiovascular category (approximately 40 to 50% of total  deaths). Thus, an
25      examination of the differences in risk estimates across specific causes requires a large population
26      and/or a long period of data collection.
27           The analyses of a 9-year data set for the whole population of the Netherlands
28      (population = 14.8 million) provided O3 (and other pollutants) risk estimates for more specific
29      causes of mortality, including COPD, pneumonia, and subcategories of cardiovascular causes
30      (Hoek et al., 2000, 2001; reanalysis Hoek, 2003).  The excess risk estimate for COPD was small
        January 2005                              7-71        DRAFT-DO NOT QUOTE OR CITE

-------
 1      and not significant (0.8% [95% CI: -2.4, 4.2] per 30 ppb 8-h avg O3), while the excess risk
 2      estimate for pneumonia (5.6% [95% CI: 1.8, 9.5]) was much larger than that for total
 3      nonaccidental mortality (1.6% [95% CI: 0.9, 2.4]). The excess risk estimates for some of the
 4      cardiovascular subcategories, including heart failure (3.8% [95% CI: 0.5, 7.3]) and thrombosis-
 5      related disease (6.0% [95% CI: 1.1, 10.8]), showed greater risk estimates than that for total
 6      mortality. However, these elevated relative risks were not specific to O3. For example, most of
 7      the pollutants examined, including PM10, BS, SO2, NO2, CO and NO3 , were significantly
 8      associated with pneumonia.  Therefore, it is difficult to make a causal inference specific to O3
 9      based on these results.
10           De Leon et al. (2003) examined the role of contributing respiratory causes in the
11      associations between air pollution and nonrespiratory mortality (circulatory and cancer) in
12      New York City during the period of 1985 to 1994.  The main finding of this study was that, for
13      the older age group (75+ years), the estimated excess mortality risks  for PM10 were higher for the
14      nonrespiratory deaths that had contributing respiratory causes, compared to the nonrespiratory
15      deaths without contributing respiratory causes.  This pattern was also seen for CO and SO2, but
16      not for O3. Therefore, this study did not suggest a role of contributing respiratory causes in the
17      association between O3 and nonrespiratory causes of deaths.
18           In summary, these studies examining specific respiratory or cardiovascular causes of death
19      often found risk estimates that were higher than those for the total or broader death cause
20      categories, but their lower statistical power in the smaller subcategories often made it difficult to
21      distinguish the contrasts in estimates.
22
23      7.4.7   O3-Mortality Risk Estimates for Specific Subpopulations
24           Some studies examined O3-mortality risk estimates in potentially susceptible
25      subpopulations,  such as those with underlying cardiopulmonary disease.  Sunyer et al. (2002)
26      examined the associations between air pollution and deaths in a cohort of patients (467 men and
27      611 women) with severe asthma in Barcelona, Spain during the period of 1986 to 1995. A case-
28      crossover study  design was used to estimate excess odds of mortality adjusting for weather and
29      epidemics in three groups:  (1) those who had only one asthma emergency department visit;
30      (2) those who had more than one asthma emergency department visit; and (3) those who had
31      more than one asthma and COPD emergency department visit.  Those with more than one

        January 2005                             7-72        DRAFT-DO NOT QUOTE OR CITE

-------
 1      asthma emergency department visit showed the strongest associations with the examined air
 2      pollutants, with NO2 being the most significant predictor, followed by O3. Sunyer et al. (2002)
 3      reported a significant association between O3 and all cause deaths for this group during the warm
 4      season, with an odds ratio of 1.90 (95% CI: 1.09, 3.30) per 48 |ig/m3 increase in 1-h max O3,
 5      compared to an odds ratio of 1.02 (95% CI: 0.73, 1.43) for those with only one asthma
 6      emergency department visit and 1.05 (95% CI: 0.73, 1.50) for the group with a concomitant
 7      diagnosis of COPD.  The magnitude of the effect size estimates reported for patients with more
 8      than one asthma emergency department visit was large compared to the total mortality risk
 9      estimate (relative risk of 1.03  per 48 |ig/m3 increase in  1-h max O3) observed in the related study
10      by Sunyer et al. (1996). In another Barcelona study, Saez et al. (1999) examined asthma
11      mortality death among persons aged 2 to 45 years. Once again, O3 and NO2 were the only air
12      pollutants that were significantly associated with asthma mortality death. While the similarity of
13      the patterns of associations between O3 and NO2 makes it difficult to speculate on the specific
14      causal role of O3, the results of these studies suggest that individuals with severe asthma may
15      make up a subpopulation that is sensitive to these pollutants.
16           Sunyer and Basagna (2001) also performed an analysis of emergency department visits by
17      a cohort with COPD.  The results from this study suggested that PM10, but not gases were
18      associated with mortality risks for the COPD cohort. However, a Mexico City study by Tellez-
19      Rojo et al. (2000) observed a significant association between COPD  mortality and O3, along with
20      PM10, among patients living outside a medical unit. For a cumulative 5-day lag, an excess risk of
21      15.6% (95% CI: 4.0, 28.4) per 1-h max O3 was observed for COPD mortality.
22           Goldberg et al. (2003) investigated the association between air pollution and daily
23      mortality with congestive heart failure as the underlying cause of death in patients aged 65 years
24      or more in Montreal, Quebec, Canada during the period of 1984 to 1993. Analysis was stratified
25      into two groups, those whose underlying cause of death was congestive heart failure and those
26      with a diagnosis of congestive heart failure one year before their death.  They found no
27      association between daily mortality for congestive heart failure and any pollutants.  However,
28      they did find significant associations between daily mortality among those who were classified
29      as having congestive heart failure before death and coefficient of haze, SO2, and NO2.  Ozone
30      was not significantly associated but showed positive risk estimates for year-round and warm
        January 2005                              7-73        DRAFT-DO NOT QUOTE OR CITE

-------
 1      season data and a negative risk estimate for cool season data. While the 10-year study period for

 2      this data was long, the daily mean death counts for the specific subcategory chosen was

 3      relatively small (0.7/day for mortality with congestive heart failure as underlying cause of death

 4      and 4.0/day for total mortality in patients previously diagnosed with congestive heart failure),

 5      limiting the power of the study.

 6           Few studies have examined O3-mortality effects for specific subpopulations. Among those
 7      that investigated the effect of air pollution in populations with underlying cardiopulmonary

 8      diseases, associations were not unique to O3 but were shared with other pollutants.  The results

 9      from Spain (Saez et al.,  1999;  Sunyer et al., 2002) suggest that severe asthmatics may be

10      susceptible to the mortality effects associated with NO2 and O3.
11

12      7.4.8   Summary of Acute O3 Effects on Mortality

13           •  A substantial body of new data on acute mortality effects of O3 has emerged since the
                previous O3 AQCD.  While uncertainties remain in some areas, it can be concluded
                that robust associations have been identified between various measures of daily O3
                concentrations and increased risk of mortality.  The fairly small but consistent
                associations cannot be readily explained by confounding due to time, weather,
                nor copollutants.

14           •  The majority of the available O3-mortality risk estimates were  computed using all
                year data.  The results from the studies that conducted analysis by season suggest that
                the O3 risk estimates were larger in the warm season.  Some of the risk estimates  in
                the cool season were negative, possibly reflecting the negative correlation between
                low-level O3 and PM (and other primary pollutants) during that season. Thus,
                without adequate adjustment for temporal trends, the O3 risk estimates obtained for
                year-round data may be misleading and likely underestimate the effects during the
                warm season.
15           •  Some studies examined specific subcategories of mortality, but most of these studies
                had limited statistical power to detect associations due to the small daily mortality
                counts.  The large U.S. 95 communities study indicated that there was a slightly
                greater risk of cardiopulmonary mortality compared to total mortality.

16           •  Few studies examined the effect of O3 on mortality in subpopulations with underlying
                cardiopulmonary diseases.  Similar to cause-specific mortality, these population-
                specific studies had limited statistical power to detect associations. The evidence
                suggests that individuals with severe asthma may be at increased risk of O3-related
                mortality, however, similar results were seen with other pollutants.

17
       January 2005                               7-74        DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.5   CHRONIC EFFECTS OF OZONE
 2      7.5.1    Summary of Key Findings on Studies of Health Effects and Chronic
 3              O3 Exposure from the 1996 O3 AQCD
 4          The 1996 O3 AQCD concluded that there was insufficient evidence from the limited
 5      number of studies to determine whether long-term ambient O3 exposures resulted in chronic
 6      health effects. However, the aggregate evidence suggested that chronic O3 exposure, along with
 7      other environmental factors, could be responsible for health effects in exposed populations.
 8
 9      7.5.2    Introduction to Morbidity Effects of Chronic O3 Exposure
10          Several new longitudinal epidemiologic investigations have yielded information on health
11      effects of long-term O3 exposures. Epidemiologic interest in investigating long-term effects has
12      been motivated by several considerations.  Animal toxicology studies carried out from the late
13      1980s  onward demonstrated that long-term exposures can result in permanent changes in the
14      small airways of the lung, including remodeling of the airway architecture (specifically the distal
15      airways and centriacinar region) and deposition  of collagen, as discussed earlier in Chapter 5.
16      These  changes result from the damage and repair processes that occur with repeated exposure.
17      Indices of fibrosis also were found to persist after exposure in some of the studies.  Collectively,
18      these findings provide a potential pathophysiologic basis for the changes in airway function
19      observed in children in longitudinal studies.  Seasonal ambient patterns  of exposure may be of
20      greater concern than continuous daily exposure.  In the classical study by Tyler et al. (1988),
21      seasonal exposure was associated with greater increases in total lung collagen and pulmonary
22      function changes suggestive of a delay in lung maturation in animals.
23          Controlled human exposure studies clearly demonstrated acute inflammation in the lung at
24      ambient exposure levels. Epidemiologic studies could examine whether repeated exposures over
25      multiple episode periods and/or multiple years would lead to persistent inflammation and result
26      in damage to the human lung, especially in the small, terminal bronchiolar regions where
27      vulnerability is greatest. However, the  challenges to addressing these issues in epidemiology
28      studies are formidable, and as a result there exists relatively limited literature in this area.  Long-
29      term O3 concentrations tend to be correlated with long-term concentrations of other pollutants,
30      making specific attribution difficult.  Subtle pulmonary effects require health outcome measures
31      that are sensitive, and must usually be directly collected from individual human subjects, rather

        January 2005                              7-75       DRAFT-DO NOT QUOTE OR CITE

-------
 1      than from administrative data bases. Although these factors make chronic studies difficult and
 2      expensive to conduct, efforts must be made to design studies with adequate power to examine
 3      the hypothesis being tested. Epidemiologic studies are the only approach to investigate a
 4      possible link between chronic exposure to ozone and the occurrence of human health effects.
 5           Here we review studies published from 1996 onward in which health effects were tested in
 6      relation to O3 exposures extending from several weeks to many years (Table AX7-6 in the
 7      Chapter 7 Annex). The available literature falls into four general categories:  (1) studies
 8      examining seasonal changes in lung function and lung function growth as related to O3
 9      exposures in peak season; (2) studies addressing lung function growth or decline of lung
10      function over several years in relation to long-term O3 exposures; (3) studies addressing
11      respiratory inflammation in high versus low exposure groups or time periods; and (4) studies
12      addressing longitudinal and cross-sectional associations between long-term O3 exposures and
13      asthma development and prevalence.
14
15      7.5.3  Seasonal O3 Effects on Lung Function
16           While it has been well-documented in both chamber and field studies that daily, multihour
17      exposures to O3 result in transient declines in lung function,  much less is known about the effects
18      of repeated exposures to O3 over extended periods on lung function.  Several new studies
19      reported over the past decade have examined lung function changes over seasonal time periods
20      with differing levels of O3 exposures (Frischer et al., 1999; Horak et al., 2002a,b; Ihorst et al.,
21      2004; Kinney and Lippmann, 2000; Kopp et al., 2000).  These seasonal effects of O3 are
22      examined first in this section. In the next section is a discussion of effects over years, as
23      opposed to over seasons, in addition to multiyear analyses of seasonal studies.
24           In a large study, Frischer and colleagues collected repeated lung function measurements in
25      1,150 Austrian school children (mean age 7.8 years) from nine towns that differed in mean
26      O3 levels. Mean summertime O3 exposure ranged from 32.4 to 37.3 ppb during the three
27      summers. Lung function was measured in the spring and fall over a three-year period from 1994
28      to 1996, yielding six measurements per child. The seasonal  change in lung function was
29      significantly and inversely associated with seasonal mean O3 levels. FEVj increase was lower
30      by 156.6 mL (-0.029 mL/day/ppb x 90 days/year x 3 years  x 20 ppb) for each 20 ppb increase
31      in mean 24-h avg O3 concentrations over the three summers  and  129.6 mL over the three

        January 2005                              7-76       DRAFT-DO NOT QUOTE OR CITE

-------
 1      winters.  When analyses were restricted to children who had spent the whole summer period in
 2      their community, the changes were greater, with a greater O3-related reduction of 183.6 mL in
 3      FEVj growth. Other pollutants (PM10, SO2, and NO2) had less consistent associations with lung
 4      function growth.  Horak et al. (2002a,b) extended the study of Frischer et al. (1999) with an
 5      additional year of data and indicated that seasonal mean O3 was associated with a negative effect
 6      on lung function growth, confirming results from the previous three-year study.  In an editorial,
 7      Tager (1999) stated that the Frischer et al. (1999) data provided the first prospective evidence of
 8      an association between exposure to ambient air pollution and alterations in lung growth in
 9      children. Tager further noted that the prospective study design represented a substantial
10      improvement over data derived from cross-sectional studies and should be emulated.
11           Kopp et al. (2000), in a cohort of 797 children in Austria and southwestern Germany,
12      reported lower lung function growth in children exposed to high (44 to 52 ppb O3) levels of
13      ambient O3. Children residing in low O3 (24 to 33 ppb) areas experienced a 43 mL increase
14      in FEVj whereas those in high O3  areas only experienced a 16 mL increase during the summer of
15      1994.  Similar results were found in data from the summer of 1995. In another Austrian study,
16      Ihorst et al. (2004) examined 2,153 children with a median age of 7.6 years and reported  summer
17      pulmonary function results revealing a significantly lower  FVC and FEVj increase associated
18      with higher O3 exposures in the summer, but not in the winter.
19           In a pilot study (Kinney and Lippmann, 2000), 72 nonsmoking adults (mean age 20 years)
20      from the 2nd year class of students at the U.S. Military Academy  at West Point, NY provided
21      two lung function measurements, one before and one after a five-week long summer training
22      program at four locations.  There was a greater decline in FEVj among students at the Fort Dix
23      location (78 mL)  as compared to students at the other locations (31 mL).  Ozone levels at Fort
24      Dix averaged 71 ppb (mean of daily 1-h max  O3) over the summer training period versus mean
25      values of 55 to 62 ppb at the other three locations.  In addition to the higher mean O3 level, Fort
26      Dix had greater peak O3 values (23 hours > 100 ppb) compared to the  other locations (1 hour
27      > 100 ppb). Ambient levels of other pollutants, PM10 and SO2, were relatively low during the
28      study and did not vary across the four sites. Though conclusions  are limited by the small size of
29      the study, results  are consistent with a seasonal decline in lung function that may be due,  in part,
30      to O3 exposures.  Another interesting observation from this study  was  that a larger decline was
31      observed in subjects with post-summer measurements in the first two weeks after returning from

        January 2005                               7-77        DRAFT-DO NOT QUOTE OR CITE

-------
 1      training compared to those measured in the 3rd and 4th weeks, indicating that O3-related lung
 2      function declines might be reversible.
 3           Collectively, the above studies indicate that seasonal O3 exposure is associated with
 4      declines in lung function growth in children. The study by Kinney and Lippman (2000) provide
 5      limited evidence that seasonal O3 also may affect lung function in adults, though the effect may
 6      be somewhat transient.
 7
 8      7.5.4   Chronic O3 Effects on Lung Function
 9           Lung capacity grows during childhood and adolescence as body size increases, reaches
10      a maximum during the 20s, and then begins to decline steadily and progressively with age.
11      There has long been concern that long-term exposure to air pollution might lead to slower
12      growth in lung capacity, diminished maximally attained capacity, and/or more rapid decline
13      in capacity with age. The concern arises by analogy with cigarette smoking, where it is well-
14      documented that lung function declines more rapidly with age in a dose-dependent manner
15      among  adults who smoke cigarettes. Adults who stop smoking return to a normal rate of decline
16      in capacity, although there is no evidence that they regain the capacity previously lost due to
17      smoking (Burchfiel et al., 1995). Because O3 is a strong respiratory irritant, and is associated
18      with acute lung function declines as well as inflammation and re-structuring of the respiratory
19      airways, it seems plausible that there might be a negative impact of long-term O3 exposures  on
20      lung function.  Exposures that affect lung function growth during childhood, in particular, might
21      cause greater long-term risks. Thus, studies of effects on diminished rate of lung function
22      growth in children are especially important.
23           Several studies published over the past decade have examined the  relationship between
24      lung function and long-term O3 exposure. The most extensive and robust recent study of
25      respiratory effects in relation to long-term air pollution exposures among children has been the
26      Children's Health Study carried out in 12 communities of southern California starting in 1993
27      (Peters  et al., 1999a,b; Gauderman et al., 2000,  2002, 2004a,b).  No significant associations  were
28      observed between long-term O3 exposures and self-reports of respiratory symptoms or asthma
29      (Peters  et al., 1999a).  In the initial report examining the relationship between lung function  at
30      enrollment and levels of air pollution in the community, there was evidence that annual mean O3
31      levels were associated with decreased FVC, FEVj, PEF, and FEF25.75 (the latter two being

        January 2005                              7-78        DRAFT-DO NOT QUOTE OR CITE

-------
 1      statistically significant) among females but not males (Peters et al., 1999b). Among the 4th
 2      graders, a longitudinal analysis of lung function growth over eight years indicated decrements
 3      were associated significantly with PM and NO2, but not with O3 (Guaderman et al.,  2000,
 4      2004a,b). A 2nd cohort of 4th graders were recruited in 1996 and followed over four years
 5      (Gauderman et al., 2002).  Stratified analyses by time spent outdoors indicated a
 6      significant association between decreased PEF growth and O3 exposure only in children who
 7      spent more than 1.3 hours outdoors (Guaderman et al., 2002).
 8           Ihorst et al. (2004) found that there were no associations between lung function growth rate
 9      and mean summer O3 levels for FVC and FEVj over a 3.5-year period, in contrast to the
10      significant seasonal effects discussed earlier.  Unlike the smaller increase in lung function
11      parameters over the 1st two summers among children in high O3 areas, a greater increase was
12      observed during the 3rd summer and no difference in increase was observed during the 4th
13      summer. The authors then concluded that medium-term effects on schoolchildren lung growth
14      are possibly present but are not detected over a 3- to 5-year period due to partial reversibility.
15      The study by Frischer et al. (1999) showed results similar to the Ihorst et al. (2004)  study.
16      Although a significant O3-related reduction in lung function growth was observed when three
17      years were analyzed collectively, smaller changes were observed throughout the years. FEVj
18      increase was significantly lower by 34.0 mL for each 20 ppb increase in mean 24-h  avg O3 in the
19      1st year compared to a nonsignificant but greater increase of 7.3 mL in the 3rd year (Frischer
20      et al., 1999).  Results from Horak et al. (2002a) indicated that the four-year cumulative reduction
21      in FEVj was  151.2 mL with O3 levels of 20 ppb, which was less than the cumulative estimate of
22      156.6 mL from the 1 st three years, indicating that there was little if any O3-related changes in
23      lung function growth during the 4th year.
24           Evidence for a relationship between long-term O3 exposures and decrements in maximally
25      attained lung function was observed in a nationwide cohort of 520 1st year students at Yale
26      College in New Haven, CT (Galizia and Kinney 1999; Kinney et al.,  1998). Each student
27      performed  one lung function test in the spring of their 1st year at college. Ozone exposures were
28      estimated by linking 10-year mean summer season 1-h max O3 levels at the nearest monitoring
29      station to the residential locations reported each year from birth to the time of measurement.
30      Students who had lived four or more years in areas with long-term mean O3 levels above 80 ppb
31      had significantly lower FEVj (-3.07% [95% CI: -0.22, -5.92]) and FEF25.75 (-8.11% [95% CI:

        January 2005                              7-79        DRAFT-DO NOT QUOTE OR CITE

-------
 1      -2.32, -13.90]) compared to their classmates with lower long-term O3 exposures. Stratification
 2      by gender indicated that males had much larger effect estimates than females, which might
 3      reflect higher outdoor activity levels and corresponding higher O3 exposure during childhood.
 4           A similar study of 130 1st year college freshmen at the University of California at Berkeley
 5      also reported significant effects of O3 on lung function (Kiinzli et al., 1997; Tager et al., 1998).
 6      Enrollment was limited to students from either the San Francisco or Los Angeles, CA
 7      metropolitan areas. After controlling for city of origin, long-term O3 exposures were associated
 8      with declines in FEF25.75 and FEF75 (forced expiratory flow after 75% of FVC has been exhaled).
 9      No effects were seen for PM10 and NO2.  Kiinzli and colleagues noted that significant changes in
10      these mid- and end-expiratory flow measures could be considered early indicators for pathologic
11      changes that might ultimately progress to COPD, as evidenced by animal studies that show that
12      the primary site of O3 injury in the lung is the centriacinar region (Chapter 5).  In another
13      California-based study (Gong et al., 1998b), there was no relationship between long-term
14      changes in lung function (over an approximately  10-year period) and acute responsiveness to O3
15      exposure (over a two-hour period in a controlled chamber environment) among persons living in
16      high O3 communities.
17           An autopsy pathologic study examining centriacinar region inflammatory disease was part
18      of a discussion of long-term O3 effects in the animal toxicology studies in Chapter 5.  Sherwin
19      et al. (2000) examined subjects for the above pathologic outcome in Miami, FL and Los
20      Angeles, CA residents. A trend towards greater degrees of centriacinar region alterations was
21      observed in the lungs of Los Angeles residents compared to Miami residents, independent of a
22      smoking effect.  The results suggest that the greater extent and severity of centriacinar region
23      alterations might be related to the higher O3 levels in Los Angeles.  Beyond the challenge of
24      differentiating the  lifetime of exposure for subjects in the two cities, various confounding factors
25      also can impact this study.  The pathogenesis of centriacinar region alteration is undoubtedly
26      multifactorial with respiratory infection and adverse environmental influences being two major
27      considerations.  In addition, Sherwin et al. (2000) noted that the study was limited due to the
28      relatively small number of cases available. Nonetheless, as observed by Tager (1993), the use of
29      human postmortem specimens is of interest in future epidemiology  studies.
30
31

        January 2005                              7-80       DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.5.5  Chronic O3 Exposure and Respiratory Inflammation
 2           As noted in Chapter 6, human chamber studies have demonstrated that brief (2 to
 3      6.6 hours) exposures to O3 while exercising result in inflammation in the lung, including the
 4      alveolar region where gas exchange takes place.  This acute effect is potentially important for
 5      chronic effects because repeated inflammation can result in the release of substances from
 6      inflammatory cells that can damage the sensitive cells lining the lung. Over extended periods,
 7      repeated insults of this kind could lead to permanent damage to and re-structuring of the small
 8      airways and alveoli.  In addition, since inflammation is a fundamental feature of asthma, there is
 9      concern that O3-induced inflammation can exacerbate existing asthma or perhaps promote the
10      development of asthma among genetically pre-disposed individuals.  Several studies are
11      discussed next, examining different outcomes related to inflammation.
12           In a study by Kinney et al. (1996), bronchoalveolar lavage fluids were collected in the
13      summer and winter from a group of 19 adult joggers living and working on an island in
14      New York harbor. The mean  1-h max O3 for the three-month periods were 58 ppb in the
15      summer and 32 ppb in the winter. PM10 and NO2 concentrations were similar across the two
16      seasons. There was little evidence for acute inflammation in bronchoalveolar lavage fluids
17      collected during the summer as compared to that collected from the same subjects in the winter.
18      However, there was evidence  of enhanced cell damage, as measured by lactate dehydrogenase,
19      in the summer lavage fluids. These results indicate that acute inflammatory responses may
20      diminish with repeated exposures over the course of a summer (which have been demonstrated
21      in multiday chamber exposures, Chapter 6, Section 6.9) but cell damage may be ongoing.
22           Pollution effects in the nose can be viewed as a potential surrogate  measure for effects that
23      may occur in the lungs, though doses to nasal tissues are usually higher for a given pollutant
24      concentration. In Chapter 5, morphological effects of O3 on the upper respiratory tract indicated
25      quantitative changes in the nasal transitional respiratory epithelium.  The persistent nature of the
26      O3-induced mucous cell  metaplasia in rats, as discussed in Chapter 5, suggests that O3 exposure
27      may have the potential to induce similar long-lasting alterations in the airways of humans.
28      A series of interesting studies  in Mexico City have demonstrated inflammation and genetic
29      damage to cells in the nasal passages of children chronically exposed to O3 and other air
30      pollutants (Calderon-Garciduefias et al., 1995, 1997, 1999). Nasal lavage samples and nasal
31      biopsies from children living in Mexico City were compared to those from children living in a

        January 2005                              7-81         DRAFT-DO  NOT QUOTE OR CITE

-------
 1      clean coastal town with no detectable air pollutants. In the first study, urban children (n = 38)
 2      from Mexico City were found to have significantly higher polymorphonuclear leukocyte counts
 3      and abnormal nasal cytologies compared to nonurban children (n = 28) (Calderon-Garciduefias
 4      et al., 1995). A later study observed that cells collected from the lining of the nose had
 5      significantly higher amounts of DNA damage in the urban children in Mexico City (n = 129)
 6      versus nonurban children (n = 19) (Calderon-Garciduefias et al., 1997). Among exposed
 7      children, DNA damage was greater with increasing age, suggesting an accumulation of damage
 8      over time with ongoing pollution exposures.  Another study of 86 urban and 12 nonurban
 9      children reported similar findings, in addition to increased levels of specific DNA mutations
10      (Calderon-Garciduefias et al., 1999).  They also noted far higher respiratory symptom prevalence
11      in the urban children. Fortoul et al. (2003) examined DNA strand breaks in nasal epithelial cells
12      from asthmatic and nonasthmatic medical students in Mexico City and noted greater genotoxic
13      damage in asthmatics.  These results indicate that asthmatics may have a greater vulnerability for
14      DNA damage, or a decreased ability to repair it, compared to nonasthmatic subjects.
15          Another outcome of inflammation was examined in a study by Frischer et al. (2001).
16      In this cross-sectional study, urinary eosinophil protein was analyzed as a marker of eosinophil
17      activation in 877  school children living in nine Austrian communities with varying O3 exposure.
18      The results indicated that O3 exposure was significantly associated with eosinophil
19      inflammation.
20          In the Mexico City studies, specific attribution of these adverse respiratory and genotoxic
21      effects to O3 is difficult given the complex mixture of pollutants present in the ambient air.
22      In particular, the DNA effects seem more plausibly related to other components  of urban air,
23      such as semi-volatile organic compounds. However, the inflammatory  changes such as
24      increased eosinophil levels observed in the Austrian study would be consistent with known
25      effects of O3.
26
27      7.5.6   Risk of Asthma  Development
28          Recent reports from longitudinal cohort studies have reported associations between
29      the onset of asthma and long-term O3 exposures (McConnell et al., 2002; McDonnell et al.,
30      1999). Significant associations between new cases of asthma among adult males and long-term
31      O3 exposure were observed in a cohort of nonsmoking adults in California (Greer et  al., 1993;

        January 2005                              7-82        DRAFT-DO NOT QUOTE OR CITE

-------
 1     McDonnell et al., 1999). The Advent!st Health and Smog (AHSMOG) study cohort of 3,914
 2     (age 27-87 years, 36% male) was drawn from nonsmoking, non-Hispanic white California
 3     seventh day Adventists.  Subjects were surveyed in 1977, 1987, and 1992. To be eligible,
 4     subjects had to have lived 10 or more years within 5 miles of their current residence in 1977.
 5     Residences from 1977 onward were followed and linked in time and space to interpolated
 6     concentrations of O3, PM10,  SO2, and NO2. New asthma cases were defined as self-reported
 7     doctor-diagnosed asthma at  either the 1987 or 1992 follow-up questionnaire among those who
 8     had not reported having asthma upon enrollment in 1977. During the  10 year follow-up (1977-
 9     1987), the incidence of new asthma was 2.1% for males and 2.2% for females (Greer et al.,
10     1993). A relative risk of 3.12 per 10 ppb increase in annual mean O3 was observed in males,
11     compared to a non-significant relative risk of 0.94 in females. In the 15-year follow-up study
12     (1977-1992), 3.2% of the eligible males and a slightly greater 4.3% of the eligible females
13     developed adult asthma (McDonnell et al., 1999). For males, the relative risk of developing
14     asthma was 2.27 per 30 ppb increase in 8-h avg O3 (9 a.m.-5  p.m.).  Once again, there was no
15     evidence of an association between O3 and new-onset asthma in females (relative risk of 0.85).
16     The lack of an association does not necessarily indicate no effect of O3 on the development of
17     asthma among females. For example, differences in time-activity patterns in females and males
18     may influence relative exposures to O3, leading to greater misclassification of exposure in
19     females. The consistency of the results in the two studies with different follow-up times and
20     indices of O3 exposure provides evidence that long-term O3 exposure may be associated with
21     asthma incidence in adult males.  However, as the AHSMOG cohort was drawn from a narrow
22     subject definition, the representativeness of this cohort to the general U.S. population may be
23     limited.
24           A similar study of incident asthma cases in relation to O3 among children was carried out
25     in the Children's Health Study (McConnell et al., 2002).  Annual surveys of 3,535 initially
26     nonasthmatic children (ages 9 to 16 years at enrollment) enabled  identification of new-onset
27     asthma cases through 1998.  Communities were stratified by pollution levels, with six high-O3
28     communities (mean 1-h max O3 of 75.4 ppb over four years) and  six low-O3 communities
29     (50.1 ppb). Asthma risk was not higher for residents of the six high-O3 communities versus
30     residents of the six low-O3 communities. However, within the high-O3 communities, asthma risk
31     was 3.3 times greater for children who played three or more sports as  compared with children

       January 2005                              7-83        DRAFT-DO NOT QUOTE OR CITE

-------
 1      who played no sports. This association was absent in the low-O3 communities (relative risk of
 2      0.8).  No associations with asthma were seen for PM10, PM2 5, NO2, or inorganic acid vapors.
 3      These results suggest effect modification of the impacts of O3 on asthma risk by physical
 4      activity. Playing sports may indicate outdoor activity when O3 levels are higher and an
 5      increased ventilation rate, which may lead to increased O3 exposure.  Replication of these
 6      findings in other cohorts would lend greater weight to a causal interpretation.
 7           Recent cross-sectional surveys have detected no associations between long-term O3
 8      exposures and asthma prevalence, asthma-related symptoms, or allergy to common aeroallergens
 9      in children after controlling for covariates (Charpin et al., 1999; Kuo et al., 2002; Ramadour
10      et al., 2000). However, reported O3 levels were quite low in all cases, with a range of 16 to
11      27 ppb for 8-h max O3. In addition, compared to the longitudinal study design, which observes
12      new onset of asthma prospectively, the cross-sectional study design is inherently weaker.
13      Longitudinal studies provide the strongest evidence on the question of asthma development and
14      is the preferred approach for future research.
15
16      7.5.7  Respiratory Effects of Chronic O3 Exposure on Susceptible
17             Populations
18           Studies on the effect of long-term O3 exposure on respiratory health has mostly focused on
19      potentially susceptible populations, including children and individuals who exercise outdoors, as
20      discussed in this section.  Ozone exposure was associated significantly with declines in lung
21      function or reduced lung function growth, and respiratory inflammation in these susceptible
22      populations.
23           Other studies have investigated additional symptoms and groups of potentially susceptible
24      individuals. McConnell et al. (1999) examined the association between O3 levels and the
25      prevalence of chronic lower respiratory tract symptoms in southern California children with
26      asthma (n = 3,676). In this questionnaire-based study, bronchitis, phlegm, and cough were not
27      associated with annual mean O3 concentrations in children with asthma or wheeze.  All other
28      pollutants examined, PM10, PM2 5, NO2, and gaseous acid, was associated with  an increase in
29      phlegm, but not cough.
30           One new study examined a susceptible group not examined before.  Goss et al. (2004)
31      investigated the effect of O3 on pulmonary exacerbations and lung function in individuals with

        January 2005                              7-84        DRAFT-DO NOT QUOTE OR CITE

-------
 1      cystic fibrosis over the age of 6 years (n = 11,484). The study included patients enrolled in the
 2      Cystic Fibrosis Foundation National Patient Registry. The registry contained demographic and
 3      clinical data collected annually at accredited centers for cystic fibrosis.  In 1999 and 2000, the
 4      annual mean O3 concentration from 616 monitors in the U.S. EPA Aerometric Information
 5      Retrieval System (AIRS) was 51.0 ppb (SD 7.3). Exposure was assessed by linking air pollution
 6      values from AIRS with the patient's home zip code. No clear association was found between
 7      annual mean O3 and lung function parameters.  However, a 10 ppb increase in annual mean O3
 8      was associated with a 10% (95% CI: 3, 17) increase in the odds of two or more pulmonary
 9      exacerbations. Significant excess odds of pulmonary exacerbations also were observed with
10      increased annual mean PM10 and PM2 5 concentrations.
11           In summary, several studies have identified and investigated potentially susceptible
12      populations.  Although effects are not specific to O3 exposure, the results suggest that O3 likely
13      contributes to the adverse respiratory health responses observed in these populations.
14
15      7.5.8   Mortality Effects of Chronic O3 Exposure
16           There is inconsistent and inconclusive evidence for a relationship between long-term O3
17      exposure and increased mortality risk (Table AX7-7 in the Chapter 7 Annex).  A long-term
18      prospective cohort study (AHSMOG; 1977-1992) of 6,338 nonsmoking, non-Hispanic white
19      subjects living in California found a significant association between long-term O3 exposure and
20      increased lung cancer risk among males only (Beeson et al., 1998). The relative risk for lung
21      cancer incident among males was 3.56 (95% CI: 1.35, 9.42) per 556 hours/year when O3 levels
22      exceeded 100 ppb (Beeson et al., 1998). A stronger association was observed in males who
23      never smoked (4.48 [95% CI: 1.25,  16.04]) compared to those who smoked in the past (2.15
24      [95% CI: 0.42, 10.89]) (Beeson et al., 1998). An expanded study by Abbey et al. (1999)
25      examining mortality effects of long-term O3 exposure in the same study population confirmed
26      the results of the previous study by Beeson and colleagues. The association between lung cancer
27      mortality and chronic O3 exposure was significant in males only, with a relative risk of 4.19
28      (95% CI: 1.81, 9.69) (Abbey et al.,  1999). However, the very small numbers of lung cancer
29      deaths (12 for males and 18 for females) raise concerns in regards to the precision of the effect
30      estimate, as evidenced by the wide confidence  intervals. No other mortality outcomes were
31      found to be associated with chronic O3 exposure. A particular strength of this study was the

        January 2005                             7-85       DRAFT-DO NOT QUOTE OR CITE

-------
 1      extensive effort devoted to assessing long-term air pollution exposures, including interpolation
 2      to residential and work locations from monitoring sites over time and space. However, the
 3      observation of a lung cancer effect but no effect on cardiopulmonary mortality raises concerns.
 4      The gender-specific O3 effects may be partially attributable to the differences in activity and time
 5      spent outdoors by gender. The questionnaires indicated that males spent approximately twice as
 6      much time outdoors and performed more vigorous outdoor exercises, especially during the
 7      summer, compared to the females.
 8          Lipfert et al. (2000b, 2003) reported positive effects on mortality for peak O3 exposures
 9      (95th percentile levels) in the U.S. Veterans Cohort study of approximately 50,000 male middle-
10      aged men recruited with a diagnosis of hypertension.  The actual analysis involved smaller
11      subcohorts based on exposure and mortality follow-up periods.  Four separate exposure periods
12      were defined as follows:  1960-1974; 1975-1981;  1982-1988; and 1989-1996. Three mortality
13      follow-up periods were considered:  1976-1981; 1982-1988; and 1989-1996. In a preliminary
14      screening of regression results, Lipfert et al. (2000b) compared univariate and multivariate
15      models by mean and peak (95th percentile) O3 concentrations.  For mean O3, a significant
16      negative relationship was reported in the univariate model and a nonsignificant negative
17      relationship was found in the multivariate model. For peak O3 concentration, the univariate
18      model indicated a nonsignificant positive relationship and the multivariate model resulted in a
19      significant positive relationship.  Peak O3 was used in subsequent analyses.  The mean of the
20      peak values ranged from 85 to 140 ppb over the four exposure periods.  For concurrent exposure
21      periods, the mortality risk was significant, with a 6.1% excess risk per mean 95th percentile O3
22      less estimated background level (not stated).  When exposure periods preceding death were
23      considered, the excess mortality risk was nonsignificant (-0.2%). In a further analysis, Lipfert
24      et al. (2003) reported the strongest positive association for concurrent exposure to peak O3 for
25      the subset with low diastolic blood pressure during the period of 1982-1988.  Once again, the O3
26      effect was found to be reduced when the exposure (1982-1988)  preceded mortality (1989-1996).
27          No effect of long-term O3 concentrations on mortality risk was observed in a larger
28      prospective cohort study of approximately 500,000 U.S. adults (Pope et al., 2002). Strong and
29      consistent effects of PM2 5 were observed in this study for both lung cancer and cardiopulmonary
30      mortality.
31

        January 2005                               7-86       DRAFT-DO NOT QUOTE OR CITE

-------
 1     7.5.9   Summary of Chronic O3 Effects on Morbidity and Mortality

 2         •  In the past decade, important new longitudinal studies have examined the chronic
              effect of O3 exposure on respiratory health outcomes. Evidence from recent long-term
              morbidity studies have indicated that chronic exposure to O3 may be associated with
              declines in lung function, inflammation, and development of asthma in children and
              adults. Seasonal decrements or reduced growth in lung function measures have been
              reported in several studies, however changes appear to be transient.  Studies of lung
              function declines with longer-term or annual data are not as conclusive.
 3         •  Few studies have investigated the effect of long-term O3 exposure on mortality.
              Uncertainties regarding the exposure period of relevance, and inconsistencies across
              mortality outcomes and gender raise concerns regarding plausibility.  The current
              evidence is inconclusive for a relationship between chronic O3 exposure and increased
              risk of mortality.

 4
 5
 6     7.6   INTERPRETATIVE ASSESSMENT OF THE EVIDENCE IN
 7            EPIDEMIOLOGIC STUDIES OF OZONE HEALTH EFFECTS
 8     7.6.1   Introduction
 9          In the 1996 O3 AQCD, the epidemiology section focused primarily on individual-level
10     camp and exercise studies. These field studies indicated exposure-response relationships
11     between O3 exposure from the ambient air, and declines in pulmonary function, increases in
12     respiratory symptoms, and exacerbation of asthma, especially in children. Numerous new
13     studies provide  additional evidence for evaluating associations between O3 exposure and the
14     above respiratory health outcomes. The 1996 O3 AQCD review of aggregate population time-
15     series studies suggested an association between ambient O3 concentrations and increased
16     hospitalizations. Limited studies examined the O3-mortality relationship. The current O3 AQCD
17     further presents results from time-series studies that have addressed previously unresolved
18     issues regarding potential linkages between ambient O3 concentrations and health outcomes,
19     particularly mortality. Daily time-series studies minimize confounding by population
20     characteristics (e.g., cigarette smoking, diet, occupation) by following the same population from
21     day to day. However, confounders operating over shorter time scales can affect O3 risk estimates
22     in time-series studies.
       January 2005                             7-87        DRAFT-DO NOT QUOTE OR CITE

-------
 1           In this section, the issues and attendant uncertainties that affect the interpretation of O3
 2      health effects will be discussed. The consequences of using stationary ambient monitors as an
 3      estimate of personal exposure in epidemiology studies will be examined first and a discussion of
 4      the temporal relationship between O3 exposure and the occurrence of health effects will follow.
 5      Of particular interest are the issues arising from model specifications in time-series studies to
 6      adjust for confounding by temporal factors, meteorological effects, and copollutants.  The shape
 7      of the concentration-response relationship and heterogeneity of O3 health effects also will be
 8      discussed briefly. All of these issues are of much importance for characterizing and interpreting
 9      ambient O3-health effects associations.
10
11      7.6.2   Exposure Assessment
12           Various methods have been used to assess exposure in air pollution epidemiology studies.
13      Navidi et al. (1999) describes the two methods commonly used to ascertain personal exposure:
14      (1) the microenvironmental (indirect) approach; and (2) the personal sampling (direct) approach.
15      Both methods are associated with measurement error. To determine personal exposure using the
16      microenvironmental approach, the concentrations of the various microenvironments are
17      multiplied by the time spent in each microenvironment. Both the concentration and time
18      component contribute to the measurement error.  Although there is no time component to the
19      measurement error in the personal sampling approach, the estimation of exposure using personal
20      monitoring devices  do contribute to the error, especially in the case of O3. The passive badges
21      commonly used for personal sampling of O3 provide integrated personal exposure information.
22      Their sensitivity to wind velocity, badge placement, and interference with other copollutants
23      may result in measurement error. Results from the error analysis models developed by Navidi
24      et al. (1999) indicated that neither the microenvironmental or personal sampling approach gave
25      reliable health effect estimates when measurement errors were uncorrected. The nondifferential
26      measurement error biased the effect estimates toward zero under the model assumptions.
27      However, if the measurement error was correlated with the health response, a bias away from the
28      null could result.  The use of central ambient monitors to estimate exposure also biased the
29      estimates towards the null.  Most people  spend the majority of their time indoors, where O3
30      levels tend to be much lower than outdoor ambient levels.  Using ambient concentrations to
        January 2005                              7-88        DRAFT-DO NOT QUOTE OR CITE

-------
 1      determine exposure generally overestimates true personal O3 exposure, resulting in effect
 2      estimates biased towards the null.
 3
 4      7.6.2.1  Relationship between Ambient Concentrations and Personal Exposure to O3
 5           Several studies have examined the relationship between ambient O3 concentrations from a
 6      central monitoring site and personal O3 exposure (Avol et al., 1998; Brauer and Brook, 1997;
 7      Chang et al., 2000; Delfmo et al., 1996; Lee et al., 2004; Liard et al., 1999; Linn et al., 1996; Liu
 8      et al.,  1995, 1997; O'Neill et al., 2003; Sarnat et al., 2001).  In a Baltimore, MD study of older
 9      adults, individuals with COPD, and children, 24-h avg ambient O3 concentrations from a
10      monitoring site were not found to be significantly associated with personal O3 exposure (Sarnat
11      et al., 2001). The mixed regression effect estimates were P = 0.01 (t = 1.21) and P = 0.00
12      (t = 0.03), for summer and winter, respectively. A study by O'Neill et al. (2003), in contrast,
13      found a statistically significant association between personal and ambient O3 concentrations in
14      Mexico City outdoor workers (P =  0.56, t =  8.52). The subjects in the Sarnat et al. (2001) study
15      spent less than 6% of their time outdoors, whereas the personal exposure data from O'Neill et al.
16      (2003) were from subjects who spent the entire measurement period outdoors. A scripted
17      exposure study by Chang et al. (2000) provided supportive evidence for the conflicting results in
18      the Sarnat et al. (2001) and O'Neill et al. (2003) studies. In this study, activities were scripted to
19      simulate activities typical of older adults living in Baltimore, MD. Their activities were derived
20      from the U.S. EPA-sponsored National Human Activity Pattern Survey study (Klepeis, 1999).
21      Chang et al. (2000) compared one-hour personal and ambient O3 measurements in several
22      microenvironments. There was no correlation between personal and ambient O3 concentrations
23      in the  indoor residence (r = 0.09 and r = 0.05, for summer and winter, respectively), although a
24      moderate correlation was found in other indoor environments such as restaurants, hospitals, and
25      shopping malls (r = 0.34 in summer, r = 0.46 in winter). In comparison, the correlation in
26      outdoor environments (near and away from roads) was moderate to high (0.68 < r < 0.91) and
27      statistically significant.
28           Brauer and Brook (1997) observed that the daily averaged personal O3 measurements and
29      ambient concentrations were well-correlated after stratifying groups by time spent outdoors.
30      The clinic workers (n = 25) spent 9% of their time outdoors (24-hour samples) whereas the farm
31      workers (n = 15) spent 100% of their monitored time outdoors (6-14 hour workshift samples).

        January 2005                              7-89        DRAFT-DO NOT QUOTE OR CITE

-------
 1      The personal to ambient O3 concentration ratios were significantly different for the clinic
 2      workers (0.28) and farm workers (0.96). However, the Spearman correlation coefficients were
 3      similar in the two groups, 0.60 and 0.64 for the clinic workers and farm workers, respectively.
 4      In a Paris, France study by Liard et al. (1999), adults (n = 55) and children (n = 39) wore passive
 5      O3 monitors for 4 consecutive days during three periods.  For each period, all adults wore the O3
 6      monitors over the same 4 days.  Likewise, all children wore monitors over the same 4 days for
 7      each of the three periods, but on different days from the adults.  The ambient O3 concentrations
 8      from the stationary monitoring sites did not explain a high percentage of the variance of personal
 9      O3 exposure (non-significant [value not stated] in adults and 21% in children). However, when
10      personal measurements from all subjects were aggregated for each of the six periods, the 4-day
11      mean personal O3 exposure was found to be highly correlated with the corresponding mean
12      ambient concentration (r = 0.83, p < 0.05). Similarly, a study of Los Angeles school children by
13      Linn et al. (1996) found that daily 24-h avg ambient O3 concentrations from a central site were
14      well-correlated (r = 0.61) with daily averaged personal O3 concentrations (8-10 children/day, 132
15      total monitoring days).
16           The low correlation observed between personal  O3 exposures and ambient O3
17      concentrations in the study by Sarnat et al. (2001) suggests that O3 concentrations measured at
18      central ambient monitors do not explain the variance of individual personal exposures.
19      However, in time-series studies, daily averaged personal  exposures from the aggregate
20      population is of greater relevance than exposures from specific individuals. Although
21      unresolved issues do remain, the limited evidence indicates that ambient O3 concentrations from
22      central monitors may serve as valid surrogate measures for aggregate personal O3 exposures in
23      population time-series studies investigating mortality  and hospitalization outcomes.
24
25      7.6.2.2  Factors Affecting the Relationship between Ambient Concentrations and
26              Personal Exposures to O3
27           In cohort studies investigating acute and chronic morbidity outcomes, O3 exposure
28      assessment may be improved by accounting for the distance between home and the monitoring
29      site, time-activity patterns (e.g., percentage of time spent outdoors, type of outdoor activity, time
30      of day during outdoor activity), and elements that affect indoor air exchange rates (e.g.,
31      ventilation conditions, home characteristics) in conjunction with the ambient O3 data from
32      stationary monitor sites. Several studies (Geyh et al., 2000; Lee et al., 2004; Liu et al., 1995,

        January 2005                               7-90        DRAFT-DO NOT QUOTE OR CITE

-------
 1      1997) demonstrated that the association between personal O3 exposure and ambient O3
 2      concentrations is affected by these factors.  A study by Geyh et al. (2000) observed higher indoor
 3      and personal O3 concentrations in a southern California community with 2% air conditioned
 4      homes compared to a community with 93% air conditioned homes during the summer (high O3)
 5      months, but showed no difference in O3 levels during the winter (low O3) months.  Lee et al.
 6      (2004) observed that personal O3 exposure  was positively correlated with outdoor time (r = 0.19,
 7      p < 0.01) and negatively correlated with indoor time (r = -0.17, p < 0.01). Additional factors that
 8      affected indoor O3 levels were air conditioning, window fans, and window opening.  The O3
 9      exposure assessment study by Liu et al.  (1995) found that after adjusting for time spent in
10      various indoor and outdoor microenvironments (e.g., car with windows open, car with windows
11      closed, school, work, home, outdoors near home, outdoors other than near home), mean 12-hour
12      ambient O3 concentrations explained 32% of the variance in personal exposure in the summer.
13           Other factors, including age, gender, and occupation, also may affect general exposure
14      patterns to O3 by influencing time-activity patterns.  In a southern California study by Avol et al.
15      (1998), boys were found to spend more  time outdoors and be more physically active than girls
16      (Avol et al., 1998). Another southern California study found that boys were outdoors 30 minutes
17      longer than girls, and had higher personal O3 exposure during both high and low O3 months
18      (Geyh et al., 2000). Outdoors workers also tend to be exposed to higher levels of O3 (Brauer and
19      Brook, 1997; O'Neill et al., 2003).
20           The announcement of air quality indices also may influence personal exposures to O3.
21      Niedell (2004) examined the effect of smog alerts on the relationship between O3 and hospital
22      admissions for asthma in California children aged 0 to 18  years. Air quality episodes, or smog
23      alerts, are issued in California on days when O3 concentrations exceed 200 ppb. Smog alerts
24      were found to have a significant negative effect on asthma admissions for children aged 1 to
25      12 years, providing supportive evidence that avoidance behavior might be present on days of
26      high O3 concentrations.  Avoidance behavior may include staying indoors and exercising less on
27      days when a smog alert is announced, resulting in reduced exposure to O3. Therefore air quality
28      indices may affect the relationship between ambient and personal O3 concentrations.
29           In summary, results indicate that the relationship between ambient and personal O3
30      concentrations varies depending on factors  such as time spent outdoors, ventilation conditions,
31      personal factors, and air quality indices.  The use of questionnaires to obtain information on

        January 2005                              7-91        DRAFT-DO NOT QUOTE OR CITE

-------
 1      personal characteristics, time-activity patterns, and home characteristics may improve further the
 2      accuracy of the personal O3 exposure estimates from ambient O3 data.
 3
 4      7.6.2.3  Assessing Chronic Exposure to O3
 5           Several studies examined methods of estimating chronic exposure to O3. A pilot study
 6      (n = 14) by Gonzales et al. (2003) indicated that the use of retrospective questionnaires might be
 7      a reliable method for reconstructing past time-activity and location pattern information.  The
 8      7-day, 24-h avg O3 exposures were estimated using data from an ambient monitor (mean
 9      29.5 ppb),  and information from prospective diaries and questionnaires completed one year after
10      the monitoring period.  The O3 estimates from both prospective diaries (mean 10.6 ppb) and
11      retrospective questionnaires (mean 11.8 ppb) differed only slightly, although both estimates were
12      greater than the  personal exposure measurement (mean 5.7 ppb). A study by Kiinzli et al. (1997)
13      compared the retrospective assessments of outdoor time-activity patterns using three formats, a
14      questionnaire, a table, and a 24-hour log.  College freshmen (n = 44) noted activity patterns at
15      their last residence using two or three methods and then were retested 5-7 days later. The
16      within-subject variance in reporting moderate to heavy activity was 13% for both the activity
17      questionnaire and activity table, and 32% for the 24-hour log. The data from the activity tables
18      also were similar to data published from the California Air Resources Board (CARB) 24-hour
19      recall diary study (Jenkins et al., 1992).  Results from the above studies suggest that the use of
20      activity questionnaires or activity tables to determine time-activity patterns may be suitable in
21      large retrospective epidemiologic studies of health effects requiring estimates of chronic
22      exposure to ambient O3.
23
24      7.6.3   O3 Exposure Indices
25           Three O3 indices  were used often to indicate daily O3 exposure:  maximum 1-h average
26      (1-h max); maximum 8-h average (8-h max); and 24-h average (24-h avg). The 8-h max O3 is a
27      frequently used  index in newer epidemiologic studies, as it best reflects the current U.S. EPA
28      standard. These O3 exposure indices are highly correlated as indicated in the following  studies.
29      In the 21 European multicities acute mortality study (Gryparis et al., 2004),  1-h max O3  was
30      found to be highly correlated with 8-h max O3, with a median correlation  coefficient of 0.98
31      (range: 0.91 - 0.99).  Among single city studies, the 1-h max O3 and 8-h max O3 also were found

        January 2005                               7-92       DRAFT-DO NOT QUOTE OR CITE

-------
 1      to have correlation coefficients ranging from 0.91 to 0.99 in various cities such as Atlanta, GA
 2      (Tolbert et al., 2000; White et al., 1994); southern New England (Gent et al., 2003); Ontario,
 3      Canada (Burnett et al., 1994); and Mexico City (Loomis et al., 1996; Romieu et al., 1995).
 4      In addition, 1-h max O3 was found to be highly correlated with 24-h avg O3, as observed in the
 5      Mexico City study by Loomis et al. (1996) (r = 0.77) and in the Ontario, Canada study by
 6      Burnett et al. (1994) (r = 0.87).
 7           All studies discussed in Sections 7.2 to 7.5 were examined for presentation of the three O3
 8      exposure indices.  Several presented the concentration data and correlations among 1-h max, 8-h
 9      max, and 24-h avg O3 ambient measures.  Some presented the associated risk estimates of
10      comparable analyses for the three exposure indices.  No papers provided an analysis statistically
11      comparing the indices. Summary of the available data is provided below starting with two
12      multeity mortality studies.
13           In the large U.S. 95 communities study by Bell et al. (2004), increases in O3-associated
14      daily mortality were estimated using all three O3 indices. The increments used in this document
15      to standardize expressions of excess risks are 40 ppb for 1-h max O3, 30 ppb for 8-h max O3, and
16      20 ppb for 24-h avg O3, as discussed in Section 7.1.3.2. For these increments, the effect
17      estimates calculated by Bell et al. (2004) were 1.34%,  1.28%, and 1.04% excess risk in mortality
18      for 1-h max O3, 8-h max O3, and 24-avg O3, respectively. A statistical test examining
19      differences among these risk estimates indicated that there were no significant differences by
20      exposure index. In the European study of 21 cities (of the 23  cities, two did not have 8-h max O3
21      data), the O3-mortality effect estimate for the summer season  was slightly smaller for 8-h max
22      O3, 1.82% excess risk, compared to 1-h max O3, 2.59% excess risk, but both were statistically
23      significant (Gryparis et al., 2004). Once again, a statistical test between the two risk estimates
24      yielded no significant difference between the indices.
25           Several single city mortality studies examined  multiple  O3 exposure indices (Anderson
26      et al., 1996; Dab et al., 1996; Sunyer et al., 2002; Zmirou et al., 1996; Borja-Aburto et al., 1997).
27      These studies did not differentiate risk estimates by exposure  index as the results were
28      considered similar. Hospital admission studies also provided limited data for index
29      comparisons.  Schouten et al. (1996) showed the same positive nonsignificant association
30      between total respiratory hospitalizations and O3 using both 8-h max O3 and 1-h max O3 in the
31      summer (4.0% excess risk per standardized increment). For emergency department visits, the

        January 2005                             7-93         DRAFT-DO NOT QUOTE OR CITE

-------
 1      examples of Delfmo et al. (1998) and Weisel et al. (2002) provided data not indicative of
 2      differences between the indices. A further example, Tolbert et al. (2000) noted an increase in
 3      emergency room visits of 4.0% per standard deviation increase (approximately 20 ppb) for both
 4      1-h max O3 and 8-h max O3 as being expected since the correlation between the indices was
 5      0.99.
 6           Peak flow asthma panel studies generally only used one index in these studies, thus no
 7      comparison data is available. One respiratory symptom study (Gent et al., 2003) did examine
 8      both 1-h max O3 and 8-h max O3 but noted no differences in the results. Only one FEVj panel
 9      study examined more than one exposure index. Chen et al. (1999) examined 1-h max O3 and
10      24-h avg O3 and reported at a 1-day lag for children a decrement in FEVj of -25.6 mL (-49.1,
11      -2.1) for 1-h max O3 and -13.6 mL (-33.2, 6.0) for 24-h avg O3. For 2- and 7-day lags,  smaller
12      differences were observed between the two indices. None of these FEVj outcomes, including
13      those for a 1-day lag, upon testing were significantly different by index.
14           Limited information is available to reach conclusions for comparison of the three indices
15      1-h max O3, 8-h max O3, and 24-h avg O3. In general, for the same distributional increment
16      (e.g., interquartile range), the excess health risk estimates and significance of associations were
17      comparable for all three daily O3 indices. The similarity in effects for the three exposure indices
18      was consistent for all outcomes. This is expected due to the high correlation among the indices.
19           The relationship  between the various health endpoints and the three O3 exposure indices,
20      and the associated study designs and analyses are such  that the high correlation among the
21      indices presents a significant challenge in distinguishing the most appropriate measure for
22      epidemiologic studies. The commonly used 8-h max O3 or 8-h avg O3 index continues to be an
23      appropriate choice as no other exposure index has been demonstrated to offer a better advantage.
24
25      7.6.4   Lag Time:  Period between O3 Exposure and Observed Health Effect
26           The choice of lag days for the relationship between exposure and health effects depends on
27      the hypothesis being tested and the mechanism involved in the expression of the outcome.
28      Effects can occur acutely with exposure on the same or previous day, cumulatively over several
29      days, or after a delayed period of a few days. With knowledge of the mechanism of effect, the
30      choice of lag days can be determined prior to analysis.  For example, one can expect cough to
31      occur acutely after exposure with a lag of 0 or 1 day, as O3 can act as an immediate irritant.

        January 2005                              7-94         DRAFT-DO NOT QUOTE OR CITE

-------
 1      However, an O3-related inflammatory response may not lead to asthma exacerbation until
 2      several days later. An asthmatic may be impacted by O3 on the first day of exposure, have
 3      effects triggered further on the second day, then report to the emergency room for an asthmatic
 4      attack three days after exposure. Further, within a population of asthmatics, exacerbation of
 5      asthma symptoms may be observed over a period of several days, since each asthmatic has
 6      individual aspects of the disease and may be affected by the exposure differently depending on
 7      his/her sensitivity and disease severity.  Therefore, it may be necessary to examine longer lag
 8      periods to fully understand the relationship between O3 exposure and effect. When the
 9      mechanism of the health effect is unknown, investigating the association between outcome and
10      exposure over several days also may be informative.
11           Most of the O3 time-series studies investigated a small number of lagged days, typically
12      0 through 3 days, and/or cumulative lag periods. For outcomes of mortality and hospitalizations,
13      the largest, most significant associations with O3 concentrations were observed when using short
14      lag periods, in particular a 0-day lag (exposure on same  day) and a 1-day lag (exposure on
15      previous day).  In the U.S. 95 communities study by Bell et al. (2004), the largest risk estimate
16      for O3-mortality was obtained with a 0-day lag, followed by diminishing risk estimates with 1-,
17      2-, and 3-day lags.  In a study of 16 Canadian cities by Burnett et al. (1997a), the strongest
18      association between O3 and respiratory hospitalizations was found at a 1-day lag.  Once again,
19      there was a decline in the magnitude and significance of the effect with increasing days lagged
20      for O3. These results  suggest that O3 has a rapid effect on these respiratory health outcomes.
21           Less explored is the issue of multiday effects of O3. When associations are found at
22      multiple lag days, results from a single-day model may underestimate the  cumulative effect of
23      O3 on health outcomes. In the large U.S. 95 communities study (Bell et al., 2004), distributed
24      lag models were used to estimate the effect of O3 levels on mortality over a lag of 0 to 6 days.
25      Results indicated that when accounting for multiple days, the effect estimates were twice as large
26      as those from single-day analyses, as shown in Figure 7-15. Bell et al. (2004) estimated a
27      cumulative excess risk of 1.04% on daily mortality per 20 ppb increase in 24-h avg O3 during the
28      previous week, compared to a 0.50% excess risk associated with O3 exposure on a single day, in
29      this case a 0-day lag.  In a related  U.S. study of the 19 largest cities by Huang et al. (2004), the
30      O3 estimate for the  summer season was 1.50% excess risk of cardiopulmonary mortality with
31      current-day exposure  and 2.52% for a 7-day cumulative  lag.

        January 2005                               7-95        DRAFT-DO NOT QUOTE OR CITE

-------
                                              0.5             1
                                          % Change in Mortality
                         1.5
       Figure 7-15.  Comparison of single-day lags (0-, 1-, 2-, and 3-day) to a cumulative multiday
                     lag (0- to 6-day) for % changes in all cause mortality per 20 ppb increase in
                     24-h avg O3 in all ages.
       Source: Derived from Bell et al. (2004).
 1          Burnett et al. (2001) investigated the association between respiratory hospitalizations and
 2     O3 in children less than 2 years of age. Lags up to five days were examined after stratifying by
 3     season (Figure 7-16).  In the summer season, significant associations between O3 and daily
 4     admissions were found in several of the lags, with the largest risk estimate of 12.5% excess risk
 5     per 40 ppb increase in 1-h max O3 at a 1-day lag. In comparison, the O3-related risk estimate
 6     was 30.2% using a cumulative lag period of 5 days.  In another study, Anderson et al. (1997)
 7     investigated the association between O3 and daily hospital admissions for COPD in five
 8     European cities.  Lags up to 5 days were examined, and the largest risk estimates were found
 9     using 0- and 1-day lags. Anderson et  al. observed a 4.5% excess risk per 40 ppb increase in 1-h
10     max O3 using a single-day lag compared to a 7.7% excess risk using a 5-day cumulative lag.
11          Among the field studies, Mortimer et al. (2002) reported O3-related changes in PEF for
12     single-day lags from 1 to 6 days and a multiday lag period of 5 days. No associations were seen
13     between evening outcome measures and either single-day or multiday exposure lags. Small,
14     nonsignificant morning effects were observed at 1- and 2-day lags. The effect of O3 on morning
15     outcomes increased over  several days, with the strongest association seen using multiday lag
       January 2005
7-96
DRAFT-DO NOT QUOTE OR CITE

-------
              0.20 R
                                0          10         20          30         40
                                  % Change in Respiratory Hospitalization
                                        50
       Figure 7-16.  Comparison of single-day lags (0-, 1-, 2-, 3-, 4-, and 5-day) to a cumulative
                     multiday lag (0- to 4-day) for % changes in total respiratory hospitalizations
                     per 40 ppb increase in 1-h max O3 in children less than two years of age.
       Source: Derived from Burnett et al. (2001).
 1     periods (Figure 7-17). Unrestricted lag models suggested that the O3 exposure from 3 to 5 days
 2     prior had a greater impact on morning % PEF than more immediate exposures.  Results were
 3     similar when comparing single- and multiple-day exposure lags on the incidence of respiratory
 4     symptoms in the morning (Mortimer et al., 2002).
 5          Weisel et al. (2002) stated that a lag period of 1 to 3 days between exposure to O3 and
 6     hospital admissions or emergency department visits for asthma was plausible because it might
 7     take time for the disease to progress to the most serious responses following exposure.
 8     In addition, taking medication could delay further the progression of the adverse effect.
 9     Mortimer et al. (2002) discussed biological mechanisms for delayed effects on pulmonary
10     function, which included increased bronchial reactivity secondary to airway inflammation
11     associated with irritant exposure.  Animal toxicology and human chamber studies (see
12     Chapters 5 and 6) provide further evidence that exposure to O3 may augment cellular infiltration
13     and cellular activation, enhance release of cytotoxic inflammatory mediators, and alter
14     membrane  permeability. Examining longer lag periods allows studies to investigate the
15     cumulative O3-related effects over several days rather than  one day only. The use of longer lag
16     periods also allows for delayed effects at 3 to 6 days to be observed. However, interpretation of
       January 2005
7-97
DRAFT-DO NOT QUOTE OR CITE

-------
              2.0
               1.5
           g   1.0
               0.5
               0.0
                                     Single-day Lags
                                           2
                                        / xX
                      Multiday Lag 1-5    [  /   \
                                                        3/ /
      1
       \
      \\
       \  \
       \  \
       c\  \
       6 \  \
-3     -2.5     -2     -1.5      -1      -0.5      0
                           % Change in PEF
            0.5      1
                                                                                        1.5
       Figure 7-17.  Comparison of single-day lags (1-, 2-, 3-, 4-, 5-, and 6-day) to a cumulative
                     multiday lag (1- to 5-day) for % changes in PEF per 30 ppb increase in 8-h
                     avg O3 in urban children.
       Source: Derived from Mortimer et al. (2002).
 1     the results from multiday lags may not be as straightforward as that from single-day lag
 2     analyses. Few field studies examined or presented exposure lags of more than 3 days.  In a study
 3     of asthma symptoms, Delfino et al. (1998) stated that no long-term lag effects were seen, but did
 4     not provide the lags examined.
 5          Bias resulting from the selection of lags has not been examined specifically for O3 effects.
 6     However, the issue of lags has been investigated for PM and the results of this analysis are most
 7     likely of relevance for O3.  Lumley and Sheppard (2000) performed a simulation study to
 8     examine model selection bias in air pollution epidemiology. Sheppard et al. (1999; reanalysis
 9     Sheppard, 2003) had investigated the association between asthma hospital admissions and
10     ambient PM25 concentrations over a  eight-year period in Seattle, WA. Note that the results from
11     Lumley and Sheppard (2000) and Sheppard et al. (1999) were based on GAM using default
12     convergence criteria. A negative control analysis, using simulated data with no association
13     between PM exposure and the health outcome, and a positive control analysis, in which a
14     specified non-zero excess risk is added to the simulation, were performed for comparison.  The
15     bias from selection of best of seven lags (0 to 6 days) and residual seasonal confounding in the
16     negative control  analysis (median log relative risk of 0.0013) was approximately half the log
       January 2005
                              7-98
DRAFT-DO NOT QUOTE OR CITE

-------
 1      relative risk estimated from the observed data (0.0027), after adjusting for season and
 2      temperature. In the positive control model (true log relative risk of 0.0083), the bias was small
 3      (median log relative risk of 0.0080).  Results from these simulations indicate that bias from
 4      selection of lags may  be negligible when the true association is moderately large. However, if
 5      the relative risk is relatively small, as in the case of air pollution epidemiology, bias may be of
 6      issue. Selection of the largest risk estimate from a series of lags potentially can lead to positive
 7      bias towards finding a significant association.
 8           Selection of lag periods should depend on the hypothesis of the study and the potential
 9      mechanism of the effect. Bias can result from the reporting of only the largest and most
10      significant risk estimate, as well as the reporting of single-day lag results when significant
11      relationship are found on multiple lag days. Most studies have shown that O3 has a fairly
12      consistent, immediate effect on respiratory health with the majority finding significant
13      relationships at 0- and 1-day lags, especially for acute mortality and hospitalization outcomes.
14      Some studies indicated a greater cumulative O3 effect observed over longer lag periods,
15      suggesting that in addition to single-day lags, multiday lags should be investigated to fully
16      capture a delayed O3 effect on health outcomes. The issue of lags warrants further investigation.
17
18      7.6.5   Confounding by Temporal Trends and Meteorologic Effects
19           The challenge of analyzing acute O3 effects in time-series studies is to avoid bias due to
20      confounding by daily to seasonal temporal factors.  On a seasonal scale, the analysis must
21      remove the influence  of the strong seasonal cycles that usually exist in both health outcomes and
22      O3.  On a daily scale,  weather factors and other air pollutants also may confound the  association
23      of interest. This section discusses the interpretation of effect estimates  after adjusting for
24      temporal trends and meteorologic effects.
25
26      7.6.5.1   Assessment of O3 Effects after Adjusting for Temporal Trends and
27              Meteorologic Effects
28           The relationship between  O3 and health outcomes are significantly affected by temporal
29      trends and meterological factors, namely temperature.  Analyses of the  association between
30      health outcomes and O3 concentrations using raw data, therefore, can be misleading.  In an
31      analysis of Madrid, Spain  data by Diaz  et al. (1999), a U-shaped relationship was observed
32      between mortality and O3 concentrations, with an associated minimum  at 35 |ig/m3

        January 2005                              7-99        DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
(approximately 18 ppb) for 24-h avg O3.  The negative portion of the slope is likely due to the
opposing seasonal cycles in mortality (high in winter) and temperature (low in winter).
However, little is discussed about the interpretation of the fitted morbidity and mortality
"effects" of temperature.  See, for example, Figure 7-18, which shows the fitted nonaccidental
mortality as a function of natural  spline smoothing of mean temperature in Montreal, Quebec
(Goldberg and Burnett, 2003).  The positive slope of the temperature-mortality relationship is
fitted most tightly in the mild temperature range in which we do not expect mortality effects of
temperature. It is possible that temperature has mortality effects in the mild temperature range,
however because daily fluctuations of air pollution, especially O3, are strongly influenced by
weather conditions,  ascribing the association between temperature and mortality entirely to
temperature effects may underestimate the effects of air pollution.
                          0.20-
                       o
                       i  0.10
               o
               v>
               o
               o.
               If)
                          0.05-
                          0.00-
                          -0.05 -
                             -30      -20      -10       0       10       20
                                             Mean Temperature (°C)
                                                                        30
       Figure 7-18.   Daily nonaccidental mortality in Montreal, Canada as a function of mean
                      temperature, using natural splines with two degrees of freedom.
       Source: Goldberg and Burnett (2003).
       January 2005
                                          7-100
DRAFT-DO NOT QUOTE OR CITE

-------
 1           A 2003 HEI report investigated the impact of the selection of GAM convergence criteria to
 2      adjust for temporal trends and weather variables in PM time-series studies (HEI, 2003). These
 3      sensitivity analyses included the use of varying degrees of freedom for smoothing terms to adjust
 4      for temporal trends in the Poisson regression model.  Sensitivity analyses specifically for O3
 5      effects have not been performed, with the exception of one new study. In the U.S. 95
 6      communities data, Bell et al. (2004) performed a sensitivity analysis of the O3 excess mortality
 7      risk estimates to tripling the degrees of freedom for smoothing terms used to adjust for temporal
 8      trends. They found that varying the degrees of freedom from 7 to 21 per year did not
 9      significantly affect the O3-mortality estimates, with effect estimates ranging from 0.82 to 1.08%
10      excess risk per 20 ppb increase in 24-h avg O3 during the previous week.  Using more degrees of
11      freedom in temporal trend fitting (i.e., controlling shorter temporal fluctuations) means ascribing
12      more details of daily health outcomes to unmeasured potential confounders and possibly taking
13      away real weather and air pollution effects. However, results from this large multicity study
14      indicated that O3 effects were robust to aggressive smoothing of temporary trends.
15           Temporal cycles in daily hospital admissions or emergency department visits are often
16      considerably more episodic and variable than is usually the case for daily mortality.  As a result,
17      smoothing functions that have been developed and tuned for analyses of daily mortality data
18      may not work as well at removing cyclic patterns from morbidity counts.  Two methods are
19      commonly used for season adjustment, and an important distinction exists in the manner in
20      which these adjustments are applied in the analysis. The pre-adjustment method involves
21      applying the adjustment to both outcome and air pollution variables prior to the regression
22      analysis.  In this case, the regression is done on the residuals following subtraction of smooth
23      functions for each variable. The second method, co-adjustment involves applying the
24      adjustment as part of the regression analysis, by fitting a function of time while simultaneously
25      fitting the regression effect of air pollution and weather factors. These two approaches have
26      been viewed as largely interchangeable.  However, the co-adjustment approach may  lead to
27      biased air pollution effect estimates in cases where both outcome and pollution variables exhibit
28      strong seasonal cycles. This was demonstrated using a 15-year time-series data of daily hospital
29      admissions for acute respiratory diseases in children under 2 years old (Burnett et al., 2001).
30      Note that this  analysis used Poisson GAM with default convergence criteria.  Pre-adjustment
31      followed by regression analysis yielded a statistically significant estimate of 14.1% increase in

        January 2005                              7-101       DRAFT-DO NOT QUOTE OR CITE

-------
 1      admissions per 40 ppb increase in 1-h max O3 using year-round data. However, when the co-
 2      adjustment method was applied, there was a statistically significant 7.2% decrease in
 3      admissions. The authors suggested that the co-adjustment method allows O3 to compete with the
 4      smoothing variable to explain some of the seasonal variability in the outcome, whereas pre-
 5      adjustment eliminates the seasonal variability prior to analysis of O3 effects. Interestingly, when
 6      the authors limited the analysis to the warm season (May-August), both methods yielded similar
 7      results (32.3% versus 30.0% increase for co-adjustment and pre-adjustment, respectively)
 8      implying that  stratification by season can remove a significant amount of the confounding
 9      seasonality. This finding may be important to consider in reviewing the acute O3 mortality and
10      morbidity literature since the vast majority of studies published over the past decade have used
11      the co-adjustment method.  However, the use of pre-adjustment versus co-adjustment in time-
12      series studies is an unresolved issue.  More empirical research in different locales is needed to
13      evaluate the merits of these two methods as far as O3 is concerned, and to determine what
14      endpoints may be affected.
15           An interesting study by Schwartz (2004) examined the sensitivity of the  O3-mortality
16      relationship to methods used to control for temperature.  Using a case-crossover analysis, the
17      effect of O3 on mortality was examined in 14 cities across the U.S. from 1986  to 1993.  Control
18      days for an event were selected to be all  other days from the same month of the same year.
19      Initially, temperature lagged 0 and 1 day was controlled using nonlinear regression splines with
20      3 degrees of freedom each.  In a comparison  analysis, control days were restricted to a subset
21      that was matched on temperature. The effect estimate for all year data was a 0.8% excess risk
22      per 40 ppb increase in 1-h max O3 in the analysis using nonlinear regression splines, compared
23      to a 0.9% excess risk using temperature matched controls. The  effect estimates from the two
24      analyses were not significantly different.  Results were similar when restricting analysis to warm
25      season only data.
26           More sensitivity analysis of O3 effect estimates to the extent of adjustment for temporal
27      trends and meteorological factors is needed, but perhaps it is equally as important to evaluate the
28      epidemiological adequacy of a given adjustment.  For example, do the fitted mortality series
29      sufficiently depict influenza epidemics?  Or,  when larger degrees of freedom (e.g., 12 degrees of
30      freedom per year) are used, what "unmeasured" confounders, other than weather and pollution,
31      are the investigators trying to adjust? Even in PM studies that conducted sensitivity analyses,

        January 2005                              7-102        DRAFT-DO NOT QUOTE OR CITE

-------
 1      investigators rarely stated assumptions clearly, and not enough discussions were provided as to
 2      potential reasons for the sensitivity of results.
 3           Given their relationship to health outcomes and O3 exposure, adjusting for temporal trends
 4      and meteorologic factors is critical to obtain meaningful O3 effect estimates. While the
 5      prevailing analytical approaches fit the data flexibly, the estimated effects of meteorologic
 6      variables and their impact on the adjusted O3 effects are not adequately discussed. More work is
 7      needed in this area to reduce the uncertainty involved in the epidemiologic interpretation of O3
 8      effect estimates.
 9
10      7.6.5.2  Importance of Season-Specific Estimates of O3 Health Effects
11           Analysis of O3 health effects is further complicated as the relationship of O3 with
12      temperature and with other pollutants appears to change across seasons. Such relationships can
13      be observed in Figure 7-19 from a study by Moolgavkar et al. (1995).  In this study, Moolgavkar
14      et al. examined the relationship between daily mortality and air pollution (TSP, SO2, and O3) by
15      season in Philadelphia, PA for the period of 1973 to 1988.  During the summer, there was a
16      positive relationship between O3 and TSP, as well as O3 and SO2. In contrast, the relationship
17      between O3 and TSP, and O3 and SO2 inversed during the winter. Note that a greater range of O3
18      concentrations was observed during the summer. The analyses indicated that while both TSP
19      and SO2 showed positive and significant associations with mortality in all four seasons in single-
20      pollutant models, O3 showed positive and significant associations only in the summer when the
21      mean O3 concentration was the highest (Figure 7-20). The O3-mortality association was negative
22      (though not significantly) in the winter when the mean O3 concentration was low. The addition
23      of TSP or SO2 in the regression did not attenuate the O3 effect estimate in the summer,  and in the
24      three-pollutant model in the summer, only O3 remained significant.  In another Philadelphia
25      study by Moolgavkar and Luebeck (1996), analyzed using GAM with default convergence
26      criteria but with only one nonparametric smoothing term, O3 was also positively and
27      significantly associated with mortality in the summer.  A negative association between O3 and
28      mortality was observed during the winter in the single-pollutant model.  With all pollutants
29      (TSP, SO2, NO2, and O3) included in the model, the O3 effect in the summer remained
30      significant. Both studies did not analyze the data for the year-round data set, therefore the
31      relationship between the  excess risk estimates for each season and the year-round data  could not

        January 2005                             7-103        DRAFT-DO NOT QUOTE OR CITE

-------
                                           Summer
            100-
        ^   80-
         Q.   60-
         Q.
        8   40^
             20-
              0-
                0
             30-
n  20-|
Q.
a.
S  1H
              o-
50     100
   TSP
                            150
              50     100     150
                 TSP (|jg/m3)
                                            loo-
                                             se-
                                         s'
                                          Q.  60-
                                          Q.
                                          cp  40 -
                                         O
                                             20-
                                              0-
200
                                            Winter
                                                  Q.
                                                  a.
                                           200
                                             30-
                                                      20-
                                                  CD   10^
                                              0-
0    20    40    60   80  100
           S02 (ppb)
              0    20   40   60   80   100
                         SO2 (ppb)
      Figure 7-19.  Scatterplots of daily levels of O3 with TSP and SO2 in Philadelphia, PA
                   by season.
      Source: Derived from Moolgavkar et al. (1995).
1     be compared. The results from these studies, however, suggest that year-round analyses may
2     mask the positive (or negative) associations that may exist in particular seasons.
3          In the analyses of the U.S. 90 cities data by Samet et al. (2000; reanalysis Dominici et al.,
4     2003), the focus of the study was PM10, but O3 and other gaseous pollutants also were analyzed
5     in single- and multiple-pollutant models. In the reanalysis (Dominici et al., 2003), the combined
      January 2005
                                     7-104
               DRAFT-DO NOT QUOTE OR CITE

-------
1.4-
*
j£ 1.2 -
_>
"•5
— 1 0 -
0.8-

+S
r7 fft
ii ii
•
Spring Summer






+
+T+S

I
{

i <

1
(

)
i
+
T


3


Fall Winter
26 ppb 36 ppb 16
ppb 12 ppb
       Figure 7-20.  Estimated total (nonaccidental) mortality relative risk per 100 ppb increase
                     in 24-h avg O3 of reach seasonal data set. Within each season, the left-hand
                     estimate is for O3 alone; the estimate in the middle (+T) is with TSP, the
                     right-hand estimate (+S) is with SO2 in the model.  Seasonal mean O3
                     concentrations are noted.
 1     O3-mortality estimate for all seasons, summer only, and winter only analyses were all
 2     statistically significant at a lag of 1 day (see Figure 7-21).  However, while an excess risk in
 3     mortality was observed for all seasons and summer only analyses, a negative estimate was
 4     obtained for the winter only analysis. It should be noted that, Samet et al. and Dominici et al.'s
 5     analyses used a weather model specification that is more detailed than other studies in that it had
 6     multiple terms for temperature and dewpoint (these two variables are generally highly
 7     correlated). Thus, it is possible that the high concurvity of O3 with these weather covariates may
 8     have produced these conflicting results. Another possibility is that, as mentioned previously, the
 9     negative correlation between O3 and PM and other primary pollutants may have produced the
10     apparent negative relationship between O3 and mortality in the winter (note that PM and
11     mortality were positively associated).  In the similar U.S. 95 communities study by Bell et al.
12     (2004), analyses with only winter data were not performed, however, both the all year and
13     summer only analyses indicated statistically significant positive risk estimates (1.04% and
14     0.78% excess risk per 20 ppb increase in 24-h avg O3, respectively, using a constrained
15     distributed 7-day lag model).
16          Anderson et al. (1996) examined the relationship between air pollution (O3, NO2, BS, and
17     SO2) and daily mortality (all cause, cardiovascular, and respiratory) in London, England for the
       January 2005
7-105
DRAFT-DO NOT QUOTE OR CITE

-------
                          0)
                          o
                          '•p
                          .Q
                         Q
                          O
                         £
                          in
                          o
                         o.
	All year
	Summer
	 Winter
                             -1.5       -1.0      -0.5      0.0      0.5      1.0
                                           % Change in Mortality
       Figure 1-21.  Marginal posterior distributions of the national average estimates of O3
                     effects on total mortality per 10 ppb increase in 24-h avg O3 at a 1-day lag
                     for all year, summer (June-August), and winter (December-February)
                     analyses in 90 U.S. cities.
       Source: Derived from Dominici et al. (2003).
 1      study period of 1987 to 1992 using a Poisson GLM model.  They examined the associations
 2      using data from all year, as well as the warm season (April-September) and the cool season
 3      (October-March) separately. Their results indicated that the estimated O3 relative risks were
 4      larger in the warm season than in the cool season for all cause mortality.  The percent excess risk
 5      estimated per 30 ppb increase in 8-h avg O3 (9 a.m.-5 p.m.) was 3.05% (95% CI: 1.39, 4.7),
 6      4.37% (95% CI: 2.17, 6.62), and 0.96% (95% CI: -1.10, 3.06) for all year, warm season, and
 7      cool season, respectively.  A similar pattern was seen for cardiovascular mortality, but the
 8      estimated risk was negative (not significantly) for the cool season. For respiratory mortality, the
 9      estimated excess risks were similar between the cool and warm seasons.  Many other studies also
10      reported larger excess mortality risks in the warm (or summer) season than in the cool (or
11      winter) season (see Figure 7-12 in Section 7.4.4). These studies showed  cool season risk
12      estimates that were either smaller compared to warm season estimates or slightly negative (but
13      not significant).  Of the studies that analyzed data by season, only one study in Pittsburgh, PA
14      (Chock et al., 2000) showed negative risk estimates in the summer.
       January 2005
                  7-106
DRAFT-DO NOT QUOTE OR CITE

-------
 1           The studies that observed larger (positive) associations between O3 and mortality in warm
 2      seasons are consistent with the expectation that O3, if harmful, should have a stronger association
 3      with health outcomes in the summer when concentrations are higher.  However, the negative
 4      O3-mortality associations seen in the winter in the U.S. 90 cities study (Samet et al., 2000;
 5      reanalysis Dominici et al., 2003) and Philadelphia, PA data (Moolgavkar et al., 1995) suggest
 6      that further examination of this issue is required. Specifically, if the O3 level in the winter is
 7      shown to be negatively associated with factors (e.g., PM) that are positively associated with
 8      mortality, then these potentially spurious negative O3-mortality associations can be explained.
 9      Several examples of this phenomenon  also exist in morbidity studies investigating the effect of
10      O3 on daily hospital admissions and emergency department visits (Anderson et al., 1998; Burnett
11      et al., 2001; Prescott et al., 1998; Thompson et al., 2001). A study by Thompson and colleagues
12      (2001) in Belfast, Northern Ireland observed a significant decrease in emergency department
13      visits for childhood asthma in the cold season (November-April), but not in the warm season.
14      Ozone concentrations were found to be inversely related to benzene levels (r = -0.65).  After
15      adjusting for benzene levels, there was no significant association between O3 and asthma
16      emergency department visits.
17           Unlike the time-series studies examining outcomes of mortality, hospital admissions, and
18      emergency department visits, most acute field studies did not perform year-round analyses.
19      These acute field studies that examined the relationship between O3 and lung function,
20      respiratory symptoms, and inflammation focused primarily on the O3 effect during the warm
21      season, when O3 levels were expected to be high.
22           The potential influence of season on O3 effect estimates was examined using summary
23      density curves. The O3 effect observed in all year data was compared to effects from warm
24      season and cool season only data (Figures 7-22 and 7-23). Summary probability density curves
25      (or summary density curves) were calculated to review the effect estimates from the various
26      studies.  To calculate the summary density curve, the normal distribution function first was
27      determined for each effect estimate and corresponding standard error. Then the individual
28      normal distribution functions were summed together to obtain the pooled normal distribution
29      function. The summary density curve  is  calculated by taking the derivative of the pooled normal
30      distribution function.  Unlike a single normal density curve, the summary density curve is
31      distribution-free and may be multimodal.  The summary density curves shown in Figures  7-22

        January 2005                              7-107       DRAFT-DO NOT QUOTE OR CITE

-------
                0 bi
                   -12   -10   -8-6-4-20     2     4     6
                                            % Change in Mortality
                             8     10    12
                                                   All year
              Warm
              season
             Cool season
                  % area under the density curve and > 0      84%        89%         73%
                  Mean (SD) effect estimates            1.7% (4.4%)  2.4% (4.3%)   -0.0% (4.4%)
                  Mode effect estimates                   1.3%        1.6%         0.2%
       Figure 7-22.  Summary density curves of the % change in all cause mortality for all year
                     data and by season per 40 ppb increase in 1-h max O3 or equivalent.  Effect
                     estimates from 14 studies have been included in the summary density curves
                     (see Figure 7-12 in Section 7.4.4 for the effect estimates).
 1     and 7-23 were smoothed by adding a constant to the standard error of each effect estimate in the
 2     calculation of the individual distribution functions.  The constant is a default for normal
 3     distribution densities and is larger when the number of effect estimates is smaller, as presented
 4     by Silverman (1986).  Since the normal distribution is unimodal, this constant will oversmooth
 5     when the density is multimodal.  In Figure 7-22, the summary density curves representing the %
 6     all cause (nonaccidental) mortality associated with O3 concentrations are presented (see Figure
 7     7-12 in Section 7.4.4 for the effect estimates). The summary density curves were calculated
 8     using results from 14 studies that reported at least two of the three estimates. This figure
 9     indicates that 84% of the area under the density curve has a value greater than zero for all year
10     data compared to 89% for warm season data and 73% for cool season data.  Therefore, both all
11     year and warm season data generally indicates a significant, positive O3 effect.  The mean effect
       January 2005
7-108
DRAFT-DO NOT QUOTE OR CITE

-------
              0.08

              0.06

              0.04

              0.02

              0.00
                   -15
       0                15               30
% Change in Respiratory Hospital Admissions
                            45
                                                   All year
                              Warm season
             Cool season
                    % area under the density curve and > 0
                    Mean (SD) effect estimates
                    Mode effect estimates
                     84%         88%          53%
                  7.7% (8.8%)   9.4% (11.3%)    1.6% (7.3%)
                     1.9%         4.3%         -0.8%
       Figure 7-23.  Summary density curves of the % change in total respiratory hospital
                     admissions for all year data and by season per 40 ppb increase in 1-h max
                     O3 or equivalent.  Effect estimates from six studies have been included in
                     the summary density curves (see Figure 7-8 in Section 7.3.3 for the effect
                     estimates).
 1     estimate is a 1.7% excess risk in mortality per 40 ppb increase in 1-h max O3 using all year data,
 2     compared to a slightly larger 2.4% excess risk using warm season data. The cool season only
 3     data indicates that there is no excess risk associated with O3 concentrations.
 4           Similar observations are made when examining the O3 effect on total respiratory hospital
 5     admissions (Figure 7-23).  Six studies provided season-specific estimates as well as all year
 6     results (see Figure 7-8 in Section 7.3.3 for the effect estimates).  Once again, a large % of the
 7     area under the summary density curve is greater than zero when using all year and warm season
 8     data, 84% and 88%, respectively, compared to cool season data, 53%.  The mean O3 effect
 9     estimate also is slightly larger for warm season data only, 9.4% excess risk per 40 ppb increase
10     in 1-h max O3, compared to all year analyses, 7.7% excess risk.  A small O3 effect (1.6% excess
11     risk) is observed when using cool season data only.
       January 2005
                  7-109
DRAFT-DO NOT QUOTE OR CITE

-------
 1           Integrating seasonal influences across the various health outcomes supports the view that
 2      O3 effects are different in the cool and warm seasons, with greater effects observed during the
 3      warm season. As this relates to higher O3 levels produced during the warm season, the larger
 4      effects are an appropriate conclusion. Therefore, these results indicate that warm season data
 5      should be used to derive quantitative relationships for the effect of O3 on health outcomes.  This
 6      conclusion is supported by epidemiologic researchers who focus on warm season data as an
 7      a priori design for the studies. The results also support a rationale to view the cool season as
 8      inappropriate to derive information from in regards to the level of effect.  However, studying
 9      summer data only when all year data is available weakens the power of the study since less data
10      is analyzed.  In addition, increased adverse health outcomes are observed in the winter, some of
11      which may be attributable to O3. The O3 effect in the winter time may be masked by the effects
12      of PM due to the negative correlation between these variables (see Section 7.6.6.2 for further
13      discussion).  Therefore, analysis of all year data may be improved by adjusting for PM indices in
14      addition to adequate adjustment of meteorological factors and temporal trends. The methods of
15      statistical analysis, in addition to various other factors, need to be considered in the study design
16      stage to choose the proper time period to examine the health effects of O3. The data presented
17      here may aid in making this choice.
18           Seasonality influences the relationship between O3 and health  outcomes as it may serve as
19      an indicator for changing meteorologic factors, namely temperature, and copollutant
20      concentrations.  Given the potentially significant effect of season, O3 effect estimates computed
21      for year-round data need to be interpreted with caution. Small or no effects may simply reflect
22      the cancellation of positive associations in the summer and negative associations in the winter, or
23      the presence of confounding due to the strong seasonal character of O3 concentrations.
24
25      7.6.6   Assessment of Confounding by Copollutants
26           Potential confounding by daily variations in copollutants is another analytical issue to be
27      considered.  With respect to copollutants, daily variations in O3 tend to not correlate highly with
28      most other criteria pollutants (e.g., CO, NO2, SO2, PM10), but may be more correlated with
29      secondary fine PM (e.g., PM2 5, sulfates) measured during the summer months. Assessing the
30      independent health effects of two pollutants that are somewhat correlated over time is
31      problematic.  If high correlations between O3 and PM or other gaseous pollutants exist in a given

        January 2005                              7-110        DRAFT-DO NOT QUOTE OR CITE

-------
 1      area, then disentangling their relative individual partial contributions to observed health effects
 2      associations becomes very difficult. The changing relationship between O3 and other
 3      copollutants also is of issue. In some urban locations, the correlation between PM indices and
 4      O3 is positive in the summer and negative in the winter.  This section will further discuss the
 5      correlation between O3 and copollutants and confounding of the O3 effect by copollutants.
 6
 7      7.6.6.1  Relationship between Personal Exposure to O3 and Copollutants
 8           To be confounders of the association between O3 and adverse health effects, copollutants
 9      must be associated with both O3 exposure and the health outcome (Rothman and Greenland,
10      1998, p. 121). Many studies have shown that copollutants of O3, namely PM, NO2, SO2, and
11      CO, are associated with respiratory and, in some cases, cardiovascular health outcomes.
12      In addition, ambient levels of these copollutants, measured at central monitoring sites, have been
13      found to be highly correlated to ambient O3 concentrations.  However, few studies have
14      examined the association between personal O3 concentrations and personal exposures to other
15      copollutants. In a scripted exposure study discussed earlier, Chang et al. (2000) examined the
16      relationship between 1-hour personal O3 and personal PM25 levels in several microenvironments,
17      including indoors,  outdoors, and in vehicles.  Chang et al. (2000) did not find a significant
18      correlation between personal O3 and PM25 concentrations in any of the microenvironments, even
19      after stratifying the data by season. In a Baltimore, MD study of susceptible populations (older
20      adults, individuals with COPD, and children), Sarnat et al. (2001) found that ambient 24-h avg
21      O3 concentrations and ambient 24-h avg PM25 levels were positively correlated (r = 0.67,
22      p < 0.05) in the summer and negatively correlated (r = -0.72, p < 0.05) in the winter. However,
23      no relationship was found between 24-h avg personal  O3 and personal PM2 5 concentrations.
24      Interestingly, a significant correlation also was  observed between ambient O3 and personal
25      PM2 5, with a mixed regression effect estimate of P = 0.28 (t = 4.00) in the summer and P =
26      -0.29 (t = -4.68) in the winter. In contrast to the results from Sarnat et al. (2001), a study by
27      Delfino et al. (2004) did not find a significant correlation between ambient 8-h max O3 and
28      ambient 24-h avg PM2 5 (r = 0.24) concentrations during two warm sampling periods, August-
29      October and April-June, in Alpine, CA. Personal PM measurements were taken using a
30      nephelometer which responds mainly to PM in  the 0.1 to 10 |im range, with the highest response
31      in the fine PM range.  Ambient 8-h max O3 was not found to be correlated with personal 8-h max

        January 2005                              7-111       DRAFT-DO NOT QUOTE OR CITE

-------
 1      PM (r = 0.03) or personal 24-h avg PM (r = 0.01). Personal O3 exposure data was not available
 2      in this study.  Another study by Delfino et al. (1996) found that personal 12-h avg O3 levels were
 3      not associated with ambient PM2 5 levels (r = 0.03) in a study of asthmatics (aged 9-16 years old)
 4      in San Diego, C A.  These studies provide limited evidence for a lack of correlation between
 5      personal O3 levels and personal exposure to PM.
 6
 7      7.6.6.2  Assessment of Confounding Using Multipollutant Regression Models
 8           The multipollutant regression model often is used to determine whether the pollutant-
 9      specific effect is robust.  However, due to the multicolinearity among O3 and pollutants, and the
10      changing correlation by seasons, multipollutant regression models may not adjust for potential
11      confounding adequately, especially when using year-round data.  Results from the U.S. 90 cities
12      study (Samet et al., 2000; reanalysis Dominici et al., 2003), as depicted in Figure 7-24, indicated
13      that PM10 risk estimates were robust to including O3 and other gaseous pollutants in
14      multipollutant models. In a similar analysis, the effect of copollutants on O3-mortality risk
15      estimates also were investigated in this dataset. While the addition of PM10 in the model did not
16      substantially change the O3-mortality risk estimate, a slight decline in the O3 effect was observed
17      (Figure 7-25).  In the extended U.S. 95 communities study (Bell et al., 2004), the city-specific
18      O3-mortality effects were found to be robust to the adjustment for PM10, as indicated by the
19      nearly 1:1 ratio between  estimates with and without PM10 adjustment shown in Figure 7-26.
20      These results indicated that PM10 did not  confound the association between O3 and mortality in
21      this large study. Limited data was available to examine the potential confounding effect of PM2 5
22      on the O3-mortality relationship. A weighted second-stage linear regression indicated that there
23      was no association between long-term PM2 5 average and the community-specific O3-mortality
24      effect estimate.  Several other mortality and morbidity studies have investigated confounding of
25      O3 risk estimates using multipollutant models with year-round data, and most have reported that
26      O3 effects were robust to adjustment for copollutants (see Figures 7-9 and 7-13 in Sections 7.3.3
27      and 7.4.5, respectively).
28           Since the pollutant most correlated with O3 in the summer is sulfate (which is in the fine
29      particle size range), especially in the eastern U.S., the  potential confounder of main interest for
30      O3 is PM25 and sulfate in the summer. However, the results from two-pollutant regression
31      models with O3 and sulfate (or PM2 5) should be interpreted with caution because both of these

        January 2005                              7-112        DRAFT-DO NOT QUOTE OR CITE

-------
                              1.0-
                          t   o.sj
                          o
                          0
                          O)
                          c
                          o
                             -0.5-
                             -1.0-
                                            Lag1
                                  ABODE
                                           Models

Figure 7-24.  Posterior means and 95% posterior intervals of the national average estimate
             of PM10 effects on total mortality from non-external causes per 10 ug/m3
             increase in 24-h avg PM10 at a 1-day lag within sets of 90 U.S. cities with
             pollutant data available. Models A = PM10 only; B = PM10 + O3; C = PM10 +
             O3 + NO2; D = PM10 + O3 + SO2; E = PM10 + O3 + CO.
Source: Derived from Dominici et al. (2003).
1.0-
•4-1
.|S 0.5-
o
•- 0
U
0)
G)
C
re
g -0.5-
-1.0-
La
•
t .
4 «
i





gO
i
<
•



I


i
i i

j





La

i . -


i



31
i

> '



i




1 ,







i
•


i




•
t *
f





L<
1
> <


i

1

ig2
l

> 4




1

I
•
> <


1
1


                       ABCDEABCDEABCDE
                                         Models
Figure 7-25.  Posterior means and 95% posterior intervals of the national average estimate
             of O3 effects on total mortality from non-external causes per 10 ppb increase
             in 24-h avg O3 at 0-, 1-, and 2-day lags within sets of 90 U.S. cities with
             pollutant data available. Models A = O3 only; B = O3 + PM10; C = O3 + PM10
             + NO2; D = O3 +PM10 +SO2; E = O3 + PM10 + CO.

Source: Derived from Dominici at al. (2003).
January 2005
7-113
DRAFT-DO NOT QUOTE OR CITE

-------
                     O)
                     re
                     re
                     Q
 o
S
Q.
                     O
                     I
                         15 -
                         10 -
                          5 -
     0 -
                         -5 -
                                                          = National Average Effect
                               -5
                                          10
             15
                                         Without PM10 Adjustment
                                     (using only days with PM10 data)
       Figure 7-26.  Maximum likelihood estimates of O3-mortality for 95 U.S. communities,
                    determined using a constrained distributed lag model for lags 0 through
                    6 days.  Same dataset was used for O3 estimates with and without adjustment
                    for PM10.
       Source: Derived from Bell et al. (2004).
1     pollutants are formed under the same atmospheric condition and are both part of the "summer
2     haze" pollution mix.  A simple two-pollutant regression model does not address their possible
3     synergistic effects, and the high correlation between the two pollutants may lead to unstable and
4     possibly misleading results. In any case, most studies that analyzed O3 with PM indices did not
5     have PM2 5 data and very few examined sulfate data. The studies that did have PM2 5 data,
6     including Santa Clara County, CA (Fairley, 1999; reanalysis Fairley, 2003), Philadelphia, PA
7     (Lipfert et al., 2000a), and Detroit, MI (Lippmann et al., 2000; reanalysis Ito, 2003), examined
8     copollutant models for year-round data only, but O3 mortality risk estimates were not
      January 2005
                          7-114
DRAFT-DO NOT QUOTE OR CITE

-------
 1      substantially affected by the addition of PM25. A mortality study by Lipfert et al. (2000a) also
 2      found that O3 risk estimates were not affected by the addition of sulfate.  Amongst the morbidity
 3      studies, the two summertime studies in Toronto, Canada by Burnett et al. (1997b, 2001) found
 4      that the O3 effect was only slightly attenuated after including PM25 in the model.  In one of these
 5      studies (Burnett et al., 1997b), the effect of O3 also was adjusted for sulfate. With the addition of
 6      sulfate in the model, the risk estimate for O3 remained relatively stable, from an 11% excess risk
 7      to a 9% excess risk per 20 ppb increase in 12-h avg O3 at a 1-day lag.
 8           Other studies have estimated O3 health risks with copollutants in the model by season.
 9      Amongst the mortality studies (see Figure 7-14 in Section 7.4.5), the O3 risk estimates in the
10      warm season were mostly positive and significant, with the exception of the Pittsburgh, PA
11      analysis by Chock et al. (2000).  Adjusting for copollutants, in particular PM indices, did not
12      substantially change the O3-mortality effect estimates, with both slight reductions and increases
13      observed in the adjusted estimates.  In the analysis using cool season data only, the O3 effect
14      estimates were generally negative, but none were statistically significant.  In contrast to the
15      analyses of warm season data, the O3 risk estimates all increased slightly with the adjustment of
16      PM indices. The inverse relationship between O3 and PM during the cool season most likely
17      influenced the O3-mortality effect estimates in the single-pollutant models. Thus, the
18      confounding effect by PM indices appears to vary by season, most likely due to the changing
19      relationship between O3 and PM by season.  These results indicate that although PM does  not
20      seem to influence significantly the association between O3 and mortality during the warm
21      season, PM may be a confounder of the O3-mortality relationship in the cool season.
22           A study of respiratory hospitalizations in 16 Canadian cites by Burnett et al.  (1997a) also
23      stratified  O3 risk estimates by season. A preliminary analysis studying O3 effects found that a
24      positive association between O3 and respiratory hospitalizations was observed in the spring,
25      summer, and fall,  but not in the winter. In an analysis restricted to warmer months (April-
26      December), the pooled  O3 risk estimates for all cities were significant but attenuated with  the
27      addition of copollutants and dewpoint temperature into the model.  Of the 16 Canadian cities,
28      Montreal and Vancouver showed no association between O3 and respiratory hospitalizations
29      after adjusting for dewpoint temperature. After the exclusion of these two cities, the O3 risk
30      estimates were found to be robust with the addition of copollutants in the models.  Two
31      additional respiratory hospitalization studies in the metropolitan Toronto, Canada area by

        January 2005                              7-115       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Burnett et al. (1997b, 2001) also observed consistent O3 risk estimates with the inclusion of
 2      copollutants. The analyses in both studies were restricted to warm months (May-September).
 3           In field studies, power to assess independent O3 effects may be limited by small sample
 4      sizes and short follow-up times.  Among the field studies, the O3 effect also was found to be
 5      robust to the addition of copollutants in multipollutant models, with a few exceptions.
 6      For example, the effect of O3 on PEF was not robust to adjustments for PM2 5 and sulfate, in
 7      studies by Romieu et al. (1996) and Neas et al.  (1999).  In general, however, O3 effects on
 8      respiratory symptoms (Romieu et al., 1996), lung function parameters  (Brauer et al.,  1996, Gold
 9      et al., 1999), and asthma medication use (Gent  et al., 2003) were robust to inclusion of PM25.
10      Often, the effects for O3 were found to be stronger than those for PM.
11           Multipollutant regression analyses indicated that O3 risk estimates were not sensitive to the
12      inclusion of copollutants, including PM25 and sulfate, in both year-round and warm season data.
13      These results suggest that the effect of O3 on respiratory health outcomes appears to be robust
14      and independent of the effects of other copollutants.  However, there is concern as to whether
15      analysis of the O3 effect on health outcomes is confounded by PM indices in the cool season.
16      In addition, uncertainty remains as to the use of multipollutant regression models to assess the
17      independent health effects of pollutants that are correlated.
18
19      7.6.7  Issues  of Model Uncertainty and Multiple  Hypothesis Testing
20           Epidemiologic studies that investigated the association between O3 and various health
21      outcomes often found a significant effect. A major concern is whether these significant
22      associations are an artifact of model selection resulting from multiple hypothesis testing.
23      Testing multiple hypotheses may, at times, be appropriate. For example, developing several
24      hypotheses a priori allows researchers to explore more thoroughly potential mechanisms for an
25      O3-related health effect. Sensitivity analyses, which are critical for model validation, also use
26      multiple hypothesis testing. The basic issue with multiple hypothesis testing is that an extremely
27      large number of models are possible, any of which may turn out to give the best statistical "fit"
28      of a given set of data. Including all potentially  confounding variables  into the model is not
29      practical as this  may result in overfitting the model and inflated standard errors.  On the other
30      hand, selection of one "best" model ignores the uncertainty involved in model selection and
31      leads to an underestimation of the error.  Akaike Information Criterion and Bayes Information

        January 2005                              7-116       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Criterion are some of the statistical methods used to assist in model variable selection.  Recent
 2      attention has focused on Bayesian model averaging as an efficient method to incorporate model
 3      uncertainty into decision-making.
 4           A few authors have applied Bayesian model averaging to study the effect of air pollution
 5      on mortality. Clyde et al. (2000) and Clyde (2000) used Bayesian model averaging to analyze
 6      the relationship between mortality and PM concentrations from Phoenix, AZ and Birmingham,
 7      AL, respectively. In addition to the uncertainty of effect estimation, Bayesian model averaging
 8      incorporated uncertainty regarding the choice of confounding variables, pollutants, and lags.
 9      In the Phoenix, AZ study, Clyde et al. (2000) did not observe a PM2 5 effect on mortality, but did
10      find that coarse PM (PM10_2 5) was significantly associated with increased mortality. In a
11      reanalysis of the Birmingham, AL study (original analysis Schwartz,  1993), Clyde (2000)
12      observed that the PM10 effect originally estimated by Schwartz was plausible but Bayesian
13      model averaging results supported a smaller risk estimate. However, Clyde (2000) noted that
14      her analysis of the Birmingham data did not take into consideration factors that might bias the
15      estimated effect toward the null. For example, measurement error in  the exposure variables were
16      not considered.  In addition, the Poisson model (similar to many other regression models)
17      assumed that all individuals in a population had equal risks, including potentially susceptible
18      populations such as those with respiratory illnesses and outdoor workers.
19           Only one study using Bayesian model averaging reported a coefficient for O3-related
20      mortality.  Koop and Tole (2004) used Bayesian model averaging to analyze the effect of various
21      air pollutants, including O3,  SO2, CO, NO, NO2, PM10_2 5, and PM25, on mortality in Toronto,
22      Canada. Current values and up to 3-day lags were considered. In addition, a comprehensive set
23      of meteorological variables were included in the models. The 50+ explanatory variables
24      required the fitting of an enormous number  of potential models. Although the point estimates
25      for all pollutants were positive, very small effects were found. Sixty-six percent of the PM data
26      used to calculate these effect estimates were imputed. Ozone data was collected daily,
27      eliminating the need for imputation.  For O3, the cumulative effect on non-accidental deaths was
28      an excess of 0.054 deaths (posterior SD 0.159) per one standard deviation (9.15 ppb, 24-h avg
29      O3) increase in O3 levels. The most probable O3 model estimated the same-day effect of O3 on
30      mortality to be a statistically significant 0.526 (posterior SD 0.176) excess deaths. However, this
31      most probably model received only 0.23% of the probability. Koop and Tole concluded that the

        January 2005                             7-117       DRAFT-DO NOT QUOTE OR CITE

-------
 1      standard error of the cumulative effect was much too large to base policy advice.  However, in
 2      the context of the many interaction terms, meteorological variables, smoothing surfaces, and the
 3      relatively loose posterior distribution, it is likely that Koop and Tole have overestimated the
 4      variance of their pollution coefficients.
 5           Model diagnostics may be a way to reduce model uncertainty (George, 1999 in comments
 6      to Hoeting et al., 1999). However, Hoeting et al. state that model diagnostics is often based upon
 7      methods that use multiple testing. They believe that diagnostics should be used first to suggest
 8      better models and Bayesian model averaging should be used later to compare all models.  The
 9      models considered should have "appreciable likelihood" or be excluded from Bayesian model
10      averaging (George, 1999). A problem with Bayesian model  averaging occurs when variables  are
11      highly correlated. When this occurs, the estimated posterior effects may be diluted, resulting in
12      smaller coefficients (George, 1999). George believed that the issue of dilution could be
13      addressed by altering the prior probabilities used. In their reply to George, Hoeting et al. stated
14      that dilution is a problem when the highly correlated variables act by the same mechanism and
15      may serve as surrogates for the same variable, which may be the case for air pollutants.
16           While Bayesian model averaging can theoretically be used to take into account uncertainty,
17      claims of causality based on observational studies may be highly sensitive to the choice of prior
18      distributions and class of models under consideration (Clyde et al., 2000). Additional research in
19      this area may provide new and interesting insights into the issues of model uncertainty and
20      multiple hypothesis testing.
21
22      7.6.8   Concentration-Response Function and Threshold
23           An important consideration in determining whether a safe level of O3 can be identified is
24      whether the concentration-response relationship is linear across the full concentration range or
25      instead shows evidence of a threshold. Of particular interest is the shape of the concentration-
26      response curve in the vicinity of the current 8-h NAAQS for O3 of 80 ppb. The O3
27      concentration-response relationship has been explored in several studies.
28           To examine the shape of the concentration-response relationship between O3 and mortality,
29      Gryparis et al. (2004) used meta-smoothing to combine smooth curves across the 23 European
30      cities in a hierarchical model.  For the summer period, the estimated concentration-response
        January 2005                              7-118       DRAFT-DO NOT QUOTE OR CITE

-------
 1      curve did not appear to deviate significantly from linearity within the range of O3 concentrations
 2      commonly observed in European cities.
 3           In the U.S. 95 communities study (Bell et al., 2004), effect estimates calculated using only
 4      days with 24-h avg O3 levels less than 60 ppb were compared to those using all data.  At a lag of
 5      1 day, O3 was associated with an excess risk of 0.36% per 20 ppb increase in 24-h avg O3 using
 6      data from all days and only a slightly smaller risk of 0.30% when data was limited to days less
 7      than 60 ppb. These results suggest that if there is a threshold, it is present at 24-h avg O3 levels
 8      below 60 ppb.  Fairley (2003) reanalyzed the Santa Clara County mortality data using GAM
 9      with stringent convergence criteria and examined a new exposure index for O3.  He noted O3
10      concentrations exceeding 60 ppb each hour and calculated a daily sum of these exceedances.
11      Fairley's index incorporates measures of concentration and exposure duration. This type of
12      index is called a linear time-integrated  concentration, also known as dosage. The O3 index with
13      the 60 ppb threshold level was found to be significantly associated with mortality in single-
14      pollutant models as well as in multi-pollutant models.  Two other threshold levels were
15      examined, 40 ppb and 80 ppb. Both produced statistically significant results in single-pollutant
16      models. These results indicate that the threshold for O3-mortality effects is less than 40 ppb.
17      The implication for thresholds in terms of the three standard indices (i.e., 1-h max, 8-h max, and
18      24-h avg) is unclear, but there may be an empirical relationship.
19           Vedal et al. (2003) observed  that  the annual mean 1-h max O3 concentration of 27.3 ppb in
20      Vancouver, Canada, was lower than that in any of the 90 NMMAPS cities (Samet et al.,  2000),
21      thus a study in this city may be able to  better focus on the shape of the concentration-response
22      curve at lower levels.  In this Vancouver study, a statistically significant O3 effect was observed
23      on total mortality at a 0-day lag during the summer. Statistically significant effects on
24      respiratory mortality at a 2-day lag and marginally significant effects on cardiovascular mortality
25      at a 0-day lag also were observed for O3 in the summer. The O3 effect on mortality was found to
26      be robust in two-pollutant models.  Vedal et al. (2003) questioned if O3, other gaseous pollutants,
27      and PM were acting as surrogate markers of pollutant sources that contain more toxic
28      compounds, as the low measured concentrations were unlikely in their opinion to cause the
29      observed effects. They further stated that measurement error and interference by meteorological
30      factors might have contributed to the inability to detect a threshold.  Vedal et al.  (2003)
31      concluded that O3 concentrations were  associated with adverse effects on mortality even at low

        January 2005                             7-119       DRAFT-DO NOT QUOTE OR CITE

-------
 1      levels.  Although this study supports the argument that there is no threshold concentrations
 2      below which adverse effects cannot be detected, the results must be interpreted with caution as
 3      concerns remain.
 4           Kim et al. (2004) investigated the presence of a threshold in O3-mortality effects in Seoul,
 5      Korea by analyzing data using a log linear GAM (linear model), a cubic natural spline model
 6      (nonlinear model), and a B-mode splined model (threshold model).  Models were stratified by
 7      season and adjusted for PM10, long-term time trend, and meteorological variables. An estimated
 8      threshold value of 47 ppb was observed for 1-h max O3. None of the other pollutants examined,
 9      including PM10,  SO2, NO2, and CO, had a nonlinear association with mortality. Using summer
10      data only, the B-spline model resulted in an excess mortality risk of 7.1% (95% CI: 3.1, 11.2)
11      per 40 ppb increase in 1-h max O3, compared to an excess risk of 3.6% (95% CI, 0.5, 6.8)
12      calculated using the log linear model. If a threshold truly exists, results from the Kim et al.
13      (2004) study suggest that the use of log-linear models may underestimate the O3 effect on
14      mortality at levels above the threshold.
15           In London, England data (Anderson et al., 1996), an adjusted O3-mortality plot indicated a
16      possible threshold level around 50 ppb for 8-h avg O3.  A study by Simpson et al. (1997) in
17      Brisbane, Australia observed a significant excess risk in mortality only in the highest quintile of
18      O3 exposure, which had a mean concentration of 42 ppb for 1-h max O3.  One study by Lipfert
19      et al. (2000b) examined the presence of a threshold in the effect of chronic O3 exposure on
20      mortality in U.S. veterans. A simple concentration-response plot comparing the risk estimate in
21      the upper two tertiles to that from the lowest tertile seemed to indicate a threshold level of
22      approximately 140 ppb of 1-h max O3 during the period 1975  to 1981 for both concurrent
23      mortality (1975-1981) and delayed mortality (1982-1988).
24           Among several studies with morbidity outcomes, examination of the shape of the
25      concentration-response function indicated evidence of an effect threshold.  In a study of all-age
26      respiratory hospital admissions in Toronto, Canada, effects of O3 appeared to become apparent
27      only above approximately 30 ppb daily 1-h max O3 (Burnett et al., 1997b). In  London, England,
28      Ponce de Leon et al. (1996) observed an indication of a threshold in the O3 effect on
29      hospitalizations at 40 to 50 ppb for 8-h max O3 and 50 to 60 ppb for  1-h max O3. In a study of
30      emergency department visits for asthma in St. John, Canada, effects observed in the over 15
31      years age group were apparent only when data above the 95th percentile  (75 ppb daily 1-h max

        January 2005                             7-120       DRAFT-DO  NOT  QUOTE OR CITE

-------
 1      O3) were included (Stieb et al., 1996).  However, other morbidity studies observed a monotonic
 2      increase in the concentration-response function, suggesting that there was no threshold in O3
 3      effects on hospitalizations and emergency department visits (Burnett et al.,  1997a; Jaffe et al.,
 4      2003; Petroeschevsky et al., 2001; Tenias et al., 1998).
 5           In a field study by Mortimer et al. (2002), the association of ambient O3 levels with PEF
 6      and asthma symptoms was investigated in eight urban cities in the U.S.  The mean 8-h avg O3
 7      was 48 ppb, with less than 5% of days exceeding 80 ppb. Analysis performed using all data
 8      indicated that a 15 ppb change in 8-h avg O3 was associated with statistically significant
 9      decrements in PEF (-0.59% [95% CI:  -1.05, -0.13]) and increased incidence of respiratory
10      symptoms (odds ratio of 1.16 [95% CI: 1.02, 1.30]) over multiday lag periods. When data was
11      restricted to days when ambient O3 concentrations were less than 80 ppb, the O3 effects
12      persisted, with a significant PEF decline (-0.70% [95% CI: -1.29, -0.12])  and incidence of
13      morning symptoms (odds ratio of 1.17 [95% CI: 1.01, 1.35]).  A study by Chen et al. (1999) also
14      found that there was no clear threshold in the O3 effect on FEVj and FVC in Taiwanese school
15      children.
16           Note that adjusting for seasonal cycles does not address the issue of the changing
17      relationship between O3 concentrations and personal exposure across seasons. The ambient O3
18      levels are lower in the cold season, but people are likely to be exposed to even lower levels of
19      O3 in cold seasons due to the shorter time spent outdoors and the longer time spent indoors with
20      closed windows.  This is in contrast to what occurs with fine particles, which can effectively
21      penetrate the indoors. Thus, a more "accurate" concentration-response relationship may need to
22      be examined in a summer-only data set (which may suffer low data density in the low
23      concentration range). Even for summer data, however, an interpretation of the relationship is
24      not straightforward because of the possible influence of the use of air conditioning (an effective
25      remover of O3).  Greater use of air conditioning would be expected on hot days when the
26      O3 level is higher, but the use of air conditioning may also vary from city to city and across
27      social class within a city. These complications make it difficult to examine the existence of a
28      threshold of O3 health effects in the observational data.
29           Limited studies have examined the issue of thresholds in O3 health effects studies.  Some
30      studies have found a low level threshold while others have found no threshold in O3 effects.
31      An absence of a detectable threshold in population studies does not indicate an absence of

        January 2005                              7-121        DRAFT-DO NOT QUOTE OR CITE

-------
 1      individual thresholds. For a further discussion on thresholds in air pollutant health effects, see
 2      Section 8.4.7 in the 2004 PM AQCD. While no conclusion can be made regarding the threshold
 3      issue, the limited evidence shows that the possible threshold level may be well below the current
 4      standards. The distribution of thresholds, particularly around the NAAQS value of 80 ppb for
 5      8-h max O3, needs to be further investigated.
 6
 7      7.6.9   Spatial Variability in O3 Effect
 8           As described in Chapter 3 of this AQCD, O3 concentrations tend to be more spatially
 9      variable than PM2 5 concentrations  in urban areas. In addition, relative personal exposures to O3
10      likely vary by region. This spatial  variability in O3 concentrations and personal exposures may
11      contribute to the heterogeneity in observed O3 health effects. More than 80% of the O3-mortality
12      estimates from the various studies conducted in North America, South America, Europe, and
13      Australia were between 0 and 7% excess risk per 40 ppb increase in 1-h max O3 using year-
14      round data.  In general, the O3-mortality estimates were greater when using summer only data
15      compared to year-round data. Though not all statistically significant, most of the O3-mortality
16      estimates were greater than zero, indicating a positive relationship  between O3 exposure and
17      mortality. The O3 risk estimates from the numerous hospitalization and emergency department
18      visit studies were generally larger in magnitude and more variable  from study to study compared
19      to the mortality studies. These differences in the O3 effect estimates may be attributable to the
20      greater variability in the outcome measure, such as more subcategories of outcome and varying
21      degrees of severity, in hospitalization studies compared to mortality studies.
22           As differences in study design, population,  and data analysis  may affect risk estimates,
23      studies that were conducted in multiple cities using standardized methods were further examined
24      to investigate the spatial heterogeneity of O3 effects. Bell et al. (2004) conducted a time-series
25      analysis of O3 and mortality  in 95 U.S. communities from 1987 to 2000. A 10 ppb increase in
26      O3 in the previous week was associated with a statistically significant increase of 0.52% excess
27      risk of mortality in the pooled analysis of 95 communities.  Although some heterogeneity  was
28      observed among the communities (previously shown in Figure 7-11 of Section 7.4.3), the  range
29      of the community-specific effect estimates were fairly narrow. Of the 95 U.S. communities, 93
30      had positive O3-mortality risk estimates. Only  5 had risk estimates greater than 1% per 10 ppb
        January 2005                             7-122        DRAFT-DO NOT QUOTE OR CITE

-------
 1      increase in 24-h avg O3 during the previous week, with all communities indicating an excess
 2      mortality risk less than 2%.
 3           Greater heterogeneity was observed in the European study of 23 cities in 14 countries
 4      (Gryparis et al., 2004). In the year-round analyses, only 8 of the 23 cities had positive O3-
 5      mortality effect estimates. However, in the analyses using summer data only, the risk estimates
 6      were positive in 19  of the 23 cities, with a range of 0.8 to 8% excess risk per 40 ppb increase in
 7      1-h max O3.  The heterogeneity may be attributable to the considerable variability among
 8      countries in factors  that may influence the relationship between ambient O3 concentrations and
 9      personal exposure to O3,  such as climate, use of air conditioning, personal activity patterns, and
10      socioeconomic factors. In addition, the variability in the concentration and composition of co-
11      existing pollutants by cities or countries may contribute to the heterogeneity in the O3-mortality
12      effects. For example, concentrations of NO2 may vary widely by region, depending on the
13      differences in traffic density.
14           Among the hospitalization studies, Burnett et al. (1997a) conducted the largest multicity
15      study of 16 Canadian cities.  The mean daily 1-h max O3 was 31 ppb in the  16 cities.  The pooled
16      O3 estimate was 5.6% (95% CI: 3.4, 7.9) excess risk in respiratory hospitalization per 40 ppb
17      increase in 1-h max O3 using warm season data (April to December).  The risk estimates were
18      fairly homogenous across the 16 Canadian cities, ranging from 3.1% for Vancouver to 7.7% for
19      Quebec City.
20           Anderson et al. (1997) investigated the association between O3 and hospital admissions for
21      COPD in five European cities, London, Paris, Amsterdam, Rotterdam, and Barcelona. The
22      pooled  risk estimate was  5.0%, 4.7%, and 3.5% excess per 30 ppb increase  in 8-h max O3 for
23      year-round, warm season, and  cool season data, respectively.  Results from the APHEA study
24      showed similar variability to that from the Burnett et al. (1997a) study. The year-round excess
25      risk estimates were  lower in the two Dutch cities, 3.0%, compared to that in Paris, 9.8%.
26      In general, however, there was no significant evidence of heterogeneity in the O3 effects among
27      the five European cities.
28           Among the field studies,  various respiratory health outcomes were examined,  including
29      PEF, spirometric parameters, respiratory symptoms, and medication use.  Only one field study
30      investigated the O3 effect in several locations (Mortimer et al., 2002). Mortimer et al. (2002)
31      investigated the association between ambient O3 concentrations, and PEF and asthma symptoms

        January 2005                             7-123        DRAFT-DO NOT QUOTE OR CITE

-------
 1      in asthmatic children living in eight urban cities in the U.S. - St. Louis, MO; Chicago, IL;
 2      Detroit, MI; Cleveland, OH; Washington, DC; Baltimore, MD; East Harlem, NY; and Bronx,
 3      NY.  In the analysis pooling data from all eight cities, a 15 ppb increase in 8-h avg O3 was
 4      associated with a significant decrement of -0.59% in morning PEF for a 5-day cumulative lag
 5      period. The % changes in PEF were negative in all cities except for Baltimore, 0.24%.  Among
 6      the other seven cities, the % changes in PEF were quite homogenous, with values ranging from
 7      -0.54% for Washington, DC to -0.86% for St. Louis. A 15 ppb increase in 8-h avg O3 also was
 8      associated with an increased incidence of morning symptoms in the pooled analysis  (odds ratio
 9      of 1.16 for a 4-day cumulative lag period). In all cities except for St. Louis, there was an
10      increase in the incidence of morning symptoms.  The odds ratios for incidence of morning
11      symptoms varied more by city compared to the PEF measurements, ranging from 1.09 for
12      Chicago to 1.72 for Detroit.  The greater variance in incidence of symptoms may indicate the
13      lack of standardization in the use of symptoms as a health outcome measure.
14           Most of the multicity studies found consistent O3 effect estimates for mortality,
15      hospitalizations, and other respiratory health outcomes.  The slight heterogeneity of O3 effects
16      may be partially attributable to the use of centrally located ambient monitors to assess exposure.
17      There may be differences in relative personal exposures to O3 due to varying factors, namely use
18      of air conditioning and activity patterns, that affect the relationship between personal exposure
19      and ambient concentrations. The variability in the concentration and composition of
20      copollutants present also may contribute to the heterogeneity of the effect of O3 on health
21      outcomes as confounding by copollutants may vary by region.
22
23      7.6.10   Health Effects of O3 in Susceptible Populations
24           In this section, the effects of O3 on morbidity and mortality in potentially susceptible
25      populations will be examined.  In epidemiology studies of O3 health effects, the most widely
26      studied subpopulation was asthmatics. Also of interest were the observed health effects of O3  on
27      different age groups, particularly children and the elderly.  This section begins with  a discussion
28      of the O3-related health effects in asthmatics.
29
30
        January 2005                              7-124       DRAFT-DO NOT QUOTE OR CITE

-------
 1      7.6.10.1  Health Effects Associated with Ambient O3 Exposure in Asthmatics
 2           Epidemiological studies of health effects from acute O3 exposure in asthmatics have
 3      examined a range of outcomes: pulmonary function, respiratory symptoms, inflammation,
 4      emergency room visits, hospital admissions, and mortality.  Chronic O3 exposures have been
 5      associated with similar outcomes, with the exception of emergency room visits and
 6      hospitalizations. Both are discussed in the earlier text.  This subsection draws together this
 7      information to examine whether the evidence indicates that O3 exposure impacts asthmatics.
 8           In Germany and Mexico City, O3 exposure was associated with a decline in FEVj in
 9      asthmatic adults and children (Hoppe et al., 1995a; Romieu et al., 2002). Change in FEVj also
10      was examined in a group of asthmatic hikers in Mount Washington, NH (Korrick et al., 1998).
11      Compared to the healthy subjects, the asthmatic subjects experienced a four-fold greater decline
12      in FEVj with the same exposure to  O3. The results from the hiker study are consistent with those
13      observed in controlled human exposure studies (discussed in Chapter 6), which also indicate
14      significantly greater decrements in FEVj among mild asthmatics versus nonasthmatic subjects
15      with heavy intermittent exercise.
16           PEF was examined in panels of asthmatics in several field studies (see Figures 7-la and
17      7-lb). Collectively, all the  studies indicated decrements of morning peak flow but most of the
18      estimates were not statistically significant. One multicity study of eight urban areas in the U.S.
19      observed reductions in morning PEF that were not significant in each individual city (Mortimer
20      et al., 2002). However, the analysis combining data from all eight cities indicated a statistically
21      significant change with a cumulative lag of 1  to 5 days. Further analysis showed that the
22      incidence of > 10% decline in morning PEF was statistically significant, which was discussed by
23      the author as an  indication that O3 exposure may be associated with clinically significant changes
24      in PEF in asthmatic children. The study  examined 846 asthmatic children, the largest asthma
25      panel study reported.
26           Respiratory symptom increases in asthma panels were examined in several field studies,
27      some of which also examined PEF as discussed above.  The health indicators examined varied
28      among these studies and the analyses results were both negative and positive with a few being
29      statistically significant.  Collectively, they are suggestive of a potential effect on respiratory
30      symptoms but the evidence in the available studies is not strong.  Two U.S. studies examining
31      larger panels may be better  studies from which to draw inferences as the large sample size

        January 2005                              7-125       DRAFT-DO NOT QUOTE OR CITE

-------
 1      provides greater power to examine the effect of O3 on respiratory symptoms. The eight U.S.
 2      urban cities study mentioned above reported morning symptoms in the 846 asthmatic children to
 3      be most strongly associated with a 4-day cumulative lag period of O3 concentrations (Mortimer
 4      et al., 2002). A New England study examined 271 asthmatic children and observed a significant
 5      O3 effect on a variety of respiratory symptoms at a lag of 1 day among the 130 subjects who used
 6      maintenance asthma medications (Gent et al., 2003).
 7           Few epidemiological studies have examined airway inflammation in asthmatics.  A Mexico
 8      City study indicated that supplementation with antioxidants may modulate the impact of O3
 9      exposure on the small airways of children with moderate to severe asthma (Romieu et al., 2002).
10      A related study indicated that asthmatic children with GSTM1 null genotype were found to be
11      more susceptible to the impact of O3 exposure on small airways (Romieu et al, 2004).
12      An additional study in Mexico City examined DNA strand breaks in nasal epithelial cells in
13      asthmatic and nonasthmatics medical students and noted greater genotoxic damage in asthmatics
14      (Fortoul et al., 2003).
15           Emergency department visits for asthmatics have been examined in several studies and
16      range from negative to positive results with limited analyses providing significant results (see
17      Figure 7-6 in Section 7.3.2). Examination of the studies indicated that seasonal  summer studies
18      tended to yield positive outcomes, as expected based  on earlier discussions. Two studies in
19      Atlanta, GA (Tolbert et al., 2000)  and Valencia, Spain (Tenias et al., 1998) indicated significant,
20      positive effects in warm season analyses.  Further, a Canadian study, one of the  larger studies
21      conducted in the summer season, reported a large significant increase in asthma emergency
22      department visits when the daily 1-h max O3 concentration exceeded 75  ppb (Stieb et al., 1996).
23      A three-city study in Ohio also indicated a positive result during the summer (Jaffe et al., 2003).
24      Other studies of mostly year-long  data tended to produce nonsignificant results,  which  in some
25      cases were negative (Atkinson et al., 1999a; Castellsague et al., 1995; Thompson et al., 2001;
26      Tobias etal., 1999).
27           Hospital admission studies that specifically examined asthmatics were fewer in number
28      than those that examined total respiratory diseases. Significant effects were noted in all age
29      groups in studies conducted in Seattle, WA (Sheppard et al., 2003),  New Jersey  (Weisel et al.,
30      2002), Toronto, Canada (Burnett et al., 1999), London, England (Anderson et al., 1998),
31      Brisbane, Australia (Petroeschevsky et al., 2001), and Hong Kong (Wong et al.,  1999a).

        January 2005                             7-126        DRAFT-DO NOT QUOTE OR CITE

-------
 1      However, several other studies, mostly examining the effect on asthmatic children, did not
 2      observe a significant relationship (Gouveia and Fletcher, 2000a; Lin et al., 2003; Morgan et al.,
 3      1998; Nauenberg and Basu, 1999; Schouten et al., 1996).
 4          Acute mortality related to asthma was examined in Barcelona, Spain (Saez et al., 1999;
 5      Sunyer et al., 2002). Severe asthmatics with more than one asthma emergency visit showed the
 6      strongest mortality associations with air pollutants, NO2 being the most significant predictor
 7      followed by O3 (Sunyer et al., 2002).
 8          Recent reports from longitudinal cohort studies in California have reported associations
 9      between the onset of asthma and long-term O3 exposures (Greer et al., 1993; McConnell et al.,
10      2002; McDonnell et al., 1999).  Significant associations were seen in males but not females
11      (Greer et al., 1993; McDonnell et al.,  1999). In six high O3 communities, asthma risk was
12      elevated for children who played three or more sports as compared with children who played no
13      sports (McConnell et al., 2002). Playing sports may indicate outdoor activity and an increased
14      ventilation rate which may lead to increased exposure.  These outcomes would benefit from
15      replication in other cohorts in regards to indicating weight of a causal interpretation.
16          A few studies provide limited discussion of concentration-response functions and
17      thresholds.  In the eight urban areas U.S. study, the odds ratios for incidence of > 10% decline in
18      morning PEF and incidence of morning symptoms when excluding  days with 8-h avg O3 greater
19      than 80 ppb were nearly identical to those including data from all days (Mortimer et al., 2002)
20      In the New England asthma panel  study (Gent et al., 2003), some of the significant associations
21      for symptoms occurred at 1-h max O3 levels below 60 ppb. In the St. John, Canada study (Stieb
22      et al., 2003), a significant effect of O3 on emergency department visits was reported with
23      evidence of a threshold somewhere in the range below a 1 -h max O3 of 75 ppb in the 15 years
24      and over age group.
25          Overall, asthma subjects have been examined across most health endpoints of interest. The
26      results reported in these studies range from negative to positive estimates with some indicating a
27      significant positive excess risk associated with O3.  While no endpoint in itself seems to indicate
28      an unquestionable demonstration of an association, studies with adequate sample size and
29      understandable power consistently provide strong positive and significant results, especially
30      during the summer months when higher O3 levels occur. This view is strengthened as positive
31      results are obtained cohesively across the varied outcomes. Therefore, based on the evidence it

        January 2005                              7-127       DRAFT-DO NOT QUOTE OR CITE

-------
 1      seems prudent to consider asthmatics as a potentially susceptible group that requires protection
 2      from O3 exposures.
 3           A study by Niedell (2004) examined the relationship between air pollutants and asthma
 4      hospitalizations in California. The most recent EPA O3 report (U.S. Environmental Protection
 5      Agency, 2004b) indicated that O3 levels in the pacific southwest region had decreased by 9%
 6      from 1990 to 2003. This downward trend in O3 levels was mostly influenced by the
 7      improvements in Los Angeles and other southern California metropolitan areas.  As shown in
 8      Figure AX3-60 of the Chapter 3 Annex, O3 concentrations decreased by over 30% in Los
 9      Angeles from 1992 to 1998. Results from this study noted declines in levels of air pollutants
10      since 1992 and decreased asthma admissions in 1998 for children aged 1 to 18 years ranging
11      from 5 to 14%, depending on the age group. The greatest decline (> 10%) in air pollution-
12      related asthma admissions was observed among 3 to 12 year old children. Although this benefit
13      analysis was not specific to O3, it provides evidence of decreased morbidity resulting from
14      reduced air pollutant concentrations, including O3. Many studies have reported short-term
15      associations between O3 and morbidity outcomes, yet a largely unaddressed question remains as
16      to the extent to which reductions in ambient O3 actually lead to reductions in adverse health
17      outcomes attributable to O3. This question is not only important in terms of "accountability"
18      from the regulatory point of view, but it is also a scientific question that challenges the predictive
19      validity of statistical models and their underlying assumptions used thus far to estimate excess
20      health effects due to ambient O3.
21
22      7.6.10.2  Age-Related Differences in O3 Effects
23           Several mortality studies have investigated age-related differences in O3 effects.  Among
24      the studies that observed a significant association between O3 and mortality, a comparison of all
25      age or younger age (< 65 years of age) O3-mortality risk estimates to that of the elderly
26      population (> 65 years) indicates that, in general, the elderly population is more susceptible to
27      O3 effects (Borja-Aburto et al. 1997; Bremner et al., 1999; Gouveia and Fletcher 2000b;  O'Neill
28      et al., 2004; Simpson et al., 1997; Sartor et al., 1995; Sunyer et al., 2002). For example,  a study
29      by Gouveia and Fletcher (2000b) examined the O3-mortality effect by age in Sao Paulo, Brazil.
30      There were 151,756 deaths for all non-violent causes over the period of 1991 to 1993, of which
31      49% occurred in the elderly. Among all ages, O3 was associated with a non-significant 0.6%

        January 2005                             7-128       DRAFT-DO NOT QUOTE OR CITE

-------
 1      (95% CI: -0.8, 2.0) excess risk in all cause mortality per 40 ppb increase in 1-h max O3.
 2      In comparison, in the elderly population, the O3-mortality risk estimate was nearly three-fold
 3      greater, 1.7% (95% CI: 0.0, 3.3).  Similarly, a Mexico City study found that O3-mortality risk
 4      estimates were 1.3% and 2.8% per 20 ppb increase in 24-h avg O3 concentration in all ages and
 5      the elderly, respectively (O'Neill et al., 2004).
 6           The large U.S. 95 communities study (Bell et al., 2004) did not find evidence of significant
 7      heterogeneity in risk across three age groups, < 65 years, 65 to 74 years, and > 75 years of age.
 8      Effect estimates were only slightly higher for those 65 to 74 years, 1.40% excess risk per 20 ppb
 9      increase in 24-h avg O3, compared to individuals less than 65 years and 75  years or greater,
10      1.00% and 1.04%, respectively.  However, Bell et al. (2004) noted that despite similar effect
11      estimates, the absolute effect of O3 is substantially greater in the elderly population due to the
12      higher underlying mortality rates, which leads to a larger number of extra deaths for the elderly
13      compared to the general population.
14           Few mortality studies examined another potentially  susceptible age group,  young children
15      under the age of 5 years. The results were mixed, with one Mexico City study showing a lower
16      risk of O3-related all cause mortality in young children compared to  all ages and the elderly
17      (Borja-Aburto et al., 1997) and another study showing a greater risk in respiratory mortality in
18      young children compared to the elderly (Gouveia and Fletcher, 2000b).  It should be noted that
19      approximately 10% of mortality occurred in young children, thus the statistical power to study
20      the O3 effect in this age group was limited.
21           With respect to age-specificity of associations between O3 and acute respiratory
22      hospitalizations or emergency department visits, no clear pattern emerges from recent studies.
23      Significant associations have been reported for all ages (Anderson et al.,  1997; Burnett et al.,
24      1995, 1997b, 1999; Weisel et al., 2002), adults or elderly (Burnett et al., 1997a; Delfino et al.,
25      1997, 1998; Moolgavkar et al., 1997; Schwartz et al., 1996; Yang et al., 2003), and children
26      (Burnett et al., 2001; Gouveia and Fletcher, 2000a; Lin et al., 1999; Ponka  and Virtanen,  1996;
27      Tolbert et al., 2000; Yang et al., 2003).  Interestingly, studies that have examined effects in
28      multiple age strata often have seen effects only in non-pediatric strata (Delfino et al., 1997,
29      1998; Stieb et al., 1996; Jones et al., 1995). Several studies that focused on children did not
30      report significant O3 effects, though in some cases these studies are limited by small size,
31      inadequate control  of seasonal patterns,  or very low O3 levels (Lierl  and Hornung, 2003; Lin

        January 2005                             7-129       DRAFT-DO NOT QUOTE OR CITE

-------
 1      et al., 2003; Thompson et al., 2001). If O3 is causally related to exacerbations of respiratory
 2      diseases leading to hospital usage, one would expect to see effects most prominently among
 3      children, for whom asthma is most prevalent and exposures may be greater.
 4          Many of the field studies focused on the effect of O3 on the respiratory health of school
 5      children, however, none have compared the results from children to that in other age groups.
 6      In general, children experienced significant decrements in pulmonary function parameters,
 7      including PEF, FEVl9 and FVC (Castillejos et al., 1995; Chen et al., 1999; Gielen et al., 1997;
 8      Gold et al.,  1999; Jalaludin et al., 2000; Mortimer et al., 2002; Romieu et al., 1996; Thurston
 9      et al., 1997), and some experienced increases in respiratory symptoms (Delfmo et al., 2003;
10      Gold et al.,  1999; Neas et al., 1995; Romieu et al., 1996, 1997; Thurston et al., 1997) and asthma
11      medication use (Delfmo et al., 1996; Just et al., 2002; Ostro et al., 2001). These respiratory
12      heath effects were observed in both healthy and asthmatic children.
13          Collectively, there is supporting evidence of age-related differences in susceptibility to O3
14      health effects. The elderly population (> 65 years of age) appear to be at increased risk of
15      O3-related mortality and hospitalizations, and children (< 18 years of age) experience other
16      potentially adverse respiratory health outcomes with increased O3 exposure.
17
18      7.6.11   Summary of Key Findings and Conclusions Derived From O3
19               Epidemiologic Studies
20          In the previous 1996 O3 AQCD, there was considerable evidence of O3-related respiratory
21      health effects from individual-level camp and exercise studies, as well as some consistent
22      evidence from time-series studies of emergency room visits and hospitalizations.  Since the  1996
23      document, more field studies have been conducted, with some emphasis on additional outcome
24      markers such as  respiratory symptoms  and asthma medication use.  Another significant addition
25      to the current O3 AQCD is the substantial number of short-term O3 mortality studies, which is in
26      part due to the increase in the number of studies that examined PM-mortality associations.
27      Considering the wide variability in possible study designs and statistical model specification
28      choices, the reported O3 risk estimates for the various health outcomes are in reasonably good
29      agreement.  In the case of O3-mortality time-series studies, combinations of choices in model
30      specifications (the number of weather terms and degrees of freedom for smoothing of mortality-
31      temporal trends) alone may explain the extent of the difference in O3 risk estimates across

        January 2005                              7-130        DRAFT-DO NOT QUOTE OR CITE

-------
 1      studies. As use of time-series studies to investigate air pollution effects has become more

 2      common, there has been a great effort to evaluate the issues surrounding these studies.

 3           In this section, conclusions regarding O3 health effects from the epidemiologic evidence

 4      and the issues that may affect the interpretation of the effect estimates are briefly summarized.

 5      A more integrative synthesis of all relevant information will be presented in Chapter 8 of this
 6      AQCD.

 7           (1)    Field/panel studies of acute O? effects.  Results from recent field/panel studies
                   continue to confirm that short-term O3 exposure is associated with acute
                   decrements in lung function, increased respiratory symptoms, and increased
                   medication use, particularly in children and asthmatics. Taken together with the
                   evidence from controlled human exposure studies, O3 is likely causally related
                   to the various respiratory health outcomes.

 8           (2)    O? effects on emergency department visits and hospitalizations.  Large multicity
                   studies,  as well as many studies from individual cities have reported a significant
                   O3 effect on total respiratory, asthma, and COPD hospital visits and admissions.
                   Studies using year-round data noted some inconsistencies in the O3 effect on daily
                   emergency department visits and hospitalizations. However, studies with data
                   restricted to the summer or warm season, in general, indicated positive and  robust
                   associations between ambient O3 concentrations and respiratory morbidity.

 9           (3)    Acute O? effects on mortality. The majority of the studies suggest an elevated risk
                   of mortality associated with acute exposure to O3, especially in the summer or
                   warm season when O3 levels are expected to be high. However, as the magnitude
                   of the O3-mortality risk estimates are generally small, bias due to the uncertainties
                   regarding model specification and adjustment for confounding may be of concern.

10           (4)    Chronic O? effects on morbidity and mortality. Few studies have investigated the
                   effect of chronic O3 exposure on morbidity and mortality. The strongest evidence
                   is for the association between O3 exposure and seasonal decrements or reduced
                   growth in lung function measures in adults and children. Less conclusive are
                   longitudinal studies investigating the association of chronic O3 exposure on yearly
                   lung function, asthma incidence, and respiratory symptoms.  Chronic O3-mortality
                   studies observed inconsistencies across exposure periods, cause-specific mortality
                   outcomes,  and gender. Based on the current evidence, the chronic effect  of O3
                   exposure on morbidity and mortality outcomes is still inconclusive.

11           (5)    Exposure assessement.  Exposure misclassification may result from the use of
                   stationary ambient monitors to determine exposure in population studies.
                   Although central ambient monitors do not explain the variance of individual
                   personal exposures, significant correlations are found between aggregate  personal
                   O3 measurements and O3 concentrations from ambient monitors. A simulation
                   study indicated that the use of ambient monitor data will tend to bias effect
                   estimates towards the null.

        January 2005                              7-131       DRAFT-DO NOT QUOTE OR CITE

-------
      (6)   (X exposure indices. The three most commonly used daily O3 exposure indices,
           1-h max O3, 8-max O3, and 24-h avg O3, were found to be highly correlated in
           studies conducted in various regions. In addition, the effect estimates and
           significance of associations across all health outcomes were comparable when
           using the same distributional increment for all three indices. The commonly
           used 8-h max O3 index, which is also reflective of the new 8-h NAAQS for O3,
           continues to be an appropriate choice.

      (7)   Selection of exposure lag structure.  Most studies did not hypothesize a priori the
           temporal relationship between O3 exposure and the occurrence of health effects.
           Bias can result from the selection of the largest, most significant effect estimates.
           However, the majority of the studies found an immediate O3 effect, with health
           effects having the strongest associations with exposure on the  same day and/or
           previous day. Some studies found greater cumulative effects of O3 over longer
           lag periods, indicating that multiday lags also should be investigated.

      (8)   Sensitivity to model specifications for temporal  trends.  Ozone effect estimates
           that were reported in studies whose  main focus was PM often were calculated
           using the same model specifications as PM. The sensitivity of the  O3 risk
           estimates to alternative model  specifications has not been throughly investigated.
           Uncertainty remains regarding the extent of confounding on estimates of O3 health
           risks, however limited evidence indicates that O3 effects were robust to various
           model specifications for temporal trend adjustment.

      (9)   Influence of seasonal trends. An evaluation of the confounding effects of
           meteorologic factors and copollutants on O3 risk estimates is complicated by their
           changing relationship with O3  across seasons. Mortality and morbidity effect
           estimates calculated using all year or cool season data are generally smaller than
           those from warm season  only data.  In locations where seasonal variability may be
           considerable, efforts should be made to determine season-specific risk estimates.

     (10)  Confounding by copollutants.  Multipollutant models most often are used to
           adjust for confounding by copollutants.  Results from these analyses indicate that
           copollutants generally do not appear to confound the association between O3 and
           acute health effects.  However, due to the varying concurvity across pollutants,
           multipollutant models may not be adequate to determine the independent effects
           of individual pollutants.  Given the limitations, results generally suggest that the
           inclusion of copollutants into the models do not substantially affect O3 risk
           estimates.

     (11)  Model uncertainty and multiple testing.  Various statistical methods have been
           used to assist model selection.  While Bayesian model averaging is a useful tool
           that incorporates model uncertainty  into the effect estimates, its use may  be limited
           due to the large number of variables typically considered in air pollution  health
           effects and the high degree of  correlation between the various  air pollutants.
January 2005                              7-132       DRAFT-DO NOT QUOTE OR CITE

-------
            (12)   Concentration-response function. Supporting evidence for an effect threshold
                  is provided by the numerous studies where O3 effects are seen only in the warm
                  months when O3 levels are higher and more variable. However, in the few
                  mortality and morbidity studies that have specifically examined the O3
                  concentration-response relationship, there is conflicting evidence regarding the
                  presence of an effect threshold. Lack of evidence for a population-level threshold
                  does not preclude the existence of individual thresholds.

            (13)   Spatial variability in O? effects.  Consistent O3 effect estimates were observed
                  overall for mortality, hospitalizations, and other respiratory health outcomes in
                  multicity studies, indicating little heterogeneity of O3 effects by location. The
                  slight heterogeneity observed may be partially attributable to the differences in
                  relative personal exposure to O3 and the varying concentration and composition
                  of copollutants present by region.

            (14)   O3 health effects in asthmatics. The effect of O3 on asthmatics has been examined
                  widely in both time-series studies and field panel studies.  Across various
                  respiratory health outcomes, results were consistently positive and, at times,
                  statistically significant, indicating that asthmatics may be a potentially susceptible
                  population that requires protection from O3 exposures.

            (15)   Age-related differences in (X health effects. Supporting evidence exists for
                  heterogeneity in the effects of O3 by age. The elderly population (> 65  years
                  of age) appear to be at greater risk of O3-related mortality and hospitalizations
                  compared to  all age or younger populations. In addition, negative respiratory
                  health outcomes were associated with O3 exposure in children (< 18 years of age).
5

6
       January 2005                              7-133        DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Abbey, D. E.; Nishino, N.; McDonnell, W. F.; Burchette, R. J.; Knutsen, S. F.; Beeson, W. L.; Yang, J. X. (1999)
  3             Long-term inhalable particles and other air pollutants related to mortality in nonsmokers. Am. J. Respir. Crit.
  4             Care Med. 159: 373-382.
  5       American Thoracic Society. (1991) Lung function testing: selection of reference values and interpretative strategies
  6             Am. Rev. Respir. Dis. 144: 1202-1218.
  7       Anderson, H. R.; Ponce de Leon, A.; Bland, J. M; Bower, J. S.; Strachan, D. P. (1996) Air pollution and daily
  8             mortality in London: 1987-92. Br. Med. J. 312: 665-669.
  9       Anderson, H. R.; Spix, C.; Medina, S.; Schouten, J. P.; Castellsague, J.; Rossi, G.; Zmirou, D.; Touloumi, G.;
10             Wojtyniak, B.; Ponka, A.; Bacharova, L.; Schwartz, J.; Katsouyanni, K. (1997) Air pollution and daily
11             admissions for chronic obstructive pulmonary disease in 6 European cities: results from the APHEA project.
12             Eur. Respir. J.  10: 1064-1071.
13       Anderson, H. R.; Ponce de Leon, A.; Bland, J. M.; Bower, J. S.; Emberlin, J.; Strachen, D. P. (1998) Air pollution,
14             pollens, and daily admissions for asthma in London 1987-92. Thorax 53: 842-848.
15       Atkinson, R. W.; Anderson, H. R.; Strachan, D. P.; Bland, J. M.; Bremner, S. A.; Ponce de Leon, A. (1999a)
16             Short-term associations between outdoor air pollution and visits to accident and emergency departments in
17             London for respiratory complaints.  Eur. Respir. J. 13: 257-265.
18       Atkinson, R. W.; Bremner, S. A.; Anderson, H. R.; Strachan, D. P.; Bland, J.  M.; Ponce de Leon, A. (1999b)
19             Short-term associations between emergency hospital admissions for respiratory and cardiovascular disease
20             and outdoor air pollution in London. Arch. Environ. Health 54: 3 98-411.
21       Avol, E. L.; Trim, S. C.; Little, D. E.; Spier, C. E.; Smith, M. N.; Peng, R.-C.; Linn, W. S.; Hackney, J. D.; Gross,
22             K. B.; D'Arcy,  J. B.; Gibbons, D.; Higgins, I. T. T. (1990) Ozone exposure and lung function in children
23             attending a southern California summer camp.  Presented at: 83rd annual meeting and exhibition of the Air &
24             Waste Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management
25             Association; paper no. 90-150.3.
26       Avol, E. L.; Navidi, W. C.; Rappaport, E.  B.; Peters, J. M. (1998) Acute effects of ambient ozone on asthmatic,
27             wheezy, and healthy children. Cambridge, MA: Health Effects Institute; research report no. 82.
28       Ballester, F.; Tenias, J. M.; Perez-Hoyos,  S. (2001) Air pollution and emergency hospital admissions for
29             cardiovascular diseases in Valencia, Spain. J. Epidemiol. Community Health 55: 57-65.
30       Bates, D. V.; Sizto, R. (1983) Relationship between air pollutant levels and hospital admissions in Southern Ontario.
31             Can. J. Public Health 74: 117-122.
32       Bates, D. V.; Sizto, R. (1987) Air pollution and hospital admissions in southern Ontario: the acid summer haze
33             effect. Environ. Res. 43: 317-331.
34       Bates, D. V.; Sizto, R. (1989) The Ontario Air Pollution study: identification  of the causative agent. Environ. Health
35             Perspect. 79: 69-72.
36       Bates, D. V.; Baker-Anderson, M.; Sizto, R. (1990) Asthma attack periodicity: a study of hospital emergency visits
37             in Vancouver. Environ. Res. 51: 51-70.
38       Beeson, W. L.; Abbey, D. E.; Knutsen, S.  F. (1998) Long-term concentrations of ambient air pollutants and incident
39             lung cancer in California adults: results from the AHSMOG study. Environ. Health Perspect. 106: 813-823.
40       Bell, M. L.; McDermott, A.; Zeger, S. L.;  Samet, J. M.; Dominici, F. (2004) Ozone and short-term mortality in
41             95 US urban communities, 1987-2000. JAMA J. Am. Med. Assoc. 292: 2372-2378.
42       Borja-Aburto, V H.; Loomis, D. P.; Bangdiwala, S. I.; Shy,  C. M.; Rascon-Pacheco, R. A. (1997) Ozone, suspended
43             particulates, and daily mortality in Mexico City. Am.  J. Epidemiol. 145: 258-268.
44       Borja-Aburto, V H.; Castillejos, M.; Gold, D. R.; Bierzwinski, S.; Loomis, D. (1998)  Mortality and ambient fine
45             particles in southwest Mexico City, 1993-1995. Environ. Health Perspect. 106:  849-855.
46       Bourcier, T.; Viboud, C.; Cohen, J.-C.; Thomas, F.; Bury, T.; Cadiot, L.; Mestre, O.; Flahault, A.; Borderie, V;
47             Laroche, L. (2003) Effects of air pollution and climatic conditions on the frequency of ophthalmological
48             emergency examinations. Br. J. Ophthalmol. 87: 809-811.
49       Brauer, M.; Brook, J.  R. (1997) Ozone personal exposures and health effects for selected groups residing in the
50             Fraser Valley. In: Steyn, D. G.; Bottenheim, J. W., eds.  The Lower Fraser Valley Oxidants/Pacific '93 Field
51             Study. Atmos. Environ. 31: 2113-2121.
52       Brauer, M.; Blair, J.; Vedal, S. (1996) Effect of ambient ozone exposure on lung function in farm workers. Am. J.
53             Respir. Crit.  Care Med. 154: 981-987.
         January 2005                                   7-134        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Bremner, S. A.; Anderson, H. R.; Atkinson, R. W.; McMichael, A. J.; Strachan, D. P.; Bland, J. M; Bower, J. S.
  2             (1999) Short term associations between outdoor air pollution and mortality in London 1992-4. Occup.
  3             Environ. Med. 56: 237-244.
  4       Buchdahl, R.; Parker, A.; Stebbings, T.; Babiker, A. (1996) Association between air pollution and acute childhood
  5             wheezy episodes: prospective observational study. Br. Med. J. 312: 661-664.
  6       Buchdahl, R.; Willems, C. D.; Vander, M.; Babiker, A. (2000) Associations between ambient ozone, hydrocarbons,
  7             and childhood wheezy episodes: a prospective observational study in south east London. Occup. Environ.
  8             Med. 57: 86-93.
  9       Burchfiel, C. M.; Marcus, E. B.; Curb, J. D.; Maclean, C. J.; Vollmer, W. M.; Johnson, L. R.; Fong, K.; Rodriguez,
10             B. L.; Masaki, K. H.; Buist, A. S. (1995) Effects of smoking and smoking cessation on longitudinal decline in
11             pulmonary function. Am. J. Respir. Crit. Care Med.  151: 1778-1785.
12       Burnett, R. T.; Dales, R. E.; Raizenne, M. E.; Krewski, D.; Summers, P. W.; Roberts, G. R.; Raad-Young, M.;
13             Dann, T.; Brook, J. (1994) Effects of low ambient levels of ozone and sulfates on the frequency of respiratory
14             admissions to Ontario hospitals. Environ. Res. 65: 172-194.
15       Burnett, R. T.; Dales, R.; Krewski, D.; Vincent, R.; Dann, T.; Brook, J. R. (1995) Associations between ambient
16             paniculate sulfate and admissions to Ontario hospitals for cardiac and respiratory diseases. Am. J. Epidemiol.
17             142: 15-22.
18       Burnett, R. T.; Brook, J. R.; Yung, W. T.; Dales, R. E.; Krewski, D. (1997a) Association between ozone and
19             hospitalization for respiratory diseases in 16 Canadian cities. Environ. Res. 72: 24-31.
20       Burnett, R. T.; Cakmak, S.; Brook, J. R.; Krewski, D. (1997b) The role of paniculate size and chemistry in the
21             association between summertime ambient air pollution and hospitalization for cardiorespiratory diseases.
22             Environ. Health Perspect. 105: 614-620.
23       Burnett, R. T.; Smith-Doiron, M.;  Stieb, D.; Cakmak, S.; Brook, J. R. (1999) Effects of paniculate and gaseous air
24             pollution on cardiorespiratory hospitalizations. Arch. Environ. Health 54: 130-139.
25       Burnett, R. T.; Smith-Doiron, M.;  Stieb, D.; Raizenne, M. E.; Brook, J. R.; Dales, R. E.; Leech, J. A.; Cakmak, S.;
26             Krewski, D. (2001) Association between ozone and hospitalization for acute respiratory  diseases in children
27             less than 2 years of age. Am. J. Epidemiol. 153: 444-452.
28       Calderon-Garciduenas, L.; Rodriguez-Alcaraz, A.; Garcia,  R.; Ramirez, L.; Barragan, G. (1995) Nasal inflammatory
29             responses in children exposed to a polluted urban atmosphere. J. Toxicol. Environ. Health 45: 427-437.
30       Calderon-Garciduenas, L.; Osnaya, N.; Rodriguez-Alcaraz, A.; Villarreal-Calderon, A. (1997) DNA damage in nasal
31             respiratory epithelium from children exposed to urban pollution. Environ. Mol. Mutagen. 30: 11-20.
32       Calderon-Garciduenas, L.; Wen-Wang, L.; Zhang, Y.-J.; Rodriguez-Alcaraz, A.; Osnaya, N.; Villarreal-Calderon,
33             A.; Santella, R. M. (1999) 8-hydroxy-2'-deoxyguanosine, a major mutagenic oxidative DNA lesion, and DNA
34             strand breaks  in nasal respiratory epithelium of children exposed to urban pollution. Environ. Health Perspect.
35             107:469-474.
36       Castellsague, J.; Sunyer, J.; Saez, M.; Anto, J. M. (1995) Short-term association between air pollution and
37             emergency room visits for asthma in Barcelona. Thorax 50: 1051-1056.
38       Castillejos, M.; Gold, D. R.; Damokosh, A. I.; Serrano, P.;  Allen, G.; McDonnell, W. F.; Dockery, D.; Velasco,
39             S. R.; Hernandez, M.; Hayes, C. (1995) Acute effects of ozone on the pulmonary function of exercising
40             schoolchildren from Mexico City. Am. J. Respir. Crit. Care Med. 152: 1501-1507.
41       Chang, L.-T.; Koutrakis, P.; Catalano, P. J.; Suh, H. H. (2000) Hourly personal exposures to fine  particles and
42             gaseous pollutants—results from Baltimore, Maryland. J. Air Waste Manage. Assoc. 50: 1223-1235.
43       Charpin, D.; Pascal,  L.; Birnbaum, J.; Armengaud, A.; Sambuc, R.; Lanteaume, A.; Vervloet, D. (1999) Gaseous air
44             pollution and atopy. Clin. Exp. Allergy 29: 1474-1480.
45       Chen, P.-C; Lai, Y.-M.; Wang, J.-D.; Yang, C.-Y.; Hwang, J.-S.; Kuo, H.-W.;  Huang, S.-L.; Chan, C.-C. (1998)
46             Adverse effect of air pollution on respiratory health  of primary school children in Taiwan.  Environ. Health
47             Perspect. 106:331-335.
48       Chen, P.-C.; Lai, Y.-M.; Chan, C.-C.; Hwang, J.-S.; Yang,  C.-Y.; Wang, J.-D. (1999) Short-term effect of ozone on
49             the pulmonary function of children in primary school. Environ. Health Perspect. 107: 921-925.
50       Chen, L.; Jennison, B. L.; Yang, W.; Omaye, S. T. (2000) Elementary school absenteeism and air pollution.
51             Inhalation Toxicol. 12: 997-1016.
52       Chen, L.; Yang, W.; Jennison, B. L.; Goodrich, A.; Omaye, S. T. (2002) Air pollution and birth weight in northern
53             Nevada, 1991-1999. Inhalation Toxicol. 14: 141-157.
54       Chew, F. T.; Goh, D. Y. T.; Ooi, B. C.; Saharom, R.; Hui, J. K. S.; Lee, B. W. (1999) Association of ambient
55             air-pollution levels with acute asthma exacerbation among children in Singapore. Allergy (Copenhagen)
56             54:320-329.
         January 2005                                   7-135         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Chock, D. P.; Winkler, S. L.; Chen, C. (2000) A study of the association between daily mortality and ambient air
  2             pollutant concentrations in Pittsburgh, Pennsylvania. J. Air Waste Manage. Assoc. 50: 1481-1500.
  3       Cifuentes, L. A.; Vega, J.; Kopfer,  K.; Lave, L. B. (2000) Effect of the fine fraction of paniculate matter versus the
  4             coarse mass and other pollutants on daily mortality in Santiago, Chile. J. Air Waste Manage. Assoc. 50:
  5             1287-1298.
  6       Clyde, M. A. (1999) Bayesian model averaging and model search strategies. In: Bernardo, J. M.; Berger, J. O.;
  7             Dawid, A. P.; Smith, A. F. M., eds. Bayesian Statistics 6: proceedings of the Sixth Valencia International
  8             Meeting, June; pp. 157-185. Oxford,  UK. Oxford, UK: Clarendon Press.
  9       Clyde, M. (2000) Model uncertainty and health effect studies for paniculate matter. Environmetrics 11: 745-763.
10       Clyde, M. A.; Guttorp, P.; Sullivan, E. (2000) Effects of ambient fine and coarse  particles on mortality in Phoenix,
11             Arizona. Seattle, WA: University of Washington, National Research Center for Statistics and the
12             Environment; NRCSE technical report series, NRCSE-TRS no. 040. Available:
13             http://www.nrcse.washington.edu/pdf/trs40_pm.pdf [18 October, 2004].
14       Cody, R. P.; Weisel, C. P.; Birnbaum, G.; Lioy, P. J. (1992) The effect of ozone associated with summertime
15             photochemical smog on the frequency of asthma visits to hospital emergency departments. Environ. Res.
16             58: 184-194.
17       Cuijpers, C. E. J.; Swaen, G. M. H.; Wesseling, G.; Wouters, E. F. M. (1994) Acute respiratory effects of summer
18             smog in primary school children. Toxicol. Lett. 72: 227-235.
19       Dab, W.; Medina, S.; Quenel, P.; Le Moullec, Y.; Le Tertre, A.; Thelot, B.; Monteil, C.; Lameloise, P.; Pirard, P.;
20             Momas, I.; Ferry, R.; Festy,  B. (1996) Short term respiratory health effects of ambient air pollution:  results of
21             the APHEA project in Paris. In: St Leger, S., ed. The APHEA project. Short term effects of air pollution on
22             health: a European approach using epidemiological time series data. J. Epidemiol. Commun. Health
23             50(suppl. 1): S42-S46.
24       De Leon, S. F.; Thurston, G. D.; Ito, K. (2003) Contribution of respiratory disease to nonrespiratory mortality
25             associations with air pollution. Am. J. Respir. Crit. CareMed.  167:  1117-1123.
26       Delfino, R. J.; Coate, B. D.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; Koutrakis, P. (1996) Daily asthma severity in
27             relation to personal ozone exposure and outdoor fungal spores. Am. J. Respir. Crit. Care Med.  154: 633-641.
28       Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; Matteucci, R. M.; Anderson, P. R.; Koutrakis, P. (1997a)
29             The effect of outdoor fungal spore concentrations on daily asthma severity. Environ. Health Perspect.
30             105:622-635.
31       Delfino, R. J.; Murphy-Moulton, A. M.; Burnett, R. T.; Brook, J. R.; Becklake, M. R. (1997b) Effects of air
32             pollution on emergency room visits for respiratory illnesses in Montreal, Quebec. Am. J. Respir. Crit. Care
33             Med. 155: 568-576.
34       Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H. (1998a) Symptoms inpediatric asthmatics and air pollution:
3 5             differences in effects by symptom severity, anti-inflammatory medication use and paniculate averaging time.
36             Environ. Health Perspect.  106: 751-761.
37       Delfino, R. J.; Murphy-Moulton, A. M.; Becklake, M. R. (1998b) Emergency room visits for respiratory illnesses
3 8             among the elderly in Montreal: association with low level ozone exposure. Environ.  Res. 76: 67-77.
39       Delfino, R. J.; Gone, H.; Linn, W. S.; Pellizzari, E. D.; Hu, Y. (2003) Asthma symptoms in Hispanic  children and
40             daily ambient exposures to toxic and criteria air pollutants. Environ. Health Perspect. Ill: 647-656.
41       Delfino, R. J., Quintana, P. J. E.; Flora, J.; Gastanaga, V M.; Samimi, B. S.; Kleinman, M. T.; Liu, L.-J. S.;
42             Bufalino, C.; Wu, C.-F.; McLaren, C. E. (2004) Association of FEVj in asthmatic children with personal and
43             microenvironmental exposure to airborne particulate matter. Environ. Health Perspect. 112: 932-941.
44       Desqueyroux, H.; Pujet, J.-C.; Prosper, M.;  Squinazi, F.; Momas, I. (2002a) Short-term effects of low-level air
45             pollution on respiratory health of adults suffering from moderate to severe asthma. Environ. Res. A 89: 29-37.
46       Desqueyroux, H.; Pujet, J.-C.; Prosper, M.; Le Moullec, Y.; Momas, I. (2002b) Effects of air pollution on adults
47             with chronic obstructive pulmonary disease. Arch. Environ. Health 57: 554-560.
48       Detels, R.; Tashkin, D. P.; Sayre, J. W.; Rokaw, S. N.; Coulson, A. H.; Massey, F. J., Jr.; Wegman, D. H. (1987)
49             The UCLA population studies of chronic obstructive respiratory disease: 9. lung function changes associated
50             with chronic exposure to photochemical oxidants; a cohort study among never-smokers. Chest 92: 594-603.
51       Devlin, R. B.; McDonnell,  W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991)
52             Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the
53             lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
54       Diaz, J.; Garcia, R.; Ribera, P.; Alberdi, J. C.; Hernandez, E.; Pajares, M. S.; Otero, A. (1999) Modeling of air
55             pollution and its relationship with mortality and morbidity in Madrid, Spain. Int. Arch. Occup. Environ.
56             Health 72: 366-376.
         January 2005                                   7-136         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Dockery, D. W.; Schwartz, I; Spengler, J. D. (1992) Air pollution and daily mortality: associations with particulates
  2             and acid aerosols. Environ. Res. 59: 362-373.
  3       Dominici, F.; McDermott, A.; Zeger, S. L.; Samet, J. M. (2002) On the use of generalized additive models in
  4             time-series studies of air pollution and health. Am. J. Epidemiol.  156: 193-203.
  5       Dominici, F.; McDermott, A.; Daniels, M.; Zeger, S. L.; Samet, J. M. (2003) Mortality among residents of 90 cities.
  6             In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA: Health
  7             Effects Institute; pp. 9-24. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
  8       Fairley, D. (1999) Daily mortality and air pollution in Santa Clara County, California: 1989-1996. Environ. Health
  9             Perspect. 107: 637-641.
10       Fairley, D. (2003) Mortality and air pollution for Santa Clara County, California, 1989-1996. In: Revised analyses of
11             time-series studies of air pollution and health. Special report. Boston, MA: Health Effects Institute;
12             pp. 97-106. Available: http://www.healtheffects.org/news.htm [16 May, 2003].
13       Fortoul, T. L; Valverde, M.; Lopez M. D. C.; Bizarro, P.; Lopez,  L; Sanchez, I.; Colin-Barenque, L.; Avila-Costa,
14             M. R.; Rojas, E.; Ostrosky-Shejet, P. (2003) Single-cell gel electrophoresis assay of nasal epithelium and
15             leukocytes from asthmatic and nonasthmatic subjects in Mexico City. Arch. Environ. Health 58: 348-352.
16       Friedman, M. S.; Powell, K. E.; Hutwagner, L.; Graham, L. M.; Teague, W. G. (2001) Impact of changes in
17             transportation and commuting behaviors during the 1996 summer Olympic games in Atlanta on air quality and
18             childhood asthma. JAMA J. Am. Med. Assoc. 285: 897-905.
19       Frischer, T. M.; Ktihr,  J.; Pullwitt, A.; Meinert, R.; Forster, J.; Studnicka, M.; Koren, H. (1993) Ambient ozone
20             causes upper airways inflammation in children. Am. Rev. Respir. Dis. 148: 961-964.
21       Frischer, T.; Pullwitt, A.; Kuehr, K.; Meinert, R.; Haschke, N.; Studnicka, M.; Lubec, G. (1997)  Aromatic
22             hydroxylation in nasal lavage fluid following ambient ozone exposure. Free Radical Biol. Med. 22: 201-207.
23       Frischer, T.; Studnicka, M.; Gartner,  C.; Tauber, E.; Horak, F.; Veiter, A.; Spengler, J.; Ktihr, J.; Urbanek, R. (1999)
24             Lung function growth and ambient ozone: a three-year population study in school children. Am. J. Respir.
25             Crit. Care Med. 160: 390-396.
26       Frischer, T.; Studnicka, M.; Halmerbauer, G.; Horak, F.; Gartner, C.; Tauber, E.; Koller, D. Y. (2001) Ambient
27             ozone exposure is associated with eosinophil activation in healthy children. Clin. Exp. Allergy 31: 1213-1219.
28       Fuhlbrigge, A.; Kitch,  B.; Paltiel, A. D.; et al. (2001) FEV{ is associated with risk of asthma attacks in a pediatric
29             population. J. Allergy Clin. Immunol. 107: 61-67.
30       Galizia, A.; Kinney, P. L. (1999) Long-term residence in areas of high ozone: associations with respiratory health in
31             a nationwide sample of nonsmoking young adults. Environ. Health Perspect. 107: 675-679.
32       Gamble, J. L. (1998) Effects of ambient air pollution on daily mortality: a time series analysis of Dallas, Texas,
33             1990-1994. Presented at: 91st  annual meeting and exhibition of the Air & Waste Management Association;
34             June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; paper no. 98-MP26.03.
35       Garcia-Aymerich, J.; Tobias, A.; Anto, J. M.; Sunyer, J. (2000) Air pollution and mortality in a cohort of patients
36             with chronic obstructive pulmonary disease: a time series analysis. J. Epidemiol. Community Health
37             54:73-74.
38       Garty, B. Z.; Kosman,  E.; Ganor, E.;  Berger, V.; Garty, L.; Wietzen, T.; Waisman, Y.; Mimouni, M.; Waisel, Y.
39             (1998) Emergency room visits of asthmatic children, relation to air pollution, weather, and airborne allergens.
40             Ann.  Allergy Asthma Immunol. 81: 563-570.
41       Gauderman, W. J.; McConnell, R.; Gilliland, F.; London, S.; Thomas, D.; Avol, E.; Vora, H.; Berhane, K.;
42             Rappaport, E. B.; Lurmann, F.; Margolis, H.  G.; Peters, J.  (2000) Association between air pollution and lung
43             function growth in southern California children. Am. J. Respir. Crit. Care Med. 162: 1383-1390.
44       Gauderman, W. J.; Gilliland, G. F.; Vora, H.; Avol, E.; Stram, D.; McConnell, R.; Thomas, D.; Lurmann, F.;
45             Margolis, H. G.; Rappaport, E. B.; Berhane, K.; Peters, J. M. (2002) Association between air pollution and
46             lung function growth in southern California children: results from a second cohort. Am. J. Respir. Crit. Care
47             Med. 166: 76-84.
48       Gauderman, W. J.; Avol, E.; Gilliland, F.; Vora, H.; Thomas, D.; Berhane, K.; McConnell, R.; Kuenzli, N.;
49             Lurmann, F.; Rappaport, E.; Margolis, H.; Bates, D.; Peters,  J. (2004a) The effect of air pollution on lung
50             development from 10 to 18 years of age. N. Engl. J. Med.  351: 1057-1067.
51       Gauderman, W. J.; Avol, E.; Gilliland, F. (2004b) Air pollution and lung function [reply letter]. N. Engl. J. Med.
52             351:2653.
53       Gent, J. F.; Triche, E. W.; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W.  S.; Leaderer, B. P. (2003)
54             Association of low-level ozone and fine particles with respiratory symptoms in children with asthma. JAMA
55             J. Am. Med. Assoc. 290: 1859-1867.
56       George, E. I. (1999) Comment on Hoeting et al., 1999, "Bayesian model averaging: a tutorial." Stat. Sci.
57             14:409-412.


         January 2005                                   7-137         DRAFT-DO NOT QUOTE  OR CITE

-------
  1       Geyh, A. S.; Xue, I; Ozkaynak, H.; Spengler, J. D. (2000) The Harvard Southern California chronic ozone exposure
  2             study: assessing ozone exposure of grade-school-age children in two southern California communities.
  3             Environ. Health Perspect. 108: 265-270.
  4       Gielen, M. H.; VanDer Zee, S. C.; Van Wijnen, J. H.; Van Steen, C. J.; Brunekreef, B. (1997) Acute effects of
  5             summer air pollution on respiratory health of asthmatic children. Am. J. Respir. Crit. Care Med.
  6             155:2105-2108.
  7       Gilliland, F. D.; Berhane, K.; Rappaport, E. B.; Thomas, D. C.; Avol, E.; Gauderman, W. J.; London, S. J.;
  8             Margolis, H. G.; McConnell, R.; Islam, K. T.; Peters, J. M. (2001) The effects of ambient air pollution on
  9             school absenteeism due to respiratory illnesses. Epidemiology 12: 43-54.
10       Gold, D. R.; Damokosh, A. I.; Pope, C. A., Ill; Dockery, D. W.; McDonnell, W. F.; Serrano, P.; Retama, A.;
11             Castillejos, M. (1999) Paniculate and ozone pollutant effects on the respiratory function of children in
12             southwest Mexico City. Epidemiology 10: 8-16.
13       Gold, D. R.; Litonjua,  A.; Schwartz, J.; Lovett, E.; Larson, A.; Nearing, B.; Allen, G.; Verrier, M.; Cherry, R.;
14             Verrier, R. (2000) Ambient pollution and heart rate variability. Circulation 101:  1267-1273.
15       Gold, D. R.; Schwartz, J.; Litonjua, A.; Verrier, R.; Zanobetti, A. (2003) Ambient pollution and reduced heart rate
16             variability. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA:
17             Health Effects Institute; pp. 107-112. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf
18             [18 October, 2004].
19       Goldberg, M. S.; Burnett, R. T. (2003) Revised analysis of the Montreal time-series study. In:  Revised analyses of
20             time-series studies of air pollution and health. Special report. Boston, MA: Health Effects Institute;
21             pp. 113-132. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [13 August 2003).
22       Goldberg, M. S.; Burnett, R. T.; Valois, M.-F.; Flegel, K.; Bailar, J. C., Ill; Brook, J.; Vincent, R.; Radon, K. (2003)
23             Associations between ambient air pollution and daily mortality among persons with congestive heart failure.
24             Environ. Res. 91:8-20.
25       Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B.
26             (1998a) Cardiovascular effects of ozone exposure in human volunteers. Am. J. Respir. Crit. Care Med.
27             158: 538-546.
28       Gong, H., Jr.; Simmons, M.  S.; Linn, W. S.; McDonnell, W. F.; Westerdahl, D. (1998b) Relationship between acute
29             ozone responsiveness and chronic loss of lung function in residents of a high-ozone community.
30             Arch. Environ. Health 53: 313-319.
31       Gonzales, M.; Ngo, L.; Hammond, S. K.; Tager, I. (2003) Validation of a questionnaire and microenvironmental
32             model for estimating past exposures to ozone. Int. J. Environ. Health Res. 13: 249-260.
33       Goss, C. H.; Newsom, S. A.; Schildcrout, J. S.; Sheppard, L.; Kaufman, J. D. (2004) Effect of ambient air pollution
34             on pulmonary exacerbations and lung function in cystic fibrosis. Am. J. Respir. Crit. Care Med. 169: 816-821.
3 5       Gouveia, N.; Fletcher, T. (2000a) Respiratory diseases in children and outdoor air pollution in Sao Paulo, Brazil:
36             a time series analysis. Occup. Environ. Med. 57: 477-483.
37       Gouveia, N.; Fletcher, T. (2000b) Time series analysis of air pollution and mortality: effects by cause, age and
38             socioeconomic status. J. Epidemiol. Community Health 54: 750-755.
39       Gouveia, N.; Bremner, S. A.; Novaes, H. M. D. (2004) Association between ambient air pollution and birth weight
40             in Sao Paulo, Brazil. J. Epidemiol. Community Health 58: 11-17.
41       Greer, J. R.; Abbey, D. E.; Burchette, R. J. (1993) Asthma related to occupational and ambient air pollutants in
42             nonsmokers. J. Occup. Med.  35: 909-915.
43       Gryparis, A.; Forsberg, B.; Katsouyanni, K.; Analitis, A.; Touloumi, G.; Schwartz, J.; Samoli, E.; Medina, S.;
44             Anderson, H. R.; Niciu, E. M.; Wichmann, H.-E.; Kriz, B.; Kosnik, M.; Skorkovsky, J.; Vonk, J. M.;
45             Dortbudak, Z. (2004) Acute effects of ozone on mortality from the  "air pollution and health: a European
46             approach" project. Am. J. Respir. Crit. Care Med. 170: 1080-1087.
47       Hagen, J. A.; Nafstad,  P.; Skrondal, A.; Bjerkly, S.; Magnus, P. (2000) Associations between outdoor air pollutants
48             and hospitalization for respiratory diseases. Epidemiology 11: 136-140.
49       Health Effects Institute. (2003) Revised analyses of time-series studies of air pollution and health. Boston, MA:
50             Health Effects Institute; special report. Available: Available:
51             http://www.healtheffects.org/Pubs/TimeSeries.pdf [27 June 2003].
52       Hernandez-Garduno, E.; Perez-Neria, J.; Paccagnella, A. M.; Pina-Garcia, M.; Munguia-Castro, M.;
53             Catalan-Vazquez, M.; Rojas-Ramos, M. (1997) Air pollution and respiratory health in Mexico City. J. Occup.
54             Environ. Med. 39: 299-307.
55       Higgins, I. T. T.; D'Arcy, J. B.; Gibbons, D. I.; Avol, E. L.; Gross, K. B. (1990) Effect of exposures to ambient
56             ozone on ventilatory lung function in children. Am. Rev. Respir. Dis. 141: 1136-1146.
57       Hill, A.  B. (1965) The environment and disease: association or causation? Proc. R. Soc. Med.  58: 295-300.


         January 2005                                   7-138        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Hiltermann, T. J. N.; Stolk, J.; Van der Zee, S. C.; Branekreef, B.; De Bruijne, C. R.; Fischer, P. H.; Ameling, C. B.;
  2             Sterk, P. J.; Hiemstra, P. S.; Van Bree, L. (1998) Asthma severity and susceptibility to air pollution. Eur.
  3             Respir. J. 11:686-693.
  4       Hoek, G. (2003) Daily mortality and air pollution in The Netherlands. In: Revised analyses of time-series studies of
  5             air pollution and health. Special report. Boston, MA: Health Effects Institute; pp. 133-142. Available:
  6             http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
  7       Hoek, G.; Branekreef, B. (1995) Effect of photochemical air pollution on acute respiratory symptoms in children.
  8             Am. J. Respir. Crit. Care Med.  151: 27-32.
  9       Hoek, G.; Branekreef, B.; Verhoeff, A.; Van Wijnen, J.; Fischer, P. (2000) Daily mortality and air pollution in the
10             Netherlands. J. Air Waste Manage. Assoc. 50: 1380-1389.
11       Hoek, G.; Branekreef, B.; Fischer, P.;  Van Wijnen, J. (2001) The association between air pollution and heart failure,
12             arrhythmia, embolism, thrombosis, and other cardiovascular causes of death in a time series  study.
13             Epidemiology 12: 355-357.
14       Hoeting, J.  A.; Madigan, D.; Raftery, A. E.; Volinsky, C. T. (1999) Bayesian model averaging: a tutorial. Stat. Sci.
15             14:382-417.
16       Holguin, F.; Tellez-Rojo, M. M.; Hernandez, M.; Cortez, M.;  Chow, J. C.; Watson, J. G.; Mannino, D.; Romieu, I.
17             (2003) Air pollution and heart rate variability among the elderly in Mexico City. Epidemiology 14: 521-527.
18       Holmen, A.; Blomqvist, J.; Frindberg, H.; Johnelius, Y.; Eriksson, N. E.; Henricson, K. A.; Herrstrom, P.; Hogstedt,
19             B. (1997) Frequency of patients with acute asthma in relation to ozone, nitrogen dioxide, other pollutants of
20             ambient air and meteorological observations. Int. Arch. Occup. Environ. Health 69: 317-322.
21       Hoppe, P.; Praml, G.; Rabe, G.; Lindner, J.; Frahmann, G.; Kessel, R. (1995a) Environmental ozone field study on
22             pulmonary and subjective responses of assumed risk groups. Environ. Res. 71: 109-121.
23       Hoppe, P.; Lindner, J.; Praml, G.;  Bronner, N. (1995b) Effects of environmental ozone on the lung function of senior
24             citizens. Int. J. Biometeorol. 38: 122-125.
25       Horak, F., Jr.; Studnicka, M.; Gartner, C.; Spengler, J. D.; Tauber, E.; Urbanek, R.; Veiter, A.; Frischer, T. (2002a)
26             Paniculate matter and lung function growth in children: a 3-yr follow-up study in Austrian schoolchildren.
27             Eur.  Respir. J. 19: 838-845.
28       Horak, F., Jr.; Studnicka, M.; Gartner, C.; Spengler, J. D.; Tauber, E.; Urbanek, R.; Veiter, A.; Frischer, T. (2002b)
29             Paniculate matter and lung function growth in children: a 3-yr follow-up study in Austrian schoolchildren
30             [author response]. Eur. Respir. J. 20: 1355.
31       Huang, Y.;  Dominici, F.; Bell, M. L. (2004) Bayesian hierarchical distributed lag models for summer ozone
32             exposure and cardio-respiratory mortality. Environmetrics: in press.
33       Ihorst, G.; Frischer, T.; Horak, F.; Schumacher, M.; Kopp, M.; Forster, J.; Mattes, J.; Kuehr, J. (2004) Long- and
34             medium-term ozone effects on lung growth including a broad spectrum of exposure. Eur. Respir. J.
35             23:292-299.
36       Ilabaca, M.; Olaeta, I.; Campos, E.; Villaire, J.; Tellez-Rojo, M. M.;  Romieu, I. (1999) Association between levels
37             of fine paniculate and emergency visits for pneumonia and other respiratory illnesses among children in
38             Santiago, Chile. J. Air Waste Manage. Assoc. 49: 154-163.
39       International Commission on Radiological Protection (ICRP). (1994) Human respiratory  tract model for radiological
40             protection: a report of a task group of the International  Commission on Radiological Protection. Oxford,
41             United Kingdom: Elsevier  Science Ltd. (ICRP publication 66; Annals of the ICRP: v. 24, pp. 1-482).
42       Ito, K. (2003) Associations of paniculate matter components with daily mortality and morbidity in Detroit,
43             Michigan. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA:
44             Health Effects Institute; pp. 143-156. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May,
45             2004].
46       Ito, K. (2004) Revised ozone risk  estimates for daily mortality in Detroit, Michigan [personal communication with
47             attachments to Jee Young Kim]. New York, NY: New York University  School of Medicine, Nelson Institute
48             of Environmental Medicine; October 31.
49       Ito, K.; Thurston, G. D. (1996) Daily PM10/mortality associations:  an investigation of at-risk subpopulations.
50             J. Exposure Anal. Environ. Epidemiol. 6: 79-95.
51       Jaffe, D. H.; Singer, M. E.; Rimm, A. A. (2003) Air pollution and emergency department visits for asthma among
52             Ohio Medicaid recipients, 1991-1996. Environ. Res. 91: 21-28.
53       Jalaludin, B. B.; Chey, T.; O'Toole, B. I.; Smith, W. T.; Capon, A. G.; Leeder, S. R. (2000) Acute effects of low
54             levels of ambient ozone on peak expiratory flow rate in a cohort of Australian children. Int. J. Epidemiol.
55             29: 549-557.
         January 2005                                   7-139         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Jalaludin, B. B.; O'Toole, B. I.; Leader, S. R. (2004) Acute effects of urban ambient air pollution on respiratory
  2             symptoms, asthma medication use, and doctor visits for asthma in a cohort of Australian children. Environ
  3             Res. 95: 32-42.
  4       Jenkins, P. L.; Phillips, T. I; Mulberg, E. I; Hui, S. P. (1992) Activity patterns of Californians: use of and proximity
  5             to indoor pollutant sources. Atmos. Environ. Part A 26: 2141-2148.
  6       Jones, G. N.; Sletten, C.; Mandry, C.; Brantley, P. J. (1995) Ozone level effect on respiratory illness: an
  7             investigation of emergency department visits. South. Med. J. 88: 1049-1056.
  8       Just, J.; Segala, C.; Sahraoui, F.; Priol, G.; Grimfeld, A.; Neukirch, F. (2002) Short-term health effects of paniculate
  9             and photochemical air pollution in asthmatic children. Eur. Respir. J. 20: 899-906.
10       Kim, S.-Y.; Lee, J.-T.; Hong, Y.-C; Ahn, K.-J.; Kim, H. (2004) Determining the threshold effect of ozone on daily
11             mortality: an analysis of ozone and mortality in Seoul, Korea, 1995-1999. Environ. Res. 94: 113-119.
12       Kinney, P. L.; Lippmann, M. (2000) Respiratory effects of seasonal exposures to ozone and particles. Arch. Environ.
13             Health 55: 210-216.
14       Kinney, P. L.; Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County.
15             Environ. Res. 54:  99-120.
16       Kinney, P. L.; Ito, K.; Thurston, G. D. (1995) A sensitivity analysis of mortality/PM10 associations in Los Angeles.
17             In: Phalen, R. F.; Bates, D. V., eds. Proceedings of the colloquium on paniculate air pollution and human
18             mortality and morbidity; January 1994; Irvine, CA. Inhalation Toxicol. 7: 59-69.
19       Kinney, P. L.; Nilsen, D. M.; Lippmann, M.; Brescia, M.; Gordon, T.; McGovern, T.; El Fawal, H.; Devlin, R. B.;
20             Rom, W. N. (1996) Biomarkers of lung inflammation in recreational joggers exposed to ozone. Am. J. Respir.
21             Crit. Care Med. 154: 1430-1435.
22       Kinney, P. L.; Aggarwal, M.; Nikiforov, S. V.; Nadas, A. (1998) Methods development for epidemiologic
23             investigations of the health effects of prolonged ozone exposure. Part III: an approach to retrospective
24             estimation of lifetime ozone exposure using a questionnaire and ambient monitoring data (U.S. sites).
25             Cambridge, MA: Health Effects Institute; research report no. 81; pp. 79-108.
26       Klemm, R. J.; Mason, R. M., Jr. (2000) Aerosol Research and Inhalation Epidemiological Study (ARIES):
27             air quality and daily mortality  statistical modeling—interim results. J. Air. Waste Manage. Assoc.
28             50: 1433-1439.
29       Klemm, R. J.; Lipfert, F. W.; Wyzga, R. E.; Gust, C. (2004) Daily mortality and air pollution in Atlanta: two years
30             of data from ARIES. Inhalation Toxicol. 16(suppl. 1): 131-141.
31       Klepeis, N. E. (1999) An introduction to the indirect exposure assessment approach: modeling human exposure
32             using microenvironmental measurements and the recent national human activity pattern survey. Environ.
33             HealthPerspect. Suppl.  107(2): 365-374.
3 4       Koop, G.; Tole, L. (2004) Measuring the health effects of air pollution: to what extent can we really say that people
35             are dying from bad air?  J. Environ. Econ. Manage. 47: 30-54.
36       Kopp, M. V.; Ulmer, C.; Ihorst, G.; Seydewitz, H. H.; Frischer, T.; Forster, J.; Kuehr, J. (1999) Upper airway
3 7             inflammation in children exposed to ambient ozone and potential signs of adaptation.  Eur. Respir. J.
38             14:854-861.
39       Kopp, M. V.; Bohnet, W.; Frischer, T.; Ulmer, C.; Studnicka, M.; Ihorst, G.; Gardner, C.; Forster, J.; Urbanek, R.;
40             Kuehr, J. (2000) Effects of ambient ozone  on lung function in children over a two-summer period. Eur.
41             Respir. J. 16: 893-900.
42       Koren, H. S.; Hatch, G. E.; Graham, D. E. (1990) Nasal lavage as a tool in assessing acute inflammation in response
43             to inhaled pollutants. Toxicology 60: 15-25.
44       Korrick, S. A.; Neas, L. M.; Dockery, D. W.; Gold, D. R.; Allen, G. A.; Hill, L. B.; Kimball, K. D.; Rosner, B. A.;
45             Speizer, F. E. (1998) Effects of ozone and  other pollutants on the pulmonary function of adult hikers.
46             Environ. Health Perspect. 106: 93-99.
47       Krzyzanowski, M. (1997) Methods for assessing the extent of exposure and effects of air pollution. Occup. Environ.
48             Med.  54: 145-151.
49       Krzyzanowski, M.; Quackenboss, J. J.; Lebowitz, M. D. (1992) Relation of peak expiratory flow rates and symptoms
50             to ambient ozone. Arch. Environ. Health 47: 107-115.
51       Ktinzli, N.; Lurmann, F.; Segal, M.; Ngo, L.; Balmes, J.; Tager, I. B. (1997) Association between lifetime ambient
52             ozone exposure and pulmonary function in college freshmen—results of a pilot study. Environ. Res. 72: 8-23.
53       Kuo, H. W.; Lai, J. S.; Lee, M. C.; Tai, R. C.; Lee, M. C. (2002) Respiratory effects of air pollutants among
54             asthmatics in central Taiwan. Arch. Environ. Health 57: 194-200.
         January 2005                                   7-140         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Lagerkvist, B. I; Bernard, A.; Blomberg, A.; Bergstrom, E.; Forsberg, B.; Holmstrom, K.; Karp, K.; Lundstrom,
  2             N.-G.; Segerstedt, B.; Svensson, M; Nordberg, G. (2004) Pulmonary epithelial integrity in children:
  3             relationship to ambient ozone exposure and swimming pool attendance. Environ. Health Perspect.
  4             112: 1768-1771.
  5       Langstaff, J. (2003) Percentiles of 1996-2000 ozone concentrations [memorandum to Joe Pinto]. Research Triangle
  6             Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards; September
  7             17.
  8       Lebowitz, M. D.; Camilli, A. E.; Bronnimann, D.; Quackenboss, J. (1987) The significance and meaningfulness of
  9             intraindividual changes in objective test results as responses to air contaminants. Presented at: 80th annual
10             meeting of the Air Pollution Control Association; June; New York, NY. Pittsburgh, PA: Air Pollution Control
11             Association; paper no. 87-32.1.
12       Lebowitz, M. D.; Quackenboss, J. J.; Krzyzanowski, M. (1991) Acute respiratory effects of prolonged ambient
13             ozone. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment:
14             papers from an international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste
15             Management Association; pp.  111-119. (A&WMA transactions series no. TR-19).
16       Lee, J. -T.; Schwartz, J. (1999) Reanaly sis of the effects of air pollution on daily mortality in Seoul, Korea:
17             a case-crossover design. Environ.  Health Perspect. 107: 633-636.
18       Lee, J.-T.; Shin, D.; Chung, Y. (1999) Air pollution and daily mortality in Seoul and Ulsan, Korea. Environ. Health
19             Perspect. 107:  149-154.
20       Lee, K.; Parkhurst, W. J.; Xue, J.; Ozkaynak, H.; Neuberg, D.; Spengler, J. D. (2004) Outdoor/indoor/presonal ozone
21             exposures of children in Nashville, Tennessee. J. Air Waste Manage. Assoc. 54: 352-359.
22       Liao, D.; Duan, Y.; Whitsel, E. A.; Zheng, Z.-J.; Heiss, G.; Chinchilli, V. M.; Lin, H.-M. (2004) Association of
23             higher levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based
24             study. Am. J. Epidemiol. 159: 768-777.
25       Liard, R.; Zureik, M.; Le Moullec, Y.; Soussan, D.;  Glorian, M.; Grimfeld, A.; Neukirch, F. (1999) Use of personal
26             passive samplers for measurement of NO2, NO, and O3 levels in panel studies. Environ. Res. 81: 339-348.
27       Lierl, M. B.; Hornung, R. W. (2003) Relationship of outdoor air quality to pediatric asthma exacerbations. Ann.
28             Allergy Asthma Immunol. 90:  28-33.
29       Lin, C. A.; Martins, M. A.; Farhat,  S. C.  L.; Pope, C. A., Ill; Conceicao, G. M. S.; Anastacio, V. M.; Hatanaka, M.;
30             Andrade, W. C.; Hamaue, W. R.;  Bohm, G. M.; Saldiva, P. H. N. (1999) Air pollution and respiratory illness
31             of children in Sao Paulo, Brazil. Paediatr. Perinat. Epidemiol. 13: 475-488.
32       Lin, M.; Chen, Y.; Burnett, R. T.; Villeneuve, P. J.; Krewski, D. (2003) Effect of short-term exposure to gaseous
3 3             pollution on asthma hospitalisation in children: a bi-directional case-crossover analysis. J. Epidemiol.
34             Community Health 57: 50-55.
35       Lin, M.; Chen, Y.; Villeneuve, P. J.; Burnett, R. T.; Lemyre, L.; Hertzman, C.; McGrail, K. M.; Krewski, D. (2004)
3 6             Gaseous air pollutants and asthma hospitalization of children with low household income in Vancouver,
37             British Columbia, Canada. Am. J. Epidemiol.  159: 294-303.
38       Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D.; Gong, H., Jr. (1996)
39             Short-term air pollution exposures and responses in Los Angeles area schoolchildren. J. Exposure Anal.
40             Environ. Epidemiol. 6: 449-472.
41       Linn, W. S.; Szlachcic, Y.; Gong, H., Jr.; Kinney, P. L.; Berhane, K. T. (2000) Air pollution and daily hospital
42             admissions in metropolitan Los Angeles. Environ. Health Perspect. 108: 427-434.
43       Lipfert, F. W.; Hammerstrom, T. (1992)  Temporal patterns in air pollution and hospital admissions. Environ. Res.
44             59: 374-399.
45       Lipfert, F. W.; Morris, S. C.; Wyzga, R.  E. (2000a) Daily mortality in the Philadelphia metropolitan area and
46             size-classified paniculate matter. J. Air Waste Manage. Assoc. 50: 1501-1513.
47       Lipfert, F. W.; Perry, H. M., Jr.; Miller, J. P.; Baty, J. D.; Wyzga, R. E.; Carmody, S. E. (2000b) The Washington
48             University-EPRI veterans' cohort  mortality study: preliminary results. In: Grant, L. D., ed. PM2000:
49             paniculate matter and health. Inhalation Toxicol.  12(suppl. 4): 41-73.
50       Lipfert, F. W.; Perry, H. M., Jr.; Miller, J. P.; Baty, J. D.; Wyzga, R. E.; Carmody, S. E. (2003) Air pollution, blood
51             pressure, and their long-term associations with mortality. Inhalation Toxicol. 15: 493-512.
52       Lippmann, M. (1988) Health significance of pulmonary function responses to airborne irritants. JAPCA
53             38:881-887.
54       Lippmann, M.; Ito, K.; Nadas, A.; Burnett, R.  T. (2000) Association of paniculate matter components with daily
55             mortality and morbidity in urban populations. Cambridge, MA: Health Effects Institute; research report
56             no. 95.
         January 2005                                   7-141         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Lipsett, M; Hurley, S.; Ostro, B. (1997) Air pollution and emergency room visits for asthma in Santa Clara County,
  2             California. Environ. Health Perspect. 105: 216-222.
  3       Liu, L.-J. S.; Koutrakis, P.; Leech, I; Broder, I. (1995) Assessment of ozone exposures in the greater metropolitan
  4             Toronto area. J. Air Waste Manage. Assoc. 45: 223-234.
  5       Liu, L.-J. S.; Delfino, R.; Koutrakis, P. (1997) Ozone exposure assessment in a southern California community.
  6             Environ. Health Perspect. 105: 58-65.
  7       Loomis, D. P.; Borja-Aburto, V. H.; Bangdiwala, S. I.; Shy, C. M. (1996) Ozone exposure and daily mortality in
  8             Mexico City: a time-series analysis. Cambridge, MA: Health Effects Institute; research report no. 75.
  9       Luginaah, I. N.; Fung, K. Y.; Gorey, K. M.; Webster, G.;  Wills, C. (2004) Association of Ambient Air Pollution
10             with Respiratory  Hospitalization in a Government Designated 'Area of Concern': The Case of Windsor,
11             Ontario. Environ. Health Perspect.  10.1289/ehp.7300. Available: http://dx.doi.org/ (14 December 2004).
12       Lumley, T.; Sheppard, L. (2000) Assessing seasonal confounding and model selection bias in air pollution
13             epidemiology using positive and negative control analyses. Environmetrics 11: 705-717.
14       Mann, J. K.; Tager, I. B.; Lurmann, F.; Segal, M.; Quesenberry, C. P., Jr.; Lugg, M. M.; Shan, J.; Van denEeden,
15             S. K. (2002) Air pollution and hospital admissions  for ischemic heart disease in persons with congestive heart
16             failure or arrhythmia. Environ. Health Perspect. 110: 1247-1252.
17       McConnell, R.; Berhane, K.; Gilliland, F.; London,  S. J.; Vora, H.; Avol, E.; Gauderman, W. J.; Margolis, H. G.;
18             Lurmann, F.; Thomas, D. C.; Peters, J. M. (1999) Air pollution and bronchitic symptoms in southern
19             California children with asthma. Environ. Health Perspect. 107: 757-760.
20       McConnell, R.; Berhane, K.; Gilliland, F.; London,  S. J.; Islam, T.; Gauderman, W. J.; Avol, E.; Margolis, H. G.;
21             Peters, J. M. (2002) Asthma in exercising children  exposed to  ozone: a cohort study. Lancet 359: 386-391.
22       McDonnell, W. F.; Abbey, D. E.; Nishino, N.; Lebowitz, M. D. (1999) Long-term ambient ozone concentration and
23             the incidence of asthma in nonsmoking adults: the ahsmog study. Environ. Res. 80: 110-121.
24       Metzger, K. B.; Tolbert, P. E.; Klein, M.;  Peel, J. L.; Flanders, W. D.; Todd, K. H.; Mulholland, J. A.; Ryan, P. B.;
25             Frumkin, H. (2004) Ambient air pollution and cardiovascular emergency department visits. Epidemiology
26             15:46-56.
27       Mitchell, H.; Senturia, Y.; Gergen, P.; Baker, D.; Joseph,  C.; McNiff-Mortimer, K.; Wedner, H.  J.; Crain, E.;
28             Eggleston, P.; Evans, R., Ill; Kattan, M.; Kercsmar, C.; Leickly, F.; Malveaux, F.; Smartt, E.; Weiss, K.
29             (1997) Design  and methods of the National Cooperative Inner-City Asthma Study. Pediatr. Pulmonol.
30             24:237-252.
31       Moolgavkar, S. H.; Luebeck, E. G. (1996) A critical review of the evidence on paniculate air pollution and
32             mortality. Epidemiology  7: 420-428.
33       Moolgavkar, S. H.; Luebeck, E. G.; Hall,  T. A.; Anderson, E. L.  (1995) Air pollution and daily mortality in
34             Philadelphia. Epidemiology 6: 476-484.
35       Moolgavkar, S. H.; Luebeck, E. G.; Anderson, E. L. (1997) Air pollution and hospital admissions for respiratory
36             causes in Minneapolis-St. Paul and Birmingham. Epidemiology 8: 364-370.
37       Morgan, G.; Corbett,  S.; Wlodarczyk, J. (1998a) Air pollution and hospital admissions in Sydney, Australia, 1990 to
38             1994. Am. J. Public Health 88: 1761-1766.
39       Morgan, G.; Corbett,  S.; Wlodarczyk, J.; Lewis, P. (1998b) Air pollution and daily mortality in Sydney, Australia,
40             1989 through 1993. Am.  J. Public Health 88:  759-764.
41       Mortimer, K. M.; Tager, I. B.; Dockery, D. W.; Neas, L. M.; Redline, S. (2000) The effect of ozone on inner-city
42             children with asthma: identification of susceptible subgroups. Am. J. Respir. Crit. Care Med. 162: 1838-1845.
43       Mortimer, K. M.; Neas,  L. M.; Dockery, D. W.; Redline, S.; Tager, I. B. (2002) The effect of air pollution on
44             inner-city children with asthma. Eur. Respir.  J. 19:  699-705.
45       Naeher, L. P.; Holford, T. R.; Beckett, W. S.; Belanger, K.; Triche, E. W.; Bracken, M. B.; Leaderer, B. P. (1999)
46             Healthy women's PEF variations with ambient summer concentrations of PM10, PN2 5, SO42", H+, and O3.
47             Am. J. Respir.  Crit. Care Med. 160: 117-125.
48       Nauenberg, E.; Basu,  K. (1999) Effect of insurance coverage on the relationship between asthma hospitalizations
49             and exposure to air pollution. Public Health Rep. 114: 135-148.
50       Navidi, W.; Thomas, D.; Langholz, B.; Stram, D. (1999) Statistical methods for epidemiologic studies of the health
51             effects of air pollution. Cambridge, MA: Health Effects Institute; research report no. 86.
52       Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud, D.  J.; Speizer, F. E. (1995) The association of ambient air
53             pollution with twice daily peak expiratory flow rate measurements in children. Am. J. Epidemiol.
54             141: 111-122.
55       Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Speizer, F. E. (1999) Fine particles and peak flow in children: acidity
56             versus mass. Epidemiology 10: 550-553.
         January 2005                                   7-142         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Neidell, M. J. (2004) Air pollution, health, and socio-economic status: the effect of outdoor air quality on childhood
  2             asthma. J. Health Econ. 23: 1209-1236.
  3       Newhouse, C. P.; Levetin, B. S.; Levetin, E. (2004) Correlation of environmental factors with asthma and rhinitis
  4             symptoms in Tulsa, OK. Ann. Allergy Asthma Immunol. 92: 356-366.
  5       Nutman, A.; Solomon, Y.; Mendel, S.; Nutman, J.; Hines, E.; Topilsky, M.; Kivity, S. (1998) The use of a neural
  6             network for studying the relationship between air pollution and asthma-related emergency room visits. Respir.
  7             Med. 92: 1199-1202.
  8       Oftedal, B.; Nafstad, P.; Magnus, P.; Bjerkly, S.; Skrondal, A. (2003) Traffic related air pollution and acute hospital
  9             admission for respiratory diseases inDrammen, Norway 1995-2000. Eur. J. Epidemiol. 18: 671-675.
10       O'Neill, M. S.; Ramirez-Aguilar, M.; Meneses-Gonzalez, F.; Hernandez-Avila, M.; Geyh, A. S.; Sienra-Monge, J. J.;
11             Romieu, I. (2003) Ozone exposure among Mexico City outdoor workers. J. Air Waste Manage. Assoc.
12             53: 339-346.
13       O'Neill, M. S.; Loomis, D.; Borja-Aburto, V. H. (2004) Ozone, area social conditions, and mortality in Mexico City.
14             Environ. Res. 94: 234-242.
15       Ostro, B. (1995) Fine paniculate air pollution and mortality in two Southern California counties. Environ. Res.
16             70:98-104.
17       Ostro, B.; Sanchez, J. M.; Aranda, C.; Eskeland, G. S. (1996) Air pollution and mortality: results from a study of
18             Santiago, Chile. In: Lippmann, M., ed. Papers from the ISEA-ISEE annual meeting; September 1994;
19             Research Triangle Park, NC.  J. Exposure Anal. Environ. Epidemiol. 6: 97-114.
20       Ostro, B.; Lipsett, M.; Mann, J.; Braxton-Owens, H.; White, M. (2001) Air pollution and exacerbation of asthma in
21             African-American children in Los Angeles. Epidemiology 12: 200-208.
22       Palli, D.; Saieva, C.; Grechi, D.; Masala, G.; Zanna, I.; Barbara, A.; Decarli, A.; Munnia, A.; Peluso, M. (2004)
23             DNA bulky adducts in a Mediterranean population correlate with environmental ozone concentration, an
24             indicator of photochemical smog. Int. J. Cancer 109: 17-23.
25       Park, S. K.; O'Neill, M. S.; Vokonas, P. S.; Sparrow, D.; Schwartz, J. (2004) Effects of air pollution on heart rate
26             variability: the VA normative aging study. Environ. Health Perspect: in press, 10.1289/ehp.7447. Available:
27             http://dx.doi.org/ [6 December, 2004].
28       Pearce, N.; Beasley, R.; Burgess, C.; Crane, J. (1998) Asthma epidemiology: principles and methods. New York,
29             NY: Oxford University Press.
30       Peel, J. L.; Tolbert, P. E.; Klein, M.; Metzger, K. B.; Planers, W. D.; Knox, T.; Mulholland, J. A.; Ryan, P. B.;
31             Frumkin, H. (2004) Ambient air pollution and respiratory emergency department visits. Epidemiology: in
32             press.
33       Pereira, L. A. A.; Loomis, D.; Conceicao, G.  M. S.; Braga, A. L. F.; Areas, R. M.; Kishi, H. S.; Singer, J. M.; Bohm,
34             G. M.; Saldiva, P. H. N. (1998)  Association between air pollution and intrauterine mortality in Sao Paulo,
35             Brazil. Environ. Health Perspect. 106:  325-329.
36       Peters, J. M.; Avol, E.; Navidi, W.; London, S. J.; Gauderman, W. J.; Lurmann, F.; Linn, W. S.; Margolis, H.;
37             Rappaport, E.; Gong, H., Jr.;  Thomas,  D. C. (1999a) A study of twelve southern California communities with
3 8             differing levels and types of air pollution. I. Prevalence of respiratory morbidity. Am. J. Respir. Crit. Care
39             Med. 159: 760-767.
40       Peters, J. M.; Avol, E.; Gauderman,  W. J.; Linn, W. S.; Navidi, W.; London, S. J.; Margolis, H.; Rappaport, E.;
41             Vora, H.; Gong, H.,  Jr.; Thomas, D. C. (1999b) A study of twelve southern California communities with
42             differing levels and types of air pollution. II. Effects on pulmonary function. Am. J. Respir. Crit. Care Med.
43             159:768-775.
44       Peters, A.; Liu, E.; Verrier, R. L.; Schwartz, J.; Gold, D. R.; Mittleman, M.; Baliff, J.; Oh, J. A.; Allen, G.;
45             Monahan, K.; Dockery, D. W. (2000a) Air pollution and incidence of cardiac arrhythmia. Epidemiology
46             11: 11-17.
47       Peters, A.; Skorkovsky, J.; Kotesovec,  F.; Brynda, J.; Spix, C.; Wichmann, H. E.; Heinrich, J. (2000b) Associations
48             between mortality and air pollution in  central Europe. Environ. Health Perspect. 108: 283-287.
49       Peters, A.; Dockery, D. W.; Muller,  J. E.; Mittleman, M. A. (2001) Increased paniculate air pollution and the
50             triggering of myocardial infarction. Circulation 103: 2810-2815.
51       Petroeschevsky, A.; Simpson, R. W.; Thalib,  L.; Rutherford, S. (2001) Associations between outdoor air pollution
52             and hospital admissions in Brisbane, Australia. Arch. Environ. Health 56: 37-52.
53       Ponce de Leon, A.; Anderson, H. R.; Bland, J. M.; Strachan, D. P.; Bower, J. (1996) Effects of air pollution on daily
54             hospital admissions for respiratory disease in London between 1987-88 and 1991-92. In: StLeger, S., ed.
5 5             The APHEA project. Short term effects of air pollution on health: a European approach using epidemiological
56             time series data. J. Epidemiol. Commun. Health 50(suppl. 1): S63-S70.
         January 2005                                   7-143         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Ponka, A.; Virtanen, M. (1996) Asthma and ambient air pollution in Helsinki. In: St Leger, S., ed. The APHEA
  2             project. Short term effects of air pollution on health: a European approach using epidemiological time series
  3             data. J. Epidemiol. Community Health 50(suppl. 1): S59-S62.
  4       Ponka, A.; Savela, M.; Virtanen, M. (1998) Mortality and air pollution in Helsinki. Arch. Environ. Health
  5             53:281-286.
  6       Pope, C. A., Ill; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. (2002) Lung cancer,
  7             cardiopulmonary mortality, and long-term exposure to fine paniculate air pollution. JAMA J. Am. Med.
  8             Assoc. 287: 1132-1141.
  9       Prescott, G. J.; Cohen, G. R.; Elton, R. A.;  Fowkes, F. G. R.; Agius, R. M. (1998) Urban air pollution and
10             cardiopulmonary ill health: a 14.5 year time series study. Occup. Environ. Med. 55: 697-704.
11       Raizenne, M.; Stern, B.; Burnett, R.;  Spengler, J. (1987) Acute respiratory function and transported air pollutants:
12             observational studies. Presented at: 80th annual meeting of the  Air Pollution Control Association; June;
13             New York, NY. Pittsburgh, PA: Air Pollution Control Association; paper no. 87-32.6.
14       Raizenne, M. E.; Burnett, R. T.; Stern, B.; Franklin, C. A.; Spengler, J. D. (1989) Acute lung function responses to
15             ambient acid aerosol exposures in children. Environ. Health Perspect. 79: 179-185.
16       Ramadour, M.; Burel, C.; Lanteaume, A.; Vervloet, D.; Charpin, D.; Brisse, F.; Dutau, H.; Charpin, D. (2000)
17             Prevalence of asthma and rhinitis in relation to long-term exposure to gaseous air pollutants. Allergy
18             (Copenhagen) 55: 1163-1169.
19       Ramsay, T. O.; Burnett, R. T.; Krewski, D. (2003) The effect of concurvity in generalized additive models linking
20             mortality to ambient paniculate matter. Epidemiology 14: 18-23.
21       Rich, K. E.; Petkau, J.; Vedal,  S.;Brauer,M. (2004) A case-crossover analysis of paniculate air pollution and
22             cardiac arrhythmia in patients with implantable cardioverter defibrillators. Inhalation Toxicol. 16: 363-372.
23       Ritz, B.; Yu, F.; Chapa, G.; Fruin, S.  (2000) Effect of air pollution on  preterm birth among children born in Southern
24             California between 1989 and 1993. Epidemiology 11: 502-511.
25       Roemer, W. H.; Van Wijnen, J. H . (2001)  Daily mortality and air pollution along busy streets in Amsterdam,
26             1987-1998. Epidemiology 12:  649-653.
27       Romieu, I.; Meneses, F.; Sienra-Monge, J.  J. L.;  Huerta, J.; Velasco, S. R.; White, M. C.; Etzel, R. A.;
28             Hernandez-Avila, M.  (1995) Effects of urban air pollutants on emergency visits for childhood asthma in
29             Mexico City. Am. J. Epidemiol. 141: 546-553.
30       Romieu, L; Meneses, F.; Ruiz, S.; Sienra, J. J.; Huerta, J.; White, M. C.; Etzel, R. A. (1996) Effects of air pollution
31             on the respiratory health of asthmatic children living in Mexico City. Am. J. Respir. Crit. Care Med.
32             154:300-307.
33       Romieu, L; Meneses, F.; Ruiz, S.; Huerta, J.; Sienra, J. J.; White, M.;  Etzel, R.; Hernandez, M. (1997) Effects of
34             intermittent ozone exposure on peak expiratory flow and respiratory  symptoms among asthmatic children in
35             Mexico City. Arch. Environ. Health 52: 368-376.
36       Romieu, L; Meneses, F.; Ramirez, M.; Ruiz, S.; Padilla, R. P.; Sienra, J. J.; Gerber, M.; Grievink, L.; Dekker, R.;
37             Walda, L; Brunekreef, B. (1998) Antioxidant supplementation and respiratory functions among workers
3 8             exposed to high levels of ozone. Am. J. Respir. Crit. Care Med. 158: 226-232.
39       Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Tellez-Rojo, M. M.; Moreno-Macias, H.; Reyes-Ruiz, N. L;
40             Del Rio-Navarro, B. E.; Ruiz-Navarro, M. X.; Hatch, G.; Slade, R.; Hernandez-Avila, M. (2002) Antioxidant
41             supplementation and lung functions  among children with asthma exposed to high levels of air pollutants. Am.
42             J. Respir. Crit. Care Med. 166: 703-709.
43       Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del
44             Rio-Navarro, B.; Hernandez-Avila, M.; London, S. J. (2004) Genetic polymorphism of GSTM1 and
45             antioxidant supplementation influence lung function in relation to ozone exposure in asthmatic children in
46             Mexico City. Thorax 59: 8-10.
47       Ross, M. A.; Persky, V. W.;  Scheff, P. A.;  Chung, J.; Curtis, L.; Ramakrishnan, V.; Wadden, R. A.; Hryhorczuk,
48             D. O. (2002) Effect of ozone and aeroallergens on the respiratory health of asthmatics. Arch. Environ. Health
49             57: 568-578.
50       Rothman, K. J.; Greenland, S., eds. (1998)  Modern epidemiology. 2nd ed. Philadelphia, PA: Lippincott-Raven
51             Publishers.
52       Saez, M.; Tobias, A.; Munoz, P.; Campbell, M. J. (1999) A GEE moving average analysis of the relationship
53             between air pollution and mortality for asthma in Barcelona, Spain. Stat. Med.  18: 2077-2086.
54       Saldiva, P. H. N.; Lichtenfels,  A. J. F. C.; Paiva, P. S.  O.; Barone, I. A.; Martins, M. A.; Massad, E.; Pereira, J.  C.
55             R.; Xavier, V. P.;  Singer, J. M.; Bohm, G. M. (1994) Association between air pollution and mortality due to
56             respiratory diseases in children in Sao Paulo, Brazil: a preliminary report. Environ. Res. 65: 218-225.
         January 2005                                   7-144         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Saldiva, P. H. N.; Pope, C. A., Ill; Schwartz, I; Dockery, D. W.; Lichtenfels, A. I; Salge, J. M; Barone, I.; Bohm,
  2             G. M. (1995) Air pollution and mortality in elderly people: a time-series study in Sao Paulo, Brazil. Arch.
  3             Environ. Health 50: 159-163.
  4       Samet, J. M.; Zeger, S. L.; Dominici, F.; Curriero, F.; Coursac, I.; Dockery, D. W.; Schwartz, J.; Zanobetti, A.
  5             (2000) The national morbidity, mortality, and air pollution study. Part II: morbidity, mortality, and air
  6             pollution in the United States. Cambridge,  MA: Health Effects Institute; research report no. 94, part II.
  7       Sarnat, J. A.; Schwartz, J.; Catalano, P. J.; Suh, H. H. (2001) Gaseous pollutants in paniculate matter epidemiology:
  8             confounders or surrogates? Environ. Health Perspect.  109: 1053-1061.
  9       Sartor, F.; Snacken, R.; Demuth, C.; Walckiers, D. (1995) Temperature, ambient ozone levels, and mortality during
10             summer, 1994, in Belgium. Environ. Res. 70: 105-113.
11       Scarlett, J. F.; Abbott, K. J.; Peacock, J. L.; Strachan, D. P.; Anderson, H. R. (1996) Acute effects of summer air
12             pollution on respiratory function in primary school children in southern England. Thorax 51:1109-1114.
13       Schouten, J. P.; Vonk, J. M.; de Graaf, A. (1996)  Short term effects of air pollution on emergency hospital
14             admissions for respiratory disease: results of the APHEA project in two major cities in The Netherlands,
15             1977-89. In: St Leger, S., ed. The APHEA  project. Short term effects of air pollution on health: a European
16             approach using epidemiological time series data. J. Epidemiol. Community Health 50(suppl. 1): S22-S29.
17       Schwartz, J. (1991) Paniculate air pollution and daily mortality in Detroit. Environ. Res. 56: 204-213.
18       Schwartz, J. (1993) Air pollution and daily mortality in Birmingham, Alabama. Am. J. Epidemiol. 137: 1136-1147.
19       Schwartz, J. (2004) How sensitive is the association between ozone and daily deaths to control for temperature?
20             Am.  J. Respir. Crit. Care Med.: in press, 10.1164/rccm.200407-933OC.
21       Schwartz, J.; Spix, C.; Touloumi, G.; Bacharova,  L.; Barumamdzadeh, T.; le Tertre, A.; Piekarksi, T.; Ponce de
22             Leon, A.; Ponka, A.; Rossi, G.; Saez, M.; Schouten, J. P. (1996) Methodological issues in studies of air
23             pollution and daily counts of deaths or hospital admissions.  In: St Leger, S., ed. The APHEA project. Short
24             term effects of air pollution on health: a European approach using epidemiological time series data.
25             J. Epidemiol. Commun. Health 50(suppl. 1): S3-S11.
26       Selwyn, B.  J.; Stock, T. H.; Hardy, R. J.; Chan, F. A.; Jenkins, D. E.; Kotchmar, D. J.; Chapman, R. S. (1985)
27             Health effects of ambient ozone exposure in vigorously exercising adults. In: Lee, S. D., ed. Evaluation of the
28             scientific basis for ozone/oxidants standards: proceedings of an APCA international specialty conference;
29             November 1984; Houston, TX. Pittsburgh,  PA: Air Pollution Control Association; pp. 281-296. (APCA
30             international specialty conference transactions: TR-4).
31       Sheppard, L. (2003) Ambient air pollution and nonelderly asthma hospital admissions in Seattle, Washington,
32             1987-1994. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston,
33             MA: Health Effects Institute; pp. 227-230.  Available:  http://www.healtheffects.org/Pubs/TimeSeries.pdf
34             [18 October, 2004].
35       Sheppard, L.; Levy, D.; Norris, G.; Larson, T. V.; Koenig, J. Q. (1999) Effects of ambient air pollution on
36             nonelderly asthma hospital admissions in Seattle, Washington, 1987-1994. Epidemiology 10: 23-30.
37       Sherwin, R. P.; Richters, V.; Kraft, P.; Richters, A. (2000) Centriacinar region inflammatory disease in young
38             individuals: a comparative study of Miami  and Los Angeles residents. Virchows Arch. 437: 422-428.
39       Shumway, R. H.; Azari, A. S.; Pawitan, Y. (1988) Modeling mortality fluctuations in Los Angeles as functions of
40             pollution and weather effects. Environ. Res. 45: 224-241.
41       Silverman, B. W. (1986) Density estimation. London, England: Chapman & Hall.
42       Simpson, R. W.; Williams, G.; Petroeschevsky, A.; Morgan, G.; Rutherford, S. (1997) Associations between outdoor
43             air pollution and daily mortality in Brisbane, Australia. Arch. Environ. Health 52: 442-454.
44       Spektor, D. M.; Lippmann, M.; Lioy, P. J.; Thurston, G. D.; Citak, K.; James, D.  J.; Bock, N.; Speizer, F. E.;
45             Hayes, C. (1988a) Effects of ambient ozone on respiratory function in active, normal children. Am. Rev.
46             Respir. Dis. 137:313-320.
47       Spektor, D. M.; Lippmann, M.; Thurston, G. D.; Lioy, P. J.; Stecko, J.; O'Connor, G.; Garshick, E.; Speizer, F. E.;
48             Hayes, C. (1988b) Effects of ambient ozone on respiratory function in healthy adults exercising outdoors.
49             Am.  Rev. Respir. Dis. 138: 821-828.
50       Spektor, D. M.; Lippmann, M. (1991) Health effects of ambient ozone on healthy children at a summer camp. In:
51             Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
52             international conference; March 1990; Los Angeles, CA. Pittsburgh,  PA: Air & Waste Management
53             Association; pp. 83-89.  (A&WMA transactions series no. TR-19).
54       Stieb, D. M.; Burnett, R. T.; Beveridge, R. C.; Brook, J. R. (1996)  Association between ozone and asthma
55             emergency department visits in Saint John, New Brunswick, Canada. Environ. Health Perspect.
56             104:  1354-1360.
         January 2005                                   7-145         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Stieb, D. M; Judek, S.; Burnett, R. T. (2002) Meta-analysis of time-series studies of air pollution and mortality:
  2             effects of gases and particles and the influence of cause of death, age, and season. J. Air Waste Manage.
  3             Assoc. 52: 470-484.
  4       Stieb, D. M.; Judek, S.; Burnett, R. T. (2003) Meta-analysis of time-series studies of air pollution and mortality:
  5             update in relation to the use of generalized additive models. J. Air Waste Manage. 53:  258-261.
  6       Sunyer, J.; Basagana, X. (2001) Particles, and not gases, are associated with the risk of death in patients with chronic
  7             obstructive pulmonary disease. Int. J. Epidemiol. 30: 1138-1140.
  8       Sunyer, J.; Castellsague, J.; Saez, M.; Tobias, A.; Anto, J. M. (1996) Air pollution and mortality in Barcelona. In:
  9             St Leger, S., ed. The APHEA project. Short term effects of air pollution on health: a European approach using
10             epidemiological time series data. J. Epidemiol. Community Health 50(suppl. 1): S76-S80.
11       Sunyer, J.; Basagana, X.; Belmonte, J.; Anto, J. M. (2002) Effect of nitrogen dioxide and ozone on the risk of dying
12             in patients with severe asthma. Thorax 57: 687-693.
13       Superko, H. R.; Adams, W. C.; Daly, P. W. (1984) Effects of ozone inhalation during  exercise in selected patients
14             with heart disease. Am. J. Med. 77: 463-470.
15       Tager, I. B. (1993) Introduction to working group on tropospheric ozone, Health Effects Institute Environmental
16             Epidemiology Planning Project. Environ. Health Perspect. 101(suppl. 4): 205-207.
17       Tager, I. B. (1999) Air pollution and lung function growth. Is it ozone? [editorial]. Am. J. Respir. Crit. Care Med.
18             160:387-389.
19       Tager, I. B.; Kunzli, N.; Lurmann, F.; Ngo, L.; Segal, M.; Balmes, J. (1998) Methods development for
20             epidemiologic investigations of the health effects of prolonged ozone exposure. Part II: an approach to
21             retrospective estimation of lifetime ozone exposure using  a questionnaire and ambient monitoring data
22             (California sites). Cambridge, MA: Health Effects Institute; research report no.  81; pp. 27-78.
23       Taggart, S. C. O.; Custovic, A.; Francis, H. C.; Faragher, E. B.; Yates, C. J.; Higgins, B. G.; Woodcock, A. (1996)
24             Asthmatic bronchial hyperresponsiveness varies with ambient levels of summertime air pollution. Eur.
25             Respir. J. 9: 1146-1154.
26       Tellez-Rojo, M. M.; Romieu, L; Ruiz-Velasco, S.; Lezana, M.-A.; Hernandez-Avila, M. M. (2000) Daily respiratory
27             mortality and PMio pollution in Mexico City: importance of considering place of death. Eur. Respir. J.
28             16:391-396.
29       Tenias, J. M.; Ballester, F.; Rivera, M. L. (1998) Association between hospital emergency visits for asthma and air
30             pollution in Valencia, Spain. Occup. Environ. Med.  55: 541-547.
31       Tenias, J. M.; Ballester, F.; Perez-Hoyos, S.; Rivera, M. L. (2002) Air pollution and hospital emergency room
32             admissions for chronic obstructive pulmonary disease in Valencia, Spain. Arch. Environ. Health 57: 41-47.
33       Thompson, A. J.; Shields, M. D.; Patterson, C. C. (2001) Acute asthma exacerbations and air pollutants in children
34             living in Belfast, Northern Ireland. Arch. Environ. Health 56: 234-241.
35       Thurston, G. D.; Ito, K. (2001) Epidemiological studies of acute  ozone exposures and  mortality.  J. Exposure Anal.
36             Environ. Epidemiol.  11: 286-294.
37       Thurston, G. D.; Ito, K.; Kinney, P. L.; Lippmann, M.  (1992) A multi-year study of air pollution and respiratory
3 8             hospital admissions in three New York State metropolitan areas: results for 1988 and 1989 summers.
39             J. Exposure Anal. Environ. Epidemiol. 2: 429-450.
40       Thurston, G. D.; Ito, K.; Hayes, C. G.; Bates, D. V.; Lippmann, M. (1994) Respiratory hospital admissions and
41             summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols. Environ. Res.
42             65:271-290.
43       Thurston, G. D.; Lippmann, M.; Scott, M. B.; Fine, J. M. (1997)  Summertime  haze air pollution and children with
44             asthma.  Am. J. Respir. Crit. Care Med. 155: 654-660.
45       Tobias, A.; Campbell, M. J.; Saez, M. (1999) Modelling asthma epidemics on the relationship between air pollution
46             and asthma emergency visits in Barcelona, Spain. Eur. J. Epidemiol. 15: 799-803.
47       Tolbert, P. E.; Mulholland, J. A.; Macintosh, D. L.; Xu, F.; Daniels, D.; Devine, O. J.; Carlin, B. P.; Klein, M.;
48             Dorley, J.; Butler, A. J.; Nordenberg, D. F.; Frumkin, H.; Ryan, P. B.; White, M. C. (2000) Air quality and
49             pediatric emergency room visits for asthma in Atlanta, Georgia. Am. J.  Epidemiol.  151: 798-810.
50       Touloumi, G.; Katsouyanni, K.; Zmirou, D.; Schwartz, J.; Spix, C.; Ponce de Leon, A.; Tobias, A.; Quennel, P.;
51             Rabczenko, D.; Bacharova, L.; Bisanti, L.; Vonk, J.  M.; Ponka, A. (1997) Short-term effects of ambient
52             oxidant exposure on mortality: a combined analysis within the APHEA project. Am. J. Epidemiol.
53             146: 177-185.
54       Tsai, S.-S.; Huang, C.-H.; Goggins, W. B.; Wu, T.-N.; Yang, C.-Y. (2003) Relationship between air pollution and
55             daily mortality in a tropical city: Kaohsiung, Taiwan. J. Toxicol. Environ. Health Part A 66: 1341-1349.
56       Tyler, W. S.; Tyler, N. K.; Last, J. A.; Gillespie, M.  J.; Barstow,  T. J. (1988) Comparison of daily and seasonal
57             exposures of young monkeys to ozone. Toxicology 50: 131-144.


         January 2005                                   7-146        DRAFT-DO NOT QUOTE OR CITE

-------
  1       U.S. Environmental Protection Agency. (1996a) Air quality criteria for ozone and related photochemical oxidants.
  2             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
  3             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
  4             www.epa.gov/ncea/ozone.htm.
  5       U.S. Environmental Protection Agency. (1996b) Air quality criteria for paniculate matter. Research Triangle Park,
  6             NC: National Center for Environmental Assessment-RTF Office; report nos. EPA/600/P-95/001aF-cF. 3v.
  7       U.S. Environmental Protection Agency. (2004a) Air quality criteria for paniculate matter. Research Triangle Park,
  8             NC: National Center for Environmental Assessment; report no. EPA/600/P-99/002aF-bF. 2v. Available:
  9             http://cfpub.epa.gov/ncea/ [9 November, 2004].
10       U.S. Environmental Protection Agency. (2004b) The ozone report: measuring progress through 2003. Cincinnati,
11             OH: U.S. Environmental Protection Agency; report no. EPA454/K-04-001.
12       Ulmer, C.; Kopp, M; Ihorst, G.; Frischer, T.; Forster, I; Kuehr, J. (1997) Effects of ambient ozone exposures during
13             the spring and summer of 1994 on pulmonary function of schoolchildren. Pediatr. Pulmonol. 23: 344-353.
14       Vedal, S.; Brauer, M.; White, R.; Petkau, J. (2003) Air pollution and daily mortality in a city with low levels of
15             pollution. Environ. Health Perspect. 111:45-51.
16       Vedal, S.; Rich, K.; Brauer, M.; White, R.; Petkau, J.  (2004) Air pollution and cardiac arrhythmias in patients with
17             implantable cardiovascular defibrillators. Inhalation Toxicol. 16:  353-362.
18       Verhoeff, A. P.; Hoek, G.; Schwartz, J.; Van Wijnen,  J. H. (1996) Air pollution and daily mortality in Amsterdam.
19             Epidemiology 7: 225-230.
20       Virnig, B. A.; McBean, M. (2001) Administrative data for public health surveillance and planning. Annu. Rev.
21             Public Health 22: 213-230.
22       Weisel, C. P.; Cody, R. P.; Lioy, P. J. (1995) Relationship between summertime ambient ozone levels and
23             emergency department visits for asthma in central New Jersey. Environ. Health Perspect.
24             103(suppl. 2): 97-102.
25       Weisel, C. P.; Cody, R. P.; Georgopoulos, P. G.; Purushothaman, V; Weiss, S. H.; Bielory, L.; Gregory, P.; Stern,
26             A. H.  (2002) Concepts in developing  health-based indicators for ozone. Int. Arch. Occup. Environ. Health
27             75:415-422.
28       White, M. C.; Etzel, R. A.; Wilcox, W. D.; Lloyd, C.  (1994) Exacerbations of childhood asthma and ozone pollution
29             in Atlanta. Environ. Res. 65: 56-68.
30       Wong, T. W.; Lau, T. S.; Yu, T. S.; Neller, A.; Wong, S. L.; Tarn, W.; Pang, S. W. (1999a) Air pollution and
31             hospital admissions for respiratory and cardiovascular diseases in Hong Kong. Occup. Environ. Med.
32             56:679-683.
33       Wong, C.-M; Ma, S.; Hedley, A. J. Lam, T.-H. (1999b) Does ozone have any effect on daily hospital admissions for
34             circulatory diseases? J. Epidemiol. Community Health 53: 580-581.
3 5       World Health Organization. (2004) Meta-analysis of time-series studies and panel studies of paniculate matter (PM)
36             and ozone (O3): report of a WHO task group. Copenhagen, Denmark: WHO Regional Office for Europe;
37             document no. EUR/04/5042688. Available: http://www.euro.who.int/document/E82792.pdf
38             [18 November, 2004].
39       Yang, Q.; Chen, Y.; Shi, Y.; Burnett, R. T.; McGrail,  K. M.; Krewski, D. (2003) Association between ozone and
40             respiratory admissions among children and the elderly in Vancouver, Canada. Inhalation Toxicol.
41             15: 1297-1308.
42       Yang, C.-Y.; Chang, C.-C.; Chuang, H.-Y.; Tsai, S.-S.; Wu, T.-N.; Ho, C.-K. (2004) Relationship between air
43             pollution and daily mortality in a subtropical city: Taipei, Taiwan. Environ. Int. 30: 519-523.
44       Zeger, S. L.; Thomas, D.; Dominici, F.; Samet, J. M.;  Schwartz, J.; Dockery, D.; Cohen, A. (2000) Exposure
45             measurement error in time-series studies of air  pollution: concepts and consequences. Environ. Health
46             Perspect. 108: 419-426.
47       Zhu, L.; Carlin, B. P.; Gelfand, A. E. (2003) Hierarchical regression with misaligned spatial data: relating ambient
48             ozone and pediatric asthma ER visits  in Atlanta. Environmetrics 14: 537-557.
49       Zmirou, D.; Barumandzadeh, T.; Balducci, F.; Ritter,  P.; Laham, G.; Ghilardi, J.-P. (1996) Short term effects of air
50             pollution on mortality in the city of Lyon, France, 1985-90. In: St Leger, S., ed. The APHEA project.  Short
51             term effects of air pollution on health: a European approach using epidemiological time series data.
52             J. Epidemiol. Community Health 50(suppl. 1):  S30-S35.
53
         January 2005                                   7-147        DRAFT-DO NOT QUOTE OR CITE

-------
               CHAPTER 7 ANNEX
     EPIDEMIOLOGICAL STUDIES OF HUMAN
  HEALTH EFFECTS ASSOCIATED WITH AMBIENT
               OZONE EXPOSURE
January 2005                AX7-1    DRAFT-DO NOT QUOTE OR CITE

-------
                    Table AX7-1.  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study                                                  Copollutants
         Location and Period     Outcomes and Methods     Mean O3 Levels    Considered
                                                                  Findings, Interpretation
                                                                                   Effects
X
to
        United States

        Mortimer et al. (2002)
        Eight urban areas in
        the U.S.:
        St. Louis, MO;
        Chicago, IL;
        Detroit, MI;
        Cleveland,  OH;
        Washington, DC;
        Baltimore,  MD;
        East Harlem, NY;
        Bronx, NY
        Jun-Aug 1993
Examined 846 asthmatic
children aged 4-9 years for
O3 exposure effects on PEF
and morning symptoms
using linear mixed effect
models and GEE.
8-h avg O3
(10 a.m.-6 p.m.):
44ppb
PM10,N02,
S02
No associations were seen between single
or multiday O3 measures and any evening
outcome measure. The effects of O3 on
morning outcomes increased over several
days with the strongest associations seen
for multiday lags. Joint modeling of O3
with NO2 or SO2 resulted in slightly
reduced estimates for each pollutant.
8-h avg O3 (per 15 ppb):

% change in morning PEF:
Lag 1-5:
All areas: -0.59% (-1.05,-0.13)
St. Louis: -0.86% (-2.10, 0.38)
Chicago:  -0.62% (-2.41,1.16)
Detroit: -0.75% (-2.36, 0.86)
Cleveland: -0.62% (-2.23, 0.99)
Washington, DC: -0.54% (-2.02,
0.93)
Baltimore: 0.24% (-0.95, 1.43)
EastHarlem: -0.73% (-1.63, 0.17)
Bronx: -0.69% (-1.54, 0.15)

Odds ratios:
Morning symptoms:
Lag 1-4:
All areas: 1.16(1.02,1.30)
St. Louis: 0.82 (0.59, 1.14)
Chicago: 1.09(0.69,1.72)
Detroit: 1.72(1.12,2.64)
Cleveland: 1.20(0.81,1.79)
Washington, DC: 1.11 (0.72, 1.72)
Baltimore: 1.19(0.89,1.60)
EastHarlem: 1.22(0.97,1.53)
Bronx: 1.23(0.98,1.54)

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                             Copollutants
  Outcomes and Methods    Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                  Effects
X
United States (cont'd)

Mortimer et al. (2000)
Eight urban areas in the
U.S.:
St. Louis, MO;
Chicago, IL;
Detroit, MI;
Cleveland, OH;
Washington, DC;
Baltimore, MD;
East Harlem, NY;
Bronx, NY
Jun-Aug 1993
        Avoletal. (1998)
        Southern California
        communities
        Spring-summer 1994
                               A cohort of 846 asthmatic
                               children aged 4-9 years
                               examined for effects of
                               summer O3 exposure on
                               PEF and morning
                               symptoms.  Two subgroups
                               were compared: (l)low
                               birth weight or premature
                               and (2) normal birth weight
                               or full-term. Analysis using
                               GEE and linear mixed
                               models.
Three panels of children
(age 10-12 years):
(1) asthmatic (n= 53);
(2) wheezy (n = 54); and
(3) healthy (n = 103).
Examined for symptoms,
medication use, outdoor
time, physical activity, and
pulmonary function
measures in relation to O3
exposure, via logistic
regression and GLM.
                           8-h avg O3
                           (10 a.m.-6 p.m.):
                           48ppb
                 None
Stratified
analysis of low
and high 24-h
avg O3:

Fixed site O3:
Low: < 100 ppb
High: > 100 ppb

Personal O3:
Low: < 15.6 ppb
High: > 32.4 ppb
None
               Low birth weight and premature
               asthmatic children had greater declines in
               PEF and higher incidence of morning
               symptoms than normal birth weight and
               full-term asthmatic children.
                                                                                   The three groups responded similarly.
                                                                                   Few pulmonary function or symptom
                                                                                   associations. Asthmatic children had
                                                                                   the most trouble breathing, the most
                                                                                   wheezing, and the most inhaler use on
                                                                                   high O3 day sin the spring. Ozone levels
                                                                                   were considered too low during the
                                                                                   period of the study.  Noncompliance
                                                                                   by subjects may have been a problem.
                                                                                   Other analysis methods may have been
                                                                                   more appropriate.
8-h avg O3 (per 15 ppb):

% change in morning PEF:
Low birth weight:
Lag 1-5: -1.83%(-2.65,-1.01)
Normal birth weight:
Lag 1-5: -0.30% (-0.79, 0.19)
Interaction term for birth weight,
p<0.05

Odds ratios:
Morning symptoms:
Low birth weight:
Lag 1-4: 1.42(1.10,1.82)
Normal birth weight:
Lag 1-4: 1.09(0.95,1.24)
Interaction term for birth weight,
p<0.05

Multiple endpoints analyzed. Few
consistent or statistically significant
responses to O3 exposure reported.

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
           Reference, Study
         Location and Period
                                              Copollutants
  Outcomes and Methods     Mean O3 Levels     Considered
                                         Findings, Interpretation
                                                                    Effects
        United States (cont'd)
X
        Gillilandetal. (2001)
        12 Southern California
        communities
        Jan-Jun 1996
        Linn etal. (1996)
        Three towns in
        California: Rubidoux,
        Upland, Torrance
        Fall-spring 1992-1993
        and 1993-1994
1,933 4th grade children
(age 9-10 years) followed
for school absences.  Each
absence classified as illness-
related or not. Among
former, classified into
respiratory
or gastrointestinal.
Respiratory absences further
classified into upper or
lower.  Pollution measured
in central site in each town.
Analysis of distributed lag
effects controlling for time,
day of week, and
temperature in a Poisson
model.

269 school children (age
unspecified), each followed
for moming/aftemoon lung
function and symptoms for
one week in fall, winter, and
spring over 2 school years.
Personal exposure
monitoring in a subset.
Analyzed afternoon
symptoms versus same day
pollution and morning
symptoms versus 1-day
lag pollution.
8-h avg O3
(10 a.m.-6 p.m.):
Levels not
reported.
PM10,N02
24-h avg O3:

Personal:
5ppb
SD3

Central site:
23ppb
SD12
PM2,,N02
O3 strongly associated with illness-related
and respiratory absences. PM10 only
associated with upper respiratory
absences.  Long distributed lag effects
for O3 raise questions about adequacy
of control for seasonal changes.
Central site O3 correlated with personal
exposures, r= 0.61.  Ozone effects
observed on lung function but only
significant for FEV[ in one analysis.
No effects on symptoms. Ozone effects
were not robust to NO2 or PM2 5. Power
may have been limited by short followup
within seasons (limiting both person-days
and variability in exposures).
8-h avg O3 (per 20 ppb):

% change in absences:

All illness:
62.9% (18.4, 124.1)
Nonrespiratory illnesses:
37.3% (5.7, 78.3)
Respiratory illnesses:
82.9% (3.9, 222.0)
Upper respiratory:
45.1% (21.3, 73.7)
Lower respiratory with wet cough:
173.9% (91.3, 292.3)
Change in lung function (per ppb):

FEV[ next morning:
-0.26 mL (SE 0.25), p = 0.30
FEV; afternoon:
-0.18 mL(SE 0.26), p = 0.49
FEV[ crossday difference:
-0.58 mL(SE 0.23), p = 0.01

FVC next morning:
-0.21 mL (SE 0.22),p = 0.34
FVC afternoon:
-0.20 mL(SE 0.29), p = 0.48
FVC crossday difference:
-0.25 mL (SE 0.25), p = 0.32

-------
              Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                           Copollutants
Outcomes and Methods    Mean O3 Levels    Considered
Findings, Interpretation
                                                                                                                                   Effects
X
United States (cont'd)

Ostroetal. (2001)
Central Los Angeles
and Pasadena, CA
Aug-Octl993
                               138 African-American
                               children aged 8-13 years
                               with doctor diagnosed
                               asthma requiring
                               medication in past year
                               followed for daily
                               respiratory symptoms and
                               medication use. Lags of
                               0 to 3 days examined.
                         l-hmaxO3:       PM10,NO2,      Correlation between PM10 and O3 was
                                          pollen, mold    r = 0.35. Significant O3 effect seen for
                         Los Angeles:                     extra medication use (above normal use).
                         59.5 ppb                         No O3 effect on symptoms in expected
                         SD 31.4                         direction observed. Inverse association
                                                         seen for cough. PM10 effects seen at a lag
                         Pasadena:                        of 3 days.  Time factors not explicitly
                         95.8 ppb                         controlled in analysis; may have led to
                         SD 49.0                         confounding of O3 effects.
                               1-h max O3 (per 40 ppb):

                               Odds ratios:

                               Extra medication use:
                               Lagl:  1.15(1.12,1.19)

                               Respiratory symptoms:
                               Shortness of breath:
                               Lag 3:  1.01 (0.92,1.10)
                               Wheeze:
                               Lag 3:  0.94(0.88,1.00)
                               Cough:
                               Lag 3:  0.93(0.87,0.99)
Delfino et al. (2003)
Los Angeles, CA
Nov 1999- Jan 2000




Delfino etal. (1997a)
Alpine, CA
May-Augl994








A panel study of
22 Hispanic children with
asthma aged 10-16 years.
Filled out symptom diaries
in relation to pollutant
levels. Analysis using GEE
model.
22 asthmatics aged
9-46 years followed for
respiratory symptoms,
morning-afternoon PEF,
and P2 agonist inhaler use.
Personal O3 measured for
12 hours/day using passive
monitors. GLM mixed
model.


l-hmaxO3:
25.4 ppb
SD9.6




Ambient:
12-h avg O3
(8 a.m.-8 p.m.):
64 ppb
SD17

Personal:
12-h avg O3:
(8 a.m.-8 p.m.)
18 ppb
SD14
NO2, SO2, Support the view that air toxics in
CO, volatile the pollutant mix from traffic may
organic have adverse effects on asthma in
compounds, children.
PM10


PM10, pollen, No O3 effects observed.
fungi









1-h max O3 (per 14.0 ppb):

Odds ratio:
Symptoms interfering with
daily activities:
LagO: 1.99(1.06,3.72)

No quantitative results for O3.











-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
           Reference, Study
         Location and Period
                                                                     Copollutants
                         Outcomes and Methods     Mean O3 Levels    Considered
                                                                    Findings, Interpretation
                                                                                                                Effects
X
        United States (cont'd)

        Delfinoetal. (1998a)
        Alpine, CA
        Aug-Octl995
Delfino et al. (2004)
Alpine, CA
Aug-Oct 1999, Apr-Jun
2000
A panel of 24 asthmatics
aged 9-17 years followed
for daily symptoms.
Analysis using  GEE model.
1 9 asthmatic children (age
9- 1 7 years) followed daily
for 2 weeks to determine
relationship between air
pollutants, namely PM, and
FEVj. Linear mixed model
used for analysis.
                                                   l-hmaxO3:
                                                   90 ppb
                                                   SD 1 8
                            8-h max O3 :
                            62. 8 ppb
                            SD 15.1
                            IQR 22.0
                                             PM10
PM2 5, PM10,
NO2
Asthma symptoms were significantly
associated with both ambient O3 and
PM10 in single-pollutant models. Ozone
effects generally robust to PM10. Current
day O3 effects strongest in asthmatics not
on anti-inflammatory medication. Effects
of O3 and PM10 were largely independent.
The largest effects for PM10 were seen for
a 5-day distributed lag. For O3 effects,
there were no lag day effects; current day
results showed the greatest effect.

Significant declines in FEV; associated
with various PM indices (personal, indoor
home, etc.), but not ambient O3 levels.
                                                       1-h max O3 (per 58 ppb):

                                                       Odds ratios:
                                                       O3 only model:
                                                       LagO:  1.54(1.02,2.33)
                                                       O3 with PM10 model:
                                                       LagO:  1.46(0.93,2.29)
                                                                                                                                   No quantitative results for O3.
        Delfinoetal. (1996)
        San Diego, CA
        Sep-Octl993
                        12 well-characterized
                        moderate asthmatics aged
                        9-16 years (7 males,
                        5 females) followed over
                        6 weeks for medication use
                        and respiratory symptoms.
                        Allergy measured at
                        baseline with skin prick
                        tests. Personal O3
                        measured with passive
                        badge. Analysis with
                        GLM mixed model.
                           Ambient:          PM25, SO42~,     No effect of ambient O3 on symptom
                            1-h max O3:       FT, HNO3,       score. Personal O3 significant for
                           68 ppb            pollen, fungal    symptoms, but effect disappeared when
                           SD 30            spores           confounding day of week effect was
                                                             controlled with weekend dummy
                           Ambient:                          variable. (32 inhaler used among
                            12-h avg O3:                       7 subjects was significantly related to
                           43 ppb                            personal O3. Results of this small study
                           SD 17                            suggest the value of personal exposure
                                                             data in providing more accurate estimates
                           Personal:                          of exposures. However, nearly  50% of
                            12-h avg O3:                       personal O3 measurements were below
                            11.6 ppb                          limits of detection, diminishing  value of
                           SD 11.2                          these data.  Pollen and fine particulate
                                                             (low levels) were not associated with
                                                             any of the outcomes.
                                                                                                    Change in (32-agonist inhaler use
                                                                                                    (per ppb personal O3):
                                                                                                    0.0152 puffs/day (SE 0.0075),
                                                                                                    p = 0.04

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                             Copollutants
  Outcomes and Methods     Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                                                                                       Effects
X
United States (cont'd)

Chen et al. (2000)
Washoe County, NV
1996-1998
School absenteeism
examined among 27,793
students (kindergarten to
6th grade) from 57
elementary schools.
First-order autoregression
models used to assess
relationship between O3 and
school absenteeism after
adjusting for weather, day
of week, month, holidays,
and time trends.  Ozone
levels from the current day,
and cumulative lags of
1-14  days, 1-21 days, and
1-28  days examined.
                                                           l-hmaxO3:       PM10, CO       Multipollutant models were examined.
                                                           37.45 ppb                         Ozone concentrations in the preceding
                                                           SD 13.37                         14 days were significantly associated
                                                                                            with school absenteeism for students
                                                                                            in grades 1 through 6, but not those in
                                                                                            kindergarten. Both PM10 and CO
                                                                                            concentrations on the concurrent day
                                                                                            were associated with school absenteeism,
                                                                                            but the estimate for PM10 was a negative
                                                                                            value.
                                                                        1-h max O3 (per 50 ppb):

                                                                        Total absence rate:
                                                                        O3 with PM10 and CO model:
                                                                        Lag 1-14:  3.79% (1.04, 6.55)
        Newhouse et al. (2004)
        Tulsa, OK
        Sep-Oct2000
24 subjects aged 9-64 years
with physician diagnosis of
asthma.  Performed PEF
twice daily (morning and
afternoon), and reported
daily respiratory symptoms
and medication use.
Forward stepwise multiple
regression models and
Pearson correlation
analyses.
24-h avg O3:      PM2 5, CO,      Among ambient air pollutants, O3 seemed
30 ppb           SO2, pollen,     to be most significant factor.  Morning
Range 10-70      fungal spores    PEF values significantly associated with
                                 average and maximum O3 levels on the
                                 previous day.  Individual symptoms,
                                 including wheezing, headache, and
                                 fatigue, also significantly related to
                                 average and maximum daily O3. Multiple
                                 regression analyses produced complex
                                 models with different predictor variables
                                 for each symptom.
                                                                                                                           Pearson correlation coefficient:

                                                                                                                           Morning PEF:
                                                                                                                           Mean O3 levels:
                                                                                                                           Lagl: -0.274, p< 0.05
                                                                                                                           Maximum O3 levels:
                                                                                                                           Lagl: -0.2W,p<0.05

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                                                     Copollutants
                         Outcomes and Methods     Mean O3 Levels    Considered
                                                                   Findings, Interpretation
                                                                                    Effects
X
oo
United States (cont'd)

Ross et al. (2002)
East Moline, IL and
nearby communities
May-Octl994
Neasetal. (1995)
Uniontown, PA
Summer 1990
                                59 asthmatics aged 5-49
                                years recruited.  19 lost to
                                follow-up, yielding
                                study population of 40.
                                Assessment of PEF
                                and respiratory symptoms.
                                Analytical methods unclear
                                in terms of control for
                                time factors.
83 4th and 5th grade
children reported twice
daily PEF and the presence
of cold, cough, or wheeze.
Relationship to pollutants
was analyzed by an
autoregressive linear
regression model/GEE.
The number of hours each
child spent outdoors during
the preceding 12-h period
was evaluated.
                           8-hmaxO3:
                           41.5 ppb
                           SD 14.2
                           IQR20
                 PM10, S02,
                 NO2, pollen,
                 fungi
12-h avg O3:

Daytime
(8 a.m.-8p.m.):
50.0 ppb

Overnight
(8 p.m.-8 a.m.):
24.5 ppb
                                                                            SO2, PM10, FT
Saw significant associations between
O3 and both PEF declines and
symptom increases. Most but not all
effects remained after controlling for
temperature, pollen and fungi. The O3
effect on morning PEF disappeared after
adjusting for temperature. No PM10
effects observed.
Evening cough was associated with O3
levels weighted by hours spent outdoors
during the prior 12 hours. A decrease in
PEF was associated with O3 levels
weighted by hours spent outdoors. When
particle-strong acidity was added to the
model, the decrement was decreased and
no longer significant.
8-h max O3 (per 20 ppb):

Change in PEF (L/min):
Morning:
Lag 0-1:  -2.29 (-4.26,-0.33)
Afternoon:
LagO:  -2.58 (-4.26,-0.89)

Symptom score (on scale of 0-3):
Morning:
Lag 1-3:  0.08(0.03,0.13)
Afternoon:
Lag 1-3:  0.08(0.04,0.12)

12-h avg O3 (per 30 ppb increment
weighted by proportion of time
spent outdoors during prior 12
hours):

Evening PEF:
-2.79 L/min (-6.7,-1.1)

Odds ratio:
Evening cough:
2.20(1.02,4.75)
        Neasetal. (1999)
        Philadelphia, PA
        Jul-Sep 1993
                        156 children aged 6-11
                        years at two summer camps
                        followed for twice-daily
                        PEF.  Analysis using mixed
                        effects models adjusting for
                        autocorrelated errors.
                           Daytime
                           12-h avg O3
                           (9 a.m-9 p.m.):

                           SW camp:
                           57.5 ppb
                           IQR 19.8

                           NE camp:
                           55.9 ppb
                           IQR 21.9
                 FT, S042-,
                 PM2,,PM1
Some O3 effects detected as well as PM
effects. Similar O3-related decrements
observed in both morning and afternoon
PEF. Ozone effects not robust to SO42~ in
two-pollutant models, whereas SO42~
effects relatively robust to O3.
12-h avg O3 (per 20 ppb):

Morning and evening PEF:

O3 only models:
LagO:
-1.38 L/min (-2.81, 0.04)
Lag 1-5:
-2.58 L/min (-4.81,-0.35)

O3 with SO42~ model:
Lag not specified:
-0.04 L/min

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
  Outcomes and Methods     Mean O3 Levels
                  Copollutants
                   Considered
                       Findings, Interpretation
                                                   Effects
X
        United States (cont'd)

        Gent et al. (2003)
        Southern New England
        Apr-Sep2001
        Korricketal. (1998)
        Mount Washington,
        NH
        Summers 1991, 1992
271 children (age
<\2 years) with active,
doctor-diagnosed asthma
followed over 183 days for
respiratory symptoms. For
analysis, cohort split into
two groups: 1 30 who used
maintenance medication
during follow-up and
141 who did not, on
assumption that medication
users had more severe
asthma.  Logistic regression
analyses performed.

Evaluated the  acute effects
of ambient O3  on pulmonary
function of exercising
adults. 530 hikers (age
15-64 years) were
examined. Analysis using a
general linear  regression
model.
l-hmaxO3:
58.6 ppb
SD 19.0

8-hmaxO3:
51.3 ppb
SD15.5
PM25
Mean O3 per
hour of hiking:
40 ppb
Range 21-74
PM2 5, smoke,
acidity
Correlation between 1-h max O3 and
daily PM2 5 was 0.77 during this warm-
season study.  Large numbers of
statistical tests performed.  Significant
associations between symptoms and O3
seen only in medication users, a subgroup
considered to be more sensitive. PM2 5
significant for some symptoms, but not
in two-pollutant models. Ozone effects
generally robust to PM2 5.
With prolonged outdoor exercise low-
level exposures to O3 were associated
with significant effects on pulmonary
function.  Hikers with asthma had a
4-fold greater responsiveness to exposure
toO3.
1-h max O3 (per 50 ppb):

Odds ratios:
Regular medication users (n = 130):

Chest tightness:
O3 only model:
Lagl:  1.26(1.00,1.48)
O3 with PM2 5 model:
Lagl:  1.42(1.14,1.78)

Shortness of breath:
O3 only model:
Lagl:  1.22(1.02,1.45)

% change in lung function
(per 50 ppb O3):
                                                           j:  -2.6% (-4.7, -0.4)
                                                      FVC: -2.2% (-3. 5, -0.8)
        Thurston et al. (1997)
        Connecticut River
        Valley, CT
        June 1991, 1992, 1993
Children (age 7-13 years)
with moderate-to- severe
asthma followed for
medication use, lung
function, and medical
symptoms at a summer
asthma camp for one week
in 1991 (n = 52), 1992 (n =
58), and 1993 (n = 56).
Analysis was conducted
using both Poisson
modeling and GLM.
l-hmaxO3:

1991: 114.0 ppb
1992: 52.2 ppb
1993: 84.6 ppb

1991-1993: 83.6
ppb
FT, S042-
O3 was most consistently associated
with acute asthma exacerbation, chest
symptoms, and lung function decrements.
Pollen was poorly associated with any
adverse effect.  Consistent results were
obtained between the aggregate and
individual analyses.
1-h max O3 (per 83.6 ppb):

Relative risks:
P2-agonist use:  1.46, p< 0.05
Chest symptoms:  1.50, p< 0.05

Change in PEF (per ppb):
-0.096 L/min,p< 0.05

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
 Outcomes and Methods    Mean O3 Levels
 Copollutants
 Considered
       Findings, Interpretation
            Effects
        United States (cont'd)
        Naeheretal. (1999)
        Vinton, VA
        Summers 1995, 1996
Relationship between O3      8-h max O3:
and daily change in PEF      53.69 ppb
studied in a sample of 473     Range
nonsmoking women aged     17.00-87.63
19-43 years who recently
delivered babies. PEF        24-havgO3:
performed twice daily for a    34.87 ppb
2-week period.  Mixed        Range
linear random coefficient      8.74-56.63
model.
PM2 5, PM10,     O3 was the only exposure related to
SO42~, FT       evening PEF with 5-day cumulative lag
               exposure showing the greatest effect.
                                      24-h avg O3 (per 30 ppb):

                                      Evening PEF:
                                      Lag 1-5:
                                      -7.65 L/min(-13.0,-2.25)
        Canada

X      Braueretal. (1996)
7"1      Fraser Valley, British
O      Columbia, Canada
        Jun-Aug 1993
                               58 berry pickers aged 10-69
                               years had lung function
                               measured before and after a
                               series of outdoor work shifts
                               (average duration =11
                               hours) over 59 days.
                               Analysis using pooled
                               regression with subject-
                               specific intercepts, with and
                               without temperature control.
                           l-hmaxO3:
                           40.3 ppb
                           SD 15.2

                           Work shift 03:
                           26.0 ppb
                           SD11.8
PM25, S042-.
N(V, NH4+,
FT
End shift FEV; and FVC significantly
diminished in relation to O3 levels.
PM2 5 also related to lung function
declines, but O3 remained significant in
2-pollutant models.  Next morning lung
function remained diminished following
high O3 days.  Ozone effects still evident
at or below 40 ppb.  There was an overall
decline of lung function of roughly
10% over course of study, suggesting
subchronic effect. Levels of other
pollutants low during study.
Change in lung function (per ppb
l-hmaxO3):

Endshift lung function:
FEVI: -3.8mL(SE0.4)
FVC:  -5.4mL(SE0.6)

Next morning function:
FEVI: -4.5 mL (SE 0.6)
FVC:  -5.2mL(SE0.7)

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
  Outcomes and Methods    Mean O3 Levels
                  Copollutants
                  Considered
Findings, Interpretation
                                                                                   Effects
X
        Canada (cont'd)

        Brauer and Brook
        (1997)
        Fraser Valley, British
        Columbia, Canada
        Jun-Aug 1993
Additional analysis of
Brauer etal., 1996 with
personal exposure presented
for three groups, stratified
by time spent outdoors.

Group 1:  25 individuals
who spent most of the day
indoors.

Group 2:  25 individuals
who spent much of the day
indoors, but still spent
several daylight hours
outdoors.

Group 3:  15 individuals
who spent the entire work
day outdoors.
l-hmaxO3:

Ambient:
40ppb
SD15
Range 13-84
                 PM2 5, SO42~,     Group 1: 9.0% sampling time (24-h)
                 NO3~, NH4+,      outdoors. Personal to ambient O3 ratio
                 FT              was 0.28.

                                 Group 2: 25.8% sampling time (24-h)
                                 outdoors. Personal to ambient O3 ratio
                                 was 0.48.

                                 Group 3: 100% sampling time (11-h
                                 workshift) outdoors.  Personal to ambient
                                 O3 ratio was 0.96.

                                 One of the first direct demonstrations that
                                 magnitude of personal exposure to O3 is
                                 related to amount of time spent outdoors.
                                 Further showed that,  on average, outdoor
                                 fixed O3 monitors were representative of
                                 day-to-day changes in O3 exposure
                                 experienced by the study population.
                               Same outcomes as reported in
                               Brauer etal., 1996.
        Europe

        Scarlett etal. (1996)
        Surrey, England
        Jun-Jul 1994
Examined 154 children
aged 7 years in a primary
school next to a major
motorway for O3 exposure
effects on PEF0 75, FVC, and
FEV[ using autoregression
for % change in function.
8-h max O3:       PM10, NO2,     No significant association was seen
50.7ppb          pollen          between pulmonary function measures
SD 24.48                        and O3 levels. No pollen effects.
                               Change in lung function (per ppb O3
                               weighted by inverse of variance):

                               FEV075:
                               Lagl: 0.01 mL (-0.12, 0.13)
                               FVC:
                               Lagl: 0.07 mL (-0.09, 0.23)
                               FEV075/FVC:
                               Lagl: -0.1%(-5.1,4.8)

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                         Outcomes and Methods     Mean O3 Levels
                                             Copollutants
                                              Considered
                                        Findings, Interpretation
                                                   Effects
X
to
Europe (cont'd)

Taggartetal. (1996)
Runcom and Widnes
in NW England
M-Sep 1993
Desqueyroux et al.
(2002a)
Paris, France
Nov 1995-Nov 1996
        Desqueyroux et al.
        (2002b)
        Paris, France
        Oct 1995-Nov 1996
Investigated the relationship
of asthmatic bronchial
hyperresponsiveness and
pulmonary function to
ambient levels of
summertime air pollution
among 38 adult nonsmoking
asthmatics (age 18-70
years) using log-linear
models. Analysis limited
to investigation of within
subject variance of the
dependent variables.

60 severe asthmatics (mean
age 55 years) were
monitored by their
physicians for asthma
attacks. Asthma attacks
were based on medical data
collected by a pulmonary
physician at time of clinical
examination. Analysis
using GEE.

39 adult patients with
severe COPD (mean age
67 years) followed over
14 months by physicians
for exacerbations. Logistic
regression with GEE,
examining exposure lags of
0 to 5 days.
                                                          1-havg O3:
                                                          Maximum 61
                                                          24-h avg O3:
                                                          Maximum 24. 5
                 SO2, NO2,
                 smoke
8-h avg O3
(10 a.m.-6 p.m.):

Summer:
41 |ig/m3
SD18

Winter:
11 ng/m3
SD10

8-h avg O3
(10 a.m.-6 p.m.):

Summer:
41 |ig/m3
SD18

Winter:
11 ng/m3
SD10
                                                                            PMU
                                                                    PM10, S02,
                                                                    NO,
No association found for O3. Changes
in bronchial hyperresponsiveness were
found to correlate significantly with
change in the levels of 24-h mean SO2,
NO2, and smoke.
Significant associations between PM10,
O3, and incident asthma attacks were
found. Low O3 levels raise plausibility
50 COPD exacerbations observed over
follow-up period. 1-, 2-, and 3-day lag
O3 significantly related to exacerbations.
No other pollutants significant.  Low O3
levels raise plausibility and confounding
concerns.
24-h avg O3 (per 10 |ig/m3):

% change in bronchial
hyperresponsiveness:
Lagl:  0.3% (-16.6,20.6)
Lag 2:  2.6% (-22.1, 34.9)
8-havgO3(perlO|ig/m3):

Odds ratio:
Lag 2:  1.20(1.03,1.41)
8-h avg O3 (per 10 ng/m3):

Odds ratio:
Lagl:  1.56(1.05,2.32)

Effects appeared larger among
smokers and those with worse gas
exchange lung function.

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
           Reference, Study
         Location and Period
                         Outcomes and Methods     Mean O3 Levels
                                              Copollutants
                                               Considered
       Findings, Interpretation
                                                                    Effects
X
Europe (cont'd)

Just et al. (2002)
Paris, France
Apr- Jim 1996
        Lagerkvist et al. (2004)
        Brussels, Belgium
        May 2002
82 medically diagnosed
asthmatic children (mean
age 10.9 years) followed for
O3 exposure and PEF,
asthmatic attacks, cough,
supplementary use of
P2-agonists, and symptoms
of airway irritation.
Analysis by GEE.
                        57 children (mean age 10.8
                        years) stratified by
                        swimming pool attendance.
                        Pulmonary function test
                        performed and Clara cell
                        protein levels measured in
                        blood before and after light
                        exercise outdoors for two
                        hours. Analysis using
                        student's t-test and Pearson
                        correlation test.  For dose
                        calculations, O3 levels
                        indoors assumed to be 50%
                        of the mean outdoor O3
                        concentration.
                                                            24-h avg O3:
                                                            58.9 |ig/m3
                                                            SD 24.5
                                                            Range
                                                            10.0-121.0
PM10, NO2
                            Daytime outdoor  None
                            Range 77-1 16
                                                            Exposure dose:
                                                            Range 352-914
                                                            |ig/m3-hour
In asthmatic children, O3 exposure was
related to the occurrence of asthma
attacks and additional bronchodilator use.
O3 was the only pollutant associated with
changes in lung function, as shown by an
increase in PEF variability and decrease
in PEF.
Ozone levels did not have any adverse
effect on FEV[ after 2 hours of outdoor
exercise. In addition,  no significant
differences were observed between Clara
cell protein levels before and after
exercise. A marginally significant
positive correlation between ambient O3
dose and Clara cell protein levels
observed among the nonswimmers,
indicating increased antioxidant activity
following O3 exposure in this group.  The
lack of a clear relationship between Clara
cell protein levels and O3 dose may be
attributable to the short period of time
between measurements and diurnal
variability of the protein levels.
24-havgO3(perlO|ig/m3):

% change in daily PEF variability:
Lag 0-2:  2.6%, p = 0.05

Odds ratio:
Supplementary use of p2-agonist
on days on which no steroids
were used:
LagO:  1.41 (1.05,1.89)

Pearson correlation:

O3 exposure dose and Clara cell
protein levels in serum:

All subjects (n = 54):
r = 0.17, p = 0.21
Nonswimmers (n = 33):
r = 0.34, p = 0.06
Swimmers (n = 21):
r =-0.08, p = 0.74

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                              Copollutants
  Outcomes and Methods     Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                   Effects
X
        Europe (cont'd)

        Frischeretal. (1993)
        Umkirch, Germany
        May-Octl991
Nasal lavage repeatedly
performed on 44 school
children (age 9-11 years)
according to protocol
published by Koren et al.
(1990).  Samples collected
morning after "low" and
"high" O3 days.  Nasal
lavage samples analyzed for
polymorphonuclear
leukocyte counts, albumin,
tryptase, eosinophil cationic
protein, and
myeloperoxidase. Analysis
using individual regression
methods.
Stratified
analysis of half
hour avg O3 at
3 p.m.:

Low:
< 140 ng/m3
High:
> 180 ng/m3
None
Significant higher polymorphonuclear
leukocyte counts after high O3 days.
In children without symptoms of rhinitis,
significantly elevated myeloperoxydase
and eosinophil cationic protein
concentrations detected.  Results suggest
that ambient O3 produces an
inflammatory response in the upper
airways of healthy children.
Children without symptoms of
rhinitis (n = 30):

Myeloperoxydase:
LowO3: median 77.3 9 |ig/L
HighO3: median 138.60 ng/L
p < 0.05; Wilcoxon sign rank test

Eosinophilic cationic protein:
LowO3: median 3.49 ng/L
High O3: median 5.39 ng/L
p < 0.05; Wilcoxon sign rank test
        Frischeretal. (1997)
        Umkirch, Germany
        May-Octl991
Examined 44 school
children aged 9-11 years for
ratio of ort/zo-tyrosine to
para- tyro sine in nasal
lavage as a marker of
hydroxyl radical attack.
Nasal lavage performed
according to protocol
published by Koren et al.
(1990). Concomitant lung
function tests performed.
Analysis using individual
regression methods.
Stratified
analysis of 54-h
avg O3 at 3 p.m.:

Low:
< 140 ng/m3
High:
> 180 ng/m3
None
Ambient O3 was associated with the
generation of hydroxyl radicals in the
upper airways of healthy children and
significant lung function decrements.
However, the ortho/para ratio was not
related to polymorphonuclear leukocyte
counts. Passive smoking was not related
to outcomes.
FEV; (% predicted):
Low: 105.4 (SD 15.6)
High: 103.9 (SD 15.0)
A: 1.5, p = 0.031

Ortho/para ratio:
Low: 0.02 (SD 0.07)
High:0.18(SD0.16)
A: 0.17, p = 0.0001

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
  Outcomes and Methods     Mean O3 Levels
                  Copollutants
                   Considered
                       Findings, Interpretation
             Effects
X
        Europe (cont'd)
        Hoppeetal. (1995a,b)
        Munich, Germany
        Apr-Sep 1992-1994
Five study groups (age
12-95 years): (1) senior
citizens (n = 41);
(2) juvenile asthmatics
(n = 43); (3) forestry
workers (n = 41);
(4) athletes (n = 43); and
(5) clerks (n = 40) as a
control group. Examined
for lung function (FVC,
FEVj, PEF) and questions
on irritated airways. Each
subject tested 8 days, 4 days
with elevated or high O3 and
4 days with low O3.
Analysis using Wilcoxon
matched pairs signed rank
test and linear regression.
                           !/2-h max O3
                           (1 p.m.-4 p.m.):

                           Seniors:
                           High: 70 ppb
                           Low:  31 ppb
                           Asthmatics:
                           High: 74 ppb
                           Low:  34 ppb
                           Forestry
                           workers:
                           High: 64 ppb
                           Low:  32 ppb
                           Athletes:
                           High: 71 ppb
                           Low:  28 ppb
                           Clerks:
                           High: 68 ppb
                           Low:  15 ppb
                 None
               No indication that senior citizens
               represent a risk group in this study.
               Senior citizens had the lowest ventilation
               rate (mean 10 L/min). Athletes and
               clerks experienced significant decrements
               in lung function parameters.  Well-
               medicated juvenile asthmatics have a
               trend towards large pulmonary
               decrements.  Forestry workers were
               exposed to motor tool exhaust, which
               might be a potential promoting factor.
!/2-h max O3 (per 100 ppb):

Change in lung function:

Seniors:
FEVI:  0.034 L(SD 0.101)
PEF: 0.006 L/s (SD 0.578)
Asthmatics:
FEVj:  -0.210 L (SD 0.281)
PEF: -0.712 L/s (SD 0.134)*
Forestry workers:
FEVj:  -0.140 L (SD 0.156)
PEF: -1.154 L/s (SD 0.885)*
Athletes:
FEVI:  -0.152 L(SD 0.136)*
PEF: -0.622 L/s (SD 0.589)*
Clerks:
FEVI:  -0.158 L(SD 0.114)*
PEF: -0.520 L/s (SD 0.486)*
        Koppetal. (1999)
        Two towns in Black
        Forest, Germany
        Mar-Octl994
170 school children (median
age 9.1 years) followed over
11 time points with nasal
lavage sampling.  Subjects
were not sensitive to
inhaled allergens. Nasal
lavage samples analyzed for
eosinophil cationic protein,
albumen, and leukocytes.
Analysis using GEE.
!/2-h max O3:

Villingen:
64 ug/m3
5%-95% 1-140

Freudenstadt:
105 ug/m3
5%-95% 45-179
PM10, NO2,     Eosinophil cationic protein and leukocyte
SO2, TSP       levels peaked soon after first major
               O3 episode of summer, but did not
               show response to later, even higher,
               O3 episodes. These observations are
               consistent with an adaptive response
               in terms of nasal inflammation.
Change in log eosinophil cationic
protein concentration (per ug/m3
03):

Early summer:
0.97(0.03,1.92)
Late summer:
-0.43 (-1.34, 0.47)

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
         Location and Period
to
o
o
                                                                     Copollutants
                         Outcomes and Methods    Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                   Effects
X
Europe (cont'd)

Ulmeretal. (1997)
Freudenstadt and
Villingen, Germany
Mar-Oct 1994
        Cuijpers et al. (1994)
        Maastricht, the
        Netherlands
        Nov-Dec 1990
        (baseline),
        Aug 8-16 1991 (smog
        episode)
                                135 children aged 8-11
                                years in two towns were
                                evaluated. Pulmonary
                                function was associated
                                with the highest O3
                                concentration in the
                                previous 24 hours.
                                An initial cross-sectional
                                analysis was followed by
                                a longitudinal analysis
                                using GEE with the data
                                at four time periods
                                (Apr, Jun, Aug, Sep).
                        During episode, 212
                        children (age unspecified)
                        randomly chosen from
                        535 reexamined for lung
                        function and symptoms.
                        Corrected baseline lung
                        function compared by
                        paired t-test. Difference in
                        prevalence of respiratory
                        symptoms examined.
!/2-h max O3:

Freudenstadt:
Median 50.6 ppb
90th% interval
22.5-89.7

Villingen:
Median 32.1 ppb
90th% interval
0.5-70.1
Baseline:
8-h avg O3:
Range 2-56
ug/m3

Smog episode:
l-hmaxO3:
Exceeded
160 ug/m3 on
11 days
None
PM10, S02,
N02
                In the cross-sectional analysis,
                a significant negative association between
                O3 exposure and FVC was only shown
                at the June testing.  ForFEVl5
                no significant associations were detected.
                In contrast, the longitudinal analysis
                obtained a statistically  significant
                negative correlation between O3
                exposure, and FVC and FEV[ for the
                subpopulation living in the town with
                higher O3 levels, Freudenstadt. The
                associations were more pronounced
                in males than females.
                Small decrements in FEV[ and FEF25.75
                were found in the 212 children.
                However, significant decreases in
                resistance parameters also were noted.
                Each day a different group of 30 children
                were measured. The results of the lung
                function are contradictory in that
                spirometry suggest airflow obstruction
                while impedance measurement suggest
                otherwise.  Respiratory symptoms
                impacted by low response rate of 122 of
                212 children due to summer holidays.
                No increase was observed.
Change in lung function (per ug/m3
Vi-hmaxO;,):
Freudenstadt:
-1. 13 mL,p = 0.002
Villingen:
-0.19mL,p = 0.62

FVC:
Freudenstadt:
- 1. 23 mL,p = 0.002
Villingen:
0.02 mL,p = 0.96

Change in lung function and
impedance between baseline and
smog episode:

FEV •
-0.032L(SD0.226\p<0.05
FEF25.75:
-0.086 L/s (SD 0.415), p < 0.01

Resistence at 8 Hz:
-0.47 cmH20/(L/s) (SD 1.17),
p < 0.05
        Gielenetal. (1997)
        Amsterdam, the
        Netherlands
        Apr-Jul 1995
                        61 children aged 7-13 years
                        from two special schools for
                        chronically ill children,
                        followed for twice-daily
                        PEF, symptoms, and
                        medication usage. 77% of
                        cohort had doctor-diagnosed
                        asthma.
l-hmaxO3:
77.3 ug/m3
SD15.7

8-h max O3:
67.0 ug/m3
SD 14.9
PM10, BS,       Morning PEF significantly associated
pollen           with 8-h max O3 at a lag of 2 days.
                BS also associated with PEF.  Among
                14 symptom models tested, only one
                yielded a significant O3 finding (for
                upper respiratory symptoms).  PM10
                and BS, but not O3, were related to
                P2-agonist inhaler use.
                                                       8-h max O3 (per 83.2 ug/m3):

                                                       % change in PEF:
                                                       Morning:
                                                       Lag 2:  -1.86% (-3.58,-0.14)
                                                       Afternoon:
                                                       Lag 2:  -1.88% (-3.94, 0.18)

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                         Outcomes and Methods     Mean O3 Levels
                  Copollutants
                   Considered
                      Findings, Interpretation
                                                   Effects
X
Europe (cont'd)

Hilterman et al. (1998)
Bilthoven, the
Netherlands
M-Octl995
        Hoek and Brunekreef
        (1995)
        Deume and Enkhuizen,
        the Netherlands
        Mar-M 1989
                               60 adult nonsmoking
                               intermittent to severe
                               asthmatics (age 18-55
                               years) followed over
                               96 days. Measured morning
                               and afternoon PEF,
                               respiratory symptoms, and
                               medication use. Analysis
                               controlled for time trends,
                               aeroallergens,
                               environmental tobacco
                               smoke exposures, day of
                               week, temperature. Lags of
                               0 to 2 days examined.
                       The occurrence of acute
                       respiratory symptoms
                       investigated in children
                       aged 7-11 years (Deume
                       n = 241; Enkhuizen n = 59).
                       Symptoms included cough,
                       shortness of breath, upper
                       and lower respiratory
                       symptoms, throat and eye
                       irritation, headache and
                       nausea.  Ozone-related
                       symptom prevalence and
                       incidence were examined.
                       Lags of 0 and 1 day, and
                       mean O3 concentration from
                       previous week were
                       investigated. Analyses
                       using Ist-order
                       autoregressive models and
                       logistic regression models.
8-hmaxO3:
80.1 ug/m3
Range 6-94
PM10,N02,
S02, BS
l-hmaxO3:

Deurne:
57ppb
SD20
Range 22-107

Enkhuizen:
59ppb
SD14
Range 14-114
PM10,N02.
SO,
O3 had strongest association with
symptoms of any pollutant analyzed.
PEF lower with O3 but not statistically
significant.  No effect on medication use.
No effect modification by steroid use or
hyperresponsiveness.
No consistent association between
ambient O3 concentrations and the
prevalence or incidence of symptoms in
either city. The one significant positive
coefficient in Enkhuizen for prevalence
of upper respiratory symptoms was not
confirmed by the Deurne results.
No associations of daily symptom
prevalence or incidence found with any
of the other copollutants examined.
8-h max O3 (per 100 ug/m3):

Odds ratios:
Respiratory symptoms:

Shortness of breath:
LagO:  1.18(1.02,1.36)
Sleep disturbed by breathing:
LagO:  1.14(0.90,1.45)
Pain on deep inspiration:
LagO:  1.44(1.10,1.88)
Cough of phlegm:
LagO:  0.94(0.83,1.07)
Bronchodilator use:
LagO:  1.05(0.94,1.19)

1-h max O3 (per 50 ppb):

Prevalence of symptoms:

Deume:
Any respiratory symptom:
LagO:  -0.06 (SE 0.04)
Cough:
LagO:  -0.07 (SE 0.07)

Upper respiratory symptoms:
LagO:  -0.06 (SE 0.05)

Enkhuizen:
Any respiratory symptom:
LagO:  0.12(SE0.07)
Cough:
LagO:  -0.07 (SE 0.18)
Upper respiratory symptoms:
LagO:  0.18(SE0.09)*

-------
              Table AX7-1  (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                                                    Copollutants
                         Outcomes and Methods    Mean O3 Levels    Considered
                                       Findings, Interpretation
                                                   Effects
Latin America
        Castillejos et al. (1995)    Children aged 754-11 years    1-h max O3:
        SW Mexico City
        Augl990-Octl991
                       (22 males, 18 females)
                       tested up to 8 times for
                       FEV; and FVC, before and
                       after exercise. Target
                       minute ventilation was 35
                       L/min/m2.  Analysis using
                       GEE models.
                                                                   PM,f
112.3ppb
Range 0-365

5th quintile mean
229.1 ppb
The mean % decrements in lung function
were significantly greater than zero only
in the fifth quintile of O3 exposure
(183-365 ppb).
% change with exercise in
5th quintile of O3 exposure
(183-365 ppb):
                                                                                                                                     j: -2. 85% (-4.40, -1.31)
                                                                                                                                 FVC: -1.43% (-2.81, -0.06)
X
oo
        Gold etal. (1999)
        SW Mexico City
        1991
                       40 school children aged
                       8-11 years in polluted
                       community followed for
                       twice-daily PEF and
                       respiratory symptoms. PEF
                       deviations in
                       morning/afternoon from
                       child-specific means
                       analyzed in relation to
                       pollution, temperature,
                       season, and time trend.
                       Morning symptoms
                       analyzed by Poission
                       regression.
24-h avg O3:       PM2 5, PM10     Reported significant declines in PEF
52.0 ppb                         in relation to 24-h avg O3 levels.
IQR 25                          Effects did not vary by baseline symptom
                                history.  Lags chosen to maximize effects
                                and varied by outcome. Ozone generally
                                robust to PM2 5.  Morning phlegm
                                significantly related to 24-h avg O3 at
                                a 1-day lag.
                                      24-h avg O3 (per 25 ppb):

                                      % change in PEF:
                                      Morning:
                                      Lag 1-10:  -3.8%(-5.8,-1.8)
                                      Afternoon:
                                      Lag 0-9: -4.6% (-7.0,-2.1)

                                      % change in phlegm:
                                      Morning:
                                      Lagl:  1.1% (1.0, 1.3)

-------
              Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                                                   Copollutants
                         Outcomes and Methods     Mean O3 Levels    Considered
                                       Findings, Interpretation
                                                                                  Effects
X
VO
Latin America (cont'd)

Romieuetal. (1996)
N Mexico City
Apr-Jul 1991,
Nov 1991-Feb 1992
                               71 mildly asthmatic
                               children aged 5-13 years
                               followed for PEF and
                               respiratory symptoms.
                               Lower respiratory
                               symptoms included cough,
                               phlegm, wheeze and/or
                               difficulty breathing.
1-h max O3:       PM25,PM10,     O3 effects observed on both PEF and
190 ppb          NO2, SO2       symptoms. Symptom, but not PEF,
SD 80                           effects robust to PM10 in two-pollutant
                                models.  Symptoms related to O3
                                included cough and difficulty breathing.
                                                                      1-h max O3 (per 50 ppb):

                                                                      Change in PEF (L/min):
                                                                      Morning:
                                                                      Lag 0:  -2.44 (-4.40, -0.49)
                                                                      Lagl:  -0.23 (-0.41, 1.62)
                                                                      Lag 2:  -1.49 (-3.80, 0.80)
                                                                      Afternoon:
                                                                      LagO:  -0.56 (-2.70, 1.60)
                                                                      Lagl:  -1.27 (-3.20, 0.62)
                                                                      Lag 2:  -1.92 (-4.50, 0.66)

                                                                      Odds ratios:
                                                                      Lower respiratory symptoms:
                                                                      LagO:  1.09(1.03,1.15)
                                                                      Lagl:  1.10(1.04,1.17)
                                                                      Lag 2:  1.04(0.97,1.12)
        Romieuetal. (1997)
        SW Mexico City
        Apr-Jul 1991,
        Nov 1991-Feb 1992
                       Same period as Romieu
                       et al., 1996, but in different
                       section of city. 65 mildly
                       asthmatic children aged
                       5-13 years followed for
                       twice-daily PEF, and
                       respiratory symptoms.
                       Up to 2 months follow-up
                       per child. Analysis
                       included temperature  and
                       looked at 0- to 2-day lags.
                       No time controls. Lower
                       respiratory symptoms
                       included cough, phlegm,
                       wheeze and/or difficulty
                       breathing.
l-hmaxO3:
196 ppb
SD78
                 PM,f
O3 had significant effects on PEF and
symptoms, with largest effects at lags 0
and 1 day.  Symptoms related to O3
included cough and phlegm. Ozone
effects stronger than those for PM, n.
1-h max O3 (per 50 ppb):

Change in PEF (L/min):
Morning:
LagO:  -1.32 (-3.21, 0.57)
Lagl:  -0.39 (-2.24, 1.47)
Lag 2:  -0.97 (-2.94, 0.99)
Afternoon:
LagO:  -1.81 (-3.60,-0.01)
Lagl:  -2.32 (-4.17,-0.47)
Lag 2:  -0.21 (-2.44,2.02)

Odds ratios:
Lower respiratory symptoms:
LagO:  1.11 (1.05,1.19)
Lagl:  1.08(1.01,1.15)
Lag 2:  1.07(1.02,1.13)

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
          Reference, Study
         Location and Period
to
o
o
                                                                    Copollutants
                         Outcomes and Methods     Mean O3 Levels    Considered
                                                                   Findings, Interpretation
            Effects
X
to
o
Latin America (cont'd)

Romieuetal. (1998)
Mexico City
Mar-May 1996
(1st phase)
Jun-Aug 1996
(2nd phase)
        Romieu et al. (2002)
        Mexico City
        Octl998-Apr2000
47 street workers aged
18-58 years randomly
selected to take a daily
supplement (vitamin C,
vitamin E, and beta
carotene) or placebo during
1 st phase of study.
Following washout period,
the use of supplements and
placebos was reversed
during 2nd phase.
Pulmonary function test
performed twice a week at
end of workday. Plasma
concentrations of beta
carotene and a-tocopherol
measured. Analysis using
GEE models.
                        158 asthmatic children aged    1 -h max O3:
                        6-16 years randomly given a   102ppb
                        vitamin (C and E)            SD 47
                        supplement or placebo
                        followed for 12 weeks.
                        Peak flow was measured
                        twice a day and spirometry
                        was performed twice per
                        week in the morning.
                        Double blind study. Plasma
                        concentration of vitamin E
                        levels measured.  Analysis
                        using GEE models.
                                                           l-hmaxO3:        PM10,NO2      During the 1 st phase, O3 levels were
                                                           55% of days                      significantly associated with declines in
                                                           exceeded 110                     lung function parameters. No
                                                           ppb                             associations were observed in the daily
                                                                                           supplement group. A significant
                                                           Workday hourly                   supplement effect was observed. Ozone-
                                                           average during                    related decrements also were observed
                                                           workday prior to                   during the 2nd phase, however the
                                                           pulmonary                        associations were not significant.
                                                           function test:                      Supplementation with antioxidants during
                                                           67.3 ppb                          the 1 st phase may have had a residual
                                                           SD 24                           protective effect on the lung.
                                             PM10, NO2      Ozone levels were significantly
                                                            correlated with decrements in FEF25.75
                                                            in the placebo group, but not in the
                                                            supplement group. When analysis was
                                                            restricted to children with moderate-to-
                                                            severe asthma, amplitudes of decrements
                                                            were larger and significant for FEV1;
                                                            FEF25.75, and PEF in the placebo group.
                                                            Supplementation with antioxidants may
                                                            modulate the impact of O3 exposure on
                                                            the small airways of children with
                                                            moderate to severe asthma.
1-h max O3 (per 10 ppb):

Placebo group:

1st phase:
FEV •
LagO:  -17.9 mL (SE 5.4)*
FVC:
LagO:  -14.8 mL(SE 7.1)*

2nd phase:
FEVI:
LagO:  -3.3mL(SE6.5)
FVC:
LagO:  -0.27 mL (SE 7.8)

No significant associations with O3
observed when taking supplements.

1-h max O3 (per 10 ppb):

Children with moderate to severe
asthma:

Placebo group:
O3 with PM10 and NO2 models:
FEVI:
Lagl:  -4.59 mL, p = 0.04
FEF25.75:
Lagl:  -13.32mL/s,p<0.01
PEF:
Lagl:  -15.01  mL/s, p = 0.04

No association observed in the
vitamin supplement group.

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
          Reference, Study
         Location and Period
                                                                     Copollutants
                         Outcomes and Methods    Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                   Effects
X
to
Latin America (cont'd)

Romieu et al. (2004)
Mexico City
Octl998-Apr2000
                               Additional analysis of
                               Romieu et al., 2002 with
                               data on glutathion
                               S-transferase Ml
                               polymorphism (GSTM1
                               null genotype) in
                               158 asthmatic children.
                               Analysis performed using
                               GEE models, stratified by
                               GSTM1 genotype (null
                               versus positive) within the
                               two treatment groups
                               (placebo and antioxidant
                               supplemented).
l-hmaxO3:
102 ppb
SD47
None
In the placebo group, O3 exposure was
significantly and inversely associated
with FEF2 5_75 in children who had the
GSTM1 null genotype, with larger effects
seen in children with moderate-to-severe
asthma. No significant decrements were
seen in the GSTM1 positive children.
These results provide preliminary
evidence that asthmatic children who
may be genetically impaired to handle
oxidative stress are most susceptible to
the effect of O3 on small airways
function.
1-h max O3 (per 50 ppb):

FEF2 5_75 in children with moderate
to severe asthma:

Placebo group:
GSTM1 null:
Lagl:  -80.8mL/s,p = 0.002
GSTM1 positive:
Lag 1:  -34.4 mL/s, p> 0.10

Supplement group:
GSTM1 null:
Lag 1:  -7.3 mL/s, p> 0.10
GSTM1 positive:
Lag 1: 2.0 mL/s, p> 0.10
        Australia
        Jalaludin et al. (2000)
        Sydney, Australia
        Feb-Dec 1994
                        Three groups of children
                        (mean age 9.6 years):
                        (1) n = 45 with history of
                        wheeze 12 months, positive
                        histamine challenge, and
                        doctor-diagnosed asthma;
                        (2) n = 60 with history of
                        wheeze and doctor-
                        diagnosed asthma;
                        (3) n = 20 with only history
                        of wheeze.  Examined for
                        evening PEF and daily O3
                        using GEE model and
                        population regression
                        models.
Mean daytime     PM10, NO2      A significant negative association was
O3 (6 a.m.-                        found between daily mean deviation in
9 p.m.):                           PEF and same-day mean daytime O3
12 ppb                           concentration after adjusting for
SD 6.8                           copollutants, time trend, meteorological
                                 variables, pollen count, w&Alternaria
Maximum                        count.  The association was stronger in
daytime O3                        a subgroup of children with bronchial
(6 a.m.-9 p.m.):                    hyper-reactivity and doctor-diagnosed
26 ppb                           asthma. In contrast, the same-day
SD 14.4                          maximum O3 concentration was not
                                 statistically associated.
                                                       Change in PEF (per ppb mean
                                                       daytime O3):

                                                       All children (n = 125):

                                                       O3 only model:
                                                       -0.9178 (SE 0.4192), p = 0.03
                                                       O3 with PM10 model:
                                                       -0.9195 (SE 0.4199), p = 0.03
                                                       O3 with PM10 and NO2 model:
                                                       -0.8823 (SE 0.4225), p = 0.04

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
to
o
o
           Reference, Study
         Location and Period
                                                                     Copollutants
                         Outcomes and Methods     Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                   Effects
X
to
to
Australia (cont'd)

Jalaludin et al. (2004)
Sydney, Australia
Feb-Dec 1994
                                Same three groups of
                                children as studied in
                                Jalaludin et al., 2000.
                                Examined relationship
                                between O3 and evening
                                respiratory symptoms
                                (wheeze, dry cough, and
                                wet cough), evening asthma
                                medication use (inhaled
                                (32-agonist and inhaled
                                corticosteroids), and doctor
                                visits for asthma.  Analysis
                                using GEE logistic
                                regression models.
Mean daytime      PM10, NO2
O3 (6 a.m.-
9p.m.):
12ppb
SD6.8
                                                   Maximum
                                                   daytime O3
                                                   (6 a.m.-9p.m.):
                                                   26 ppb
                                                   SD 14.4
                No significant O3 effects observed on
                evening symptoms, evening asthma
                medication use, and doctors visits.
                Also, no differences in the response of
                children in the three groups. A potential
                limitation is that the use of evening
                outcome measures rather than morning
                values may have obscured the effect of
                ambient air pollutants. Only consistent
                relationship was found between mean
                daytime PM10 concentrations  and doctor
                visits for asthma.
                                       Mean daytime O3 (per 8.3 ppb):

                                       Odds ratios:
                                       All children (n = 125):

                                       Wheeze:
                                       Lagl: 1.00(0.93,1.08)
                                       Dry cough:
                                       Lagl: 1.03(0.96,1.11)
                                       Wet cough:
                                       Lagl: 0.97(0.92,1.03)
                                       Inhaled p2-agonist use:
                                       Lagl: 1.02(0.97,1.07)
                                       Inhaled corticosteroid use:
                                       Lagl: 1.02(0.99,1.04)
                                       Doctor visit for asthma:
                                       Lagl: 1.05(0.77,1.43)
        Asia
        Chen etal. (1998)
        Six communities in
        Taiwan
        1994-1995
                        4,697 school children (age
                        unspecified) from a rural
                        area (Taihsi), urban areas
                        (Keelung and Sanchung),
                        and petrochemical industrial
                        areas (Jenwu, Linyuan, and
                        Toufen) cross-sectionally
                        examined for respiratory
                        symptoms and diseases
                        using parent-completed
                        questionnaires.  Multiple
                        logistic regression models
                        used to compare odds of
                        symptoms and diseases in
                        urban or petrochemical
                        areas to the rural area after
                        controlling for potential
                        confounding factors.
24-h avg O3:

Rural area:
52.56 ppb

Urban area:
Mean range
38.34-41.90 ppb
                                                           Petrochemical
                                                           industrial area:
                                                           Mean range
                                                           52.12-60.64 ppb
SO2, CO,
PM10,N02
School children in urban communities,
but not in petrochemical industrial areas,
had significantly more respiratory
symptoms and diseases compared to
those living in the rural community.
However, mean O3 levels in the urban
communities were lower than that of the
rural community. No causal relationship
could be derived between O3 and
respiratory symptoms and diseases in this
cross-sectional study.
Urban areas compared to rural area:

Odds ratios:
Respiratory symptoms:

Morning cough:
1.33(0.98,1.80)
Day or night cough:
1.67(1.21,2.29)
Shortness of breath:
1.40(1.04,1.91)
Wheezing or asthma:
1.68(1.11,2.54)

-------
 ^             Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
 c                                                                                                                                                 ~
           Reference, Study                                                 Copollutants
^       Location and Period     Outcomes and Methods     Mean O3 Levels   Considered           Findings, Interpretation                     Effects
 to     	
 O      Asia (cont'd)

        Chen etal. (1999)        Valid lung function data       l-hmaxO3:       SO2, CO,       FEV; and FVC significantly associated     Change in lung function:
        Three towns in Taiwan:    obtained once on each of      Range           PM10,NO2      with 1-day lag O3. FVC also related to
        Sanchun, Taihsi,         895 children (age 8-13        19.7-110.3 ppb                   NO2, SO2, and CO.  No PM10 effects       O3 only models:
        Linyuan                years) in three towns.                                        observed. Only O3 remained significant    Lag 1:
        May 1995-Jan 1996       Examined relation between                                   in multipollutant models. No PM10        FEVj! -0.64 mL/ppb (SE 0.34)*
                               lung function and pollution                                   effects. A significant O3 effect was not     FVC: -0.79 mL/ppb (SE 0.32)*
                               concentrations on same day                                   evident at O3 levels below 60 ppb.
                               and over previous 1, 2, and                                                                        O3 with NO2 models:
                               7 days.  Multipollutant                                                                            Lag 1:
                               models examined.                                                                                FEV^ -0.85 mL/ppb (SE 0.34)*
                                                                                                                             FVC: -0.91 mL/ppb (SE 0.37)*

                                                                                                                             *p<0.05

 X     ^^^^^^^^^^^^^^^^^^^^^^^^^^=^^^^^^^^^^^^^^^^^^^^^^^^^^=
 to

-------
                              Table AX7-2. Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
           Reference, Study
         Location and Period
  Outcomes and Methods     Mean O3 Levels
                   Copollutants
                   Considered
                       Findings, Interpretation
                                                    Effects
X
to
        United States

        Liao et al. (2004)
        Three locations in U.S.:
        Minneapolis, MN;
        Jackson, MS; Forsyth
        County, NC
        1996-1998
        Peters et al. (2000a)
        Eastern Massachusetts
        1995-1997
5,431 cohort members of
the Atherosclerosis Risk in
Communities study, men
and women aged 45-64
years at entry in 1987.
Association between O3 and
cardiac autonomic control
assessed using 5-minute
heart rate variability indices
collected over a 4-hour
period. Analysis using
multivariable linear
regression models, adjusting
for individual
cardiovascular disease risk
factors and meteorological
factors.

Records of detected
arrhythmias and therapeutic
interventions were
downloaded from
defibrillators implanted in
cardiac clinic patients aged
22-85 years (n =  100).
Analysis was restricted to
defibrillator discharges
precipitated by ventricular
tachycardias or fibrillation.
Data was analyzed by
logistic regression models
using fixed effect models
with individual intercepts.
8-h avg O3
(10 a.m.-6 p.m.):
41 ppb
SD16
PM10, CO,
S02, N02
24-h avg O3:
18.6 ppb
IQR 14.0
PM25,PM10,
BC, CO, NO2,
S02
Significant interaction between O3 and
ethnicity in relation to high-frequency
power (p < 0.05). Ambient O3
significantly associated with high-
frequency  power among whites, but not
blacks. No significant O3 effect on other
heart rate variability indices, including
low-frequency power and SD of normal
R-R intervals.  More consistent
relationships observed between PM10
and heart rate variability  indices.
No significant O3 effects observed for
defibrillator discharge interventions.
For patients with ten or more
interventions, increased arrhythmias
were associated significantly with
PM2 5, CO, and NO2 at various lag
periods, but not O3.
8-h avg O3 (per 16 ppb):

Log-transformed high-frequency
power:
White race:
Lagl:  -0.069 (SE 0.019)*
Black race:
Lagl:  0.047 (SE 0.034)

Log-transformed high-frequency
power:
Lagl:  -0.010 (SE 0.016)

SD of normal R-R intervals:
Lagl:  -0.336 (SE 0.290)

*p<0.05

24-h avg O3 (per 32 ppb):

Odds ratios:
Defibrillator discharges:

Patients with at least one event:
LagO:  0.96(0.47,1.98)
Patients with at least ten events:
LagO:  1.23(0.53,2.87)

-------
                         Table AX7-2 (cont'd). Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
           Reference, Study
         Location and Period
                          Outcomes and Methods     Mean O3 Levels
                                               Copollutants
                                               Considered
                       Findings, Interpretation
                                                                                     Effects
        United States (cont'd)

        Peters etal. (2001)
        Greater Boston area,
        MA
        Janl995-May 1996
                        Case-crossover study design
                        used to investigate
                        association between air
                        pollution and triggering of
                        myocardial infarction in
                        772 patients  (mean age
                        6 1.6 years).  For each
                        subject, one  case period
                        was matched to three
                        control periods 24 hours
                        apart. Conditional logistic
                        regression used for analysis.
                            l-hmaxO3:
                            19.8ppb

                            24-h avg O3:
                            19.9ppb
PM25,PM1
PM1
CO, NO,, SO,
                                  None of the gaseous pollutants, including    Odds ratios:
                                  O3, were significantly associated with
                                  the triggering of myocardial infarctions.
                                  Significant associations reported for
Myocardial infarctions:

2-h avg O3 (per 45 ppb):
Lag 1 hour:
1.31 (0.85,2.03)

24-h avg O3 (per 30 ppb):
Lag 24 hours:
0.94(0.60,1.49)
X
to
Park et al. (2004)
Greater Boston area,
MA
Nov 2000-Oct 2003
Effect of O3 on heart rate
variability was examined in
497 adult males (mean age
72.7 years). Subjects were
monitored during a 4-
minute rest period between
8 a.m. and 1 p.m.  Ozone
levels measured at central
site 1 km from study site.
Exposure averaging times
of 4-hours, 24-hours, and
48-hours investigated.
Modifying effects of
hypertension, ischemic
heart disease, diabetes,
and use of cardiac/
antihypertensive
medications also examined.
24-h avg O3:       PM2 5, particle    Of the pollutants examined, only PM2 5
23.0 ppb          number         and O3 showed significant associations
SD 13.0           concentration,    with heart rate variability outcomes.
                  BC, NO2, SO2,   The 4-hour averaging period was most
                  CO             strongly associated with heart rate
                                  variability indices. The O3 effect was
                                  robust in models including PM2 5. The
                                  associations between O3 and heart rate
                                  variability indices were stronger in
                                  subjects with hypertension (n = 335) and
                                  ischemic heart disease (n = 142).
                                  In addition, calcium-channel blockers
                                  significantly influenced the effect of O3
                                  on low frequency power. Major
                                  limitations of this study are the use of a
                                  short 4-minute period to monitor heart
                                  rate variability and the  lack of repeated
                                  measurements for each subject.
                                                        4-h avg O3 (per 13 ppb):

                                                        Change in low frequency power:

                                                        All subjects:
                                                        -11.5% (-21.3, -0.4)
                                                        Subjects with hypertension:
                                                        -12.6% (-25.0, 1.9)
                                                        Subjects without hypertension:
                                                        -5.4% (-21.6, 14.1)
                                                        Subjects with ischemic heart
                                                        disease:
                                                        -25.8% (-41.9,-5.3)
                                                        Subjects without ischemic heart
                                                        disease:
                                                        -4.8% (-16.7, 8.8)

-------
                        Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
           Reference, Study
         Location and Period
  Outcomes and Methods    Mean O3 Levels
                  Copollutants
                   Considered
Findings, Interpretation
Effects
X
to
        United States (cont'd)

        Gold et al. (2000;
        reanalysis Gold et al.,
        2003)
        Boston, MA
        Jun-Sep 1997
Repeated measurements
of heart rate variability in
subjects aged 53-87 years
(n = 21,163 observations).
Twenty-five minute
protocol included 5 minutes
each of rest, standing,
exercise outdoors, recovery,
and 20 cycles of slow
breathing. Ozone levels
measured at central site
4.8 miles from study site.
Analyses using random
effects models  and GAM
with stringent convergence
criteria.
l-hmaxO3:        PM25, NO2,      Increased levels of O3 were associated
25.7ppb          SO2             with reduced r-MSSD (square root of the
IQR 23.0                          mean of the squared differences between
                                  adjacent normal RR intervals) during the
                                  slow breathing period after exercise
                                  outdoors.  The estimated O3 effects were
                                  similar to those of PM2 5. Results suggest
                                  that O3 exposure may decrease vagal
                                  tone, leading to reduced heart rate
                                  variability.
                                 1-h max O3 (per 23.0 ppb):

                                 Change in r-MSSD:

                                 During first rest period:
                                 O3 only model:
                                 -3.0 ms(SE 1.9), p = 0.12

                                 During slow breathing period:
                                 O3 only model:
                                 -5.8ms(SE2.4),p = 0.02
                                 O3 with PM2 5 model:
                                 -5.4 ms(SE 2.5), p = 0.03
        Canada
        Rich et al. (2004)
        Vancouver, British
        Columbia, Canada
        Feb-Dec 2000
Case-crossover study
design used to investigate
association between air
pollution and cardiac
arrhythmia in patients aged
15-85 years (n = 34) with
implantable cardioverter
defribillators. Controls
periods were selected 7 days
before and after each case
day. Analysis using
conditional logistic
regression.
1-h max O3:        PM25, PM10,     No consistent association between any
27.5 ppb          EC, OC,         of the air pollutants, including O3, and
SD 9.7            SO42~, CO,       implantable cardioverter defribillators
IQR 13.4          NO2, SO2        discharges.  No significant association
                                  observed in all year data, however,
                                  significant relationship found in winter
                                  months at a 3-day lag.  Overall, little
                                  evidence that air pollutants affect risk
                                  of cardiac arrhythmias, however, power
                                  was limited to study subtle effects.
                                 No quantitative results for O3.

-------
                        Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
to
o
o
           Reference, Study
         Location and Period
                          Outcomes and Methods     Mean O3 Levels
                                              Copollutants
                                               Considered
                                         Findings, Interpretation
                                                   Effects
        Canada (cont'd)

        Vedal et al. (2004)
        Vancouver, British
        Columbia, Canada
        1997-2000
                        Retrospective, longitudinal
                        panel study of 50 patients
                        (age 12-77 years) with
                        implantable cardioverter
                        defribillators. Occurrence
                        of cardiac arrhythmia was
                        associated with air
                        pollutants over four-year
                        period. GEE used for
                        analysis.
                            l-hmaxO3:
                            28.2 ppb
                            SD 10.2
                            IQR 13.8
                  PM10, CO,
                  N02, S02
No consistent association between any
of the air pollutants and % change in
arrhythmia. Among patients with at least
2 arrhythmia event-days per year, a
significant negative relationship between
O3 and arrhythmias was observed at lag
3-day during the summer, but no
associations were found during the
winter. These results do not provide
evidence for an O3 effect on cardiac
arrhythmias in susceptible patients.
No quantitative results for O3.
X
to
Latin America

Holguin et al. (2003)
Mexico City
Feb-Apr 2000
Association between O3
and heart rate variability
examined in 34 elderly
subjects (mean age
79 years), in a nursing
home.  Subjects were
monitored during a
5-minute rest period
between 8 a.m. and 1 p.m.
every other day for a
3-month period. A total of
595 observations were
collected. Ambient O3
levels obtained from central
site 3 km upwind from
study site. Analysis
performed using GEE
models adjusting for
potential confounding
factors including age and
average heart rate.
1-h max O3:        PM2 5 (indoor,    Only PM2 5 and O3 were significantly
149 ppb           outdoor,         associated with heart rate variability
SD40             total), NO2,      outcomes. A significant effect of O3 on
                  SO2, CO         heart rate variability was limited to
                                  subjects with hypertension (n = 21).
                                  In two-pollutant models, the magnitude
                                  of the PM2 5 effect decreased slightly but
                                  remained significant, whereas O3 was
                                  no longer associated with heart rate
                                  variability indices.
                                        1-h max O3 (per 10 ppb):

                                        Log10 high frequency
                                        power/100,000 ms2:
                                        All subjects:
                                        -0.010 (-0.022, 0.001)
                                        Subjects with hypertension:
                                        -0.031 (-0.051,-0.012)
                                        Subjects without hypertension:
                                        0.002 (-0.012, 0.016)

                                        Log10 low frequency
                                        power/100,000 ms2:
                                        All subjects:
                                        -0.010 (-0.021, 0.001)
                                        Subjects with hypertension:
                                        -0.029 (-0.046,-0.011)
                                        Subjects without hypertension:
                                        0.001 (-0.012, 0.015)

-------
                                   Table AX7-3.  Effects of O3 on Daily Emergency Department Visits
1
to
o
o














J>
X
~I~J
to
oo


t-1
Reference, Study
Location and Period Outcomes and Design
United States

Jaffe et al. (2003) Daily time series study
Cincinnati, Cleveland, of emergency
and Columbus, OH department visits for
Jun-Aug 1991-1996 asthma among
Medicaid recipients
aged 5-34 years.















Mean O3
Levels


8-h max O3:

Cincinnati:
60 ppb
SD20

Cleveland:
50 ppb
SD17

Columbus:
57 ppb
SD16








Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation


PM10, NO2, 1 , 2, 3 Poisson regression with control
SO2 for city, day of week, week,
year, minimum temperature,
overall trend, and a dispersion
parameter. No specific effort
to control cycles, but
regression residuals were
uncorrelated, presumably due
to seasonal restriction. Results
shown for individual cities and
overall. PM10 available only
every 6th day. Positive
relationships between
emergency department visits
for asthma and 8-h max O3
levels lagged 2 to 3 days.
Results of borderline statistical
significance. Other pollutants
also related to asthma
emergency department visits
in single-pollutant models.
Effects
(Relative Risk and 95% CI)


8-h max O3 (per 30 ppb):

Cincinnati:
Lag 2: 1.16(1.00,1.37)
Cleveland:
Lag 2: 1.03(0.92,1.16)
Columbus:
Lag 3: 1.16(0.98,1.37)

Three cities: 1.09(1.00,
1.19)










Jones etal. (1995)
Baton Rouge, LA
Jun-Aug 1990
Daily emergency
department visits for
respiratory complaints
over a 3-month period
in pediatric (age 0-17
years), adult (age 18-60
years), and geriatric
(age > 60 years)
subgroups.
l-hmaxO3:
69.1 ppb
SD28.7

24-havgO3:
28.2 ppb
SD11.7
Mold, pollen      Not specified.
Relatively simple statistical
approach using ordinary least
squares regression to model
effects of O3 by itself and of O3
along with pollen counts, mold
counts, temperature, and
relative humidity. No direct
control of cycles but authors
reported non-significant
autocorrelations among model
residuals. Data restriction to
3-month period may have
reduced any cyclic behavior.
Significant associations
between respiratory emergency
department visits and O3
observed for adult age group
only in multiple regression
models.
24-h avg O3 (per 20 ppb):

Pediatric: 0.87(0.65,1.09)
Adult:  1.20(1.01,1.39)
Geriatric: 1.27(0.93,1.61)

-------
                                   Table AX7-3 (cont'd).  Effects of O3 on Daily Emergency Department Visits
X
to
VO
Reference, Study
Location and Period
United States (cont'd)
Weisel et al. (2002)
New Jersey
May-Aug 1995








Friedman et al. (2001)
Atlanta, GA
M-Aug 1996
Outcomes and Design

Daily asthma
emergency department
visits for all ages.








Emergency department
visits and hospital
admissions for asthma
Mean O3 Copollutants
Levels Considered

1-h max O3; Pollen, spores
5-h avg O3
(10a.m.-
3 p.m.); and 8-h
avg O3 (2 p.m.-
10p.m.)
analyzed.

Levels not
reported.

1-h max O3: NO2, SO2, CO,
Levels not PM10, mold
reported.
Lag Structure Method, Findings,
Examined Interpretation

0, 1 , 2, 3 No control for time, but
authors report no
autocorrelation, which
alleviates concerns about lack
of control. Significant O3
effects reported, even after
adjusting for potential
confounding by pollen. All
three O3 indices gave
essentially same results.

0, 0-1, 0-2 Analysis using Poisson GEE
models addressing serial
autocorrelation. An overall
Effects
(Relative Risk and 95% CI)

Slope estimate
(visits/day/ppb):

Excluding data from May
1995 when pollen levels are
high:

O3 only model:
LagO: 0.039, p = 0.049
O3 with pollen model:
LagO: 0.040, p = 0.014
1-h max O3 (per 50 ppb):

Pediatric emergency
in children aged 1-16
years. Outcomes
measures during 1996
Summer Olympics
were compared to a
baseline period of 4
weeks before and after
the Olympic Games.
decrease was observed when
comparing the number of visits
and hospitalizations during the
Olympic Games to the baseline
period.  However, significant
associations between O3 and
asthma events were found
during the Olympic Games.
departments:
LagO: 1.2(0.99,1.56)
Lag 0-1:  1.4(1.04,1.79)
Lag 0-2:  1.4(1.03,1.86)
Metzger et al. (2004)
Atlanta, GA
Jan 1993-Aug 2000











Emergency department
visits for total and
cause- specific
cardiovascular diseases
by age groups 19+
years and 65+ years.








8-h max O3:
Median 53.9
ppb
1 Oth % to 90th
% range 13.2 -
44.7








NO2, SO2, CO,
PM25,PM10,
PM10.2,,
ultrafine PM
count, SO42~,
LT, EC, OC,
metals,
oxygenated
hydrocarbons





0-2 Poisson GLM regression used
for analysis. A priori models
specified a lag of 0 to 2 days.
Secondary analyses performed
to assess alternative pollutant
lag structures, seasonal
influences, and age effects.
Cardiovascular visits were
significantly associated with
several pollutants, including
N02, CO, and PM2 5, but
notO3.


8-h max O3 (per 25 ppb):

All ages:
Total cardiovascular:
1.008(0.987,1.030)
Dysrhythmia:
1.008(0.967,1.051)
Congestive heart failure:
0.965(0.918,1.014)
Ischemic heart disease:
1.019(0.981,1.059)
Peripheral vascular and
cerebro vascular disease:
1.028(0.985,1.073)

-------
                            Table AX7-3 (cont'd).  Effects of O3 on Daily Emergency Department Visits
X
Reference, Study
Location and Period
United States (cont'd)
Peel et al. (2004)
Atlanta, GA
Jan 1993-Aug 2000











Tolbert et al. (2000)
Atlanta, GA
Jun-Aug 1993-1995















Outcomes and Design

Emergency department
visits for total and
cause- specific
respiratory diseases by
age groups 0-1, 2-18,
19+, and 65+ years.








Pediatric (aged 0-16
years) asthma
emergency department
visits over three
summers in Atlanta.
A unique feature of the
study was assignment
of O3 exposures to
zip code centroids
based on spatial
interpolation from nine
O3 monitoring stations.






Mean O3
Levels

8-h max O3:
55.6ppb
SD23.8











l-hmaxO3:
68.8 ppb
SD21.1

8-h max O3:
59.3 ppb
SD19.1











Copollutants
Considered

NO2, SO2, CO,
PM25,PM10,
PM10.2,,
ultrafine PM
count, SO42~,
H+, EC, OC,
metals,
oxygenated
hydrocarbons





PM10,N02,
mold, pollen
















Lag Structure Method, Findings,
Examined Interpretation

0-2 Poisson GEE and GLM
regression used for analysis.
A priori models specified a lag
of 0 to 2 days. Also performed
secondary analyses estimating
the overall effect of pollution
over the previous two weeks.
Seasonal analyses indicated
stronger associations with
asthma in the warm months.
Quantitative results not
presented for multipollutant,
age-specific, and seasonal
analyses.
1 A priori specification of
model, including a lag of 1 day
for all pollutants and
meteorological variables.
Secondary analysis using
logistic regression of the
probability of daily asthma
visits, referenced to total visits
(asthma and non-asthma).
Significant association with
O3 and PM10 in 1-, but not in
2-pollutant models (correlation
between O3 and PM10:
r = 0.75). Secondary analysis
indicated nonlinearity, with O3
effects clearly significant only
for days > 100 ppb versus days
< 50 ppb.
Effects
(Relative Risk and 95% CI)

8-h max O3 (per 25 ppb):

All ages:
Total respiratory:
1.024(1.008,1.039)
Upper respiratory infections:
1.027(1.009,1.045)
Asthma:
1.022(0.996,1.049)
Pneumonia:
1.015(0.981,1.050)
COPD:
1.029(0.977,1.084)

8-h max O3 (per 20 ppb):

Poisson regression:
O3 only model:
1.040(1.008,1.074)


Logistic regression:
O3 only model:
1.042(1.017,1.068)
O3 with PM10 model:
1.024(0.982,1.069)







-------
                                    Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
United States (cont'd)
Zhu et al. (2003)
Atlanta, GA
Jun-Aug 1993-1995
Outcomes and Design
Asthma emergency
department visits in
children (age 0-16
Mean O3
Levels
8-h max O3:
Levels not
reported.
Copollutants
Considered
None
Lag Structure
Examined
1
Method, Findings,
Interpretation
Used Bayseian hierarchical
modeling to address model
variability and spatial
Effects
(Relative Risk and 95% CI)
8-h max O3 (per 20 ppb):
Posterior median:
                             years) over three
                             summers in Atlanta.
associations. Data was
analyzed at the zip code level
to account for spatially
misaligned longitudinal data.
A positive, but nonsignificant
relationship between O3 and
emergency room visits for
asthma.
1.016(0.984,1.049)
Canada
Delfinoetal. (1997b)
Montreal, Quebec,
Canada
Jun-Sep 1992-1993












Daily time series
ecologic study of
emergency department
visits for respiratory
complaints within five
age strata (< 2, 2-1&,
19-34, 35-64, > 64
years).








l-hmaxO3: PM10, PM25,
1992: 33.2 ppb SO42-, H+
SD 12.6
1993: 36.2 ppb
SD 13.8

8-h max O3:
1992: 28. 8 ppb
SDH. 3
1993: 30.7 ppb
SDH. 5





0, 1, 2 Used ordinary least squares,
with 1 9-day weighted moving
average pre-filter to control
cycles; weather also controlled.
Significant effects reported for
1-day lag O3 in 1993 only for
age > 64 years. This O3 effect
reported to be robust in two-
pollutant models. LowO3
levels raise plausibility
concerns. Short data series,
multiple tests performed, and
inconsistent results across
years and age groups raise
possibility of chance findings.

1993 (age > 64 years):

1-h max O3 (per 36.2 ppb):
Lag 1: 1.214(1.084,1.343)

8-h max O3 (per 30.7 ppb):
Lagl: 1.222(1.091,1.354)









-------
                                   Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
X
to
Reference, Study
Location and Period
Canada (cont'd)
Delfinoetal. (1998b)
Montreal, Quebec,
Canada
Jun-Aug 1989-1990







Stiebetal. (1996)
Saint John, New
Brunswick, Canada
May-Sep 1984-1992

Outcomes and Design

Daily time series
ecologic study of
emergency department
visits for respiratory
complaints across all
ages and within four
age strata (<2, 2-34,
35-64, >64 years).



Daily emergency
department visits for
asthma in all ages,
age<15 years and
15+ years.
Mean O3 Copollutants
Levels Considered

1 -h max O3 : Estimated
1989: 44.1ppb PM25
SD 18.3
1990: 35.4 ppb
SD 12.9

8-h max O3:
1989: 37. 5 ppb
SD 15.5
1990: 29. 9 ppb
SD11.2
l-hmaxO3: SO2,NO2,
4 1.6 ppb SO42% TSP
Range 0-1 60
95th % 75

Lag Structure Method, Findings,
Examined Interpretation

0, 1, 2 Same analytical approach used
in Delfino et al, 1997. Results
presented only for 1989.
Significant effects reported for
1-day lag O3 in 1989 only for
age > 64 years. This O3 effect
reported to be robust in
2-pollutant models.



0, 1, 2, 3 Poisson analysis with control
of time based on 19-day
moving average filter. Also
controlled day of week and
weather variables. Ozone was
Effects
(Relative Risk and 95% CI)

1 989 (age > 64 years):

1-h max O3 (per 44.1 ppb):
Lag 1: 1.187(0.969,1.281)

8-h max O3 (per 37.5 ppb):
Lagl: 1.218(1.097,1.338)

No significant O3 effects in
other age groups or for 1990.

1-h max O3 > 75 ppb:

Lag 2: 1.33(1.10,1.56)


                                                                                              only pollutant consistently
                                                                                              associated with emergency
                                                                                              department visits for asthma,
                                                                                              but effect appeared nonlinear,
                                                                                              with health impacts evident
                                                                                              only above 75 ppb O3.

-------
3
                              Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
Europe
Atkinson et al.
(1999a)
London, England
1992-1994



Thompson et al.
(2001)
Belfast, N Ireland
1993-1995






Outcomes and Design

Emergency department
visits for respiratory
complaints, asthma for
all ages and age 0-14,
15-64, and 65+ years.


Asthma emergency
department admissions
in children (age
unspecified)






Mean O3
Levels

8-h max O3:
17.5 ppb
SD11.5




24-h avg O3:

Warm season:
18.7 ppb
IQR9

Cold season:
17.1 ppb
IQR12

Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

NO2, SO2, CO, 0, 1 , 2, 3 Poisson GLM regression used
PM10 0-1,0-2,0-3 for analysis. No warm season
analysis attempted. PM10
positively associated.



PM10, SO2, 0,0-1, 0-2, 0-3 GLM with sinusoids.
NO2, CO, Pre-adjustment. Very low O3
benzene levels in both seasons. No O3
effect in warm season.
Significant inverse O3
associations in full-year and
cold-season models. After
adjusting for benzene in model
O3 was no longer negatively
associated with asthma visits.
Effects
(Relative Risk and 95% CI)

8-h max O3 (per 25.7 ppb):

All ages:
Total respiratory:
Lag 1: 1.017(0.991,1.043)
Asthma:
Lagl: 1.027(0.983,1.072)
24-h avg O3 (per 10 ppb):

All year:
O3 only model:
Lag 0-1: 0.93(0.87,1.00)
O3 with benzene model:
Lag 0-1: 1.08(0.97,1.21)

Warm season:
O3 only model:
                                                                                                           Lag 0-1: 0.99(0.89,1.10)
                                                                                                           Cold season:
H
6
o
2!
•£—\
O
H
0

H

O
O
H
W


Bourcier et al. (2003) Ophthalmological 24-h avg O3: PM10, SO2, 0,1,2,3 Logistic Regression
Paris, France emergency 35.7 |ig/m3 NO2
Jan 1999-Dec 1999 examination; Range 1-97
conjunctivitis and
related ocular surface
problems.





Lag 0-1: 0.89(0.82,0.97)

Results indicate a strong
relation to NO2 and NO.

24-h avg O3 (per 69 |ig/m3):
Conjunctivitis:
LagO: 1.13(0.90,1.42)






-------
                                     Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
Europe (cont'd)
Castellsague et al.
(1995)
Barcelona, Spain
1985-1989





Tobias etal. (1999)
Barcelona, Spain
Outcomes and Design

Daily emergency
department visits for
asthma in persons
aged > 14 years.





Daily asthma
emergency department
Mean O3
Levels

l-hmaxO3:

Summer:
43ppb
IQR22

Winter:
29ppb
IQR16
Levels not
reported.
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

BS, SO2, NO2 Not specified. Poisson regression with year
and month dummy variables
and extensive control for
weather factors (minimum,
maximum, mean temperature,
relative humidity, dewpoint
temperature; continuous and
categorical parameterizations)

BS, NO2, SO2 Not specified. Poisson analysis using
APHEA methodology.
Effects
(Relative Risk and 95% CI)

1-h max O3 (per 12.7 ppb):

Summer:
0.991 (0.939, 1.045)

Winter:
1.055(0.998,1.116)


O3 results were sensitive to
method used to control
        1986-1989
visits. Investigated
sensitivity of results
to four alternative
methods for controlling
asthma epidemics.
Asthma epidemics either
not controlled, or controlled
with one, six, or individual
epidemic dummy variables.
asthma epidemics, with
regression coefficients
ranging over 5-fold
depending on the model.
Only 1 of 8 models reported
had a significant O3 effect.
O
5>
-n
H
6
o
!2
O
— ]
O
r^t
0
H
O
^—i
Tenias et al.
(1998,2002)
Valencia, Spain
1994-1995









Daily emergency
department visits for
asthma and COPD
in persons aged
> 14 years.








l-hmaxO3:

All year:
62.8 ug/m3
Warm season:
74.0 ug/m3
Cool season:
5 1.4 ug/m3





BS, NO2, SO2, 0, 1 , 2, 3, 4, 5 Poisson analysis using APHEA
CO methodology. Compared
warm and cold season effects.
GAM explored in sensitivity
analysis. For asthma, both O3
and NO2 significant in 1- and
2-pollutant models, and O3
effect larger in warm season.
For COPD, both O3 and CO
significant in both 1- and
2-pollutant models and no
difference in O3 effects by
season.
1-h max O3 (per 10

Asthma:
All year:
Lagl: 1.06(1.01,
Warm season:
Lag 1: 1.08(1.02,
Cold season:
Lagl: 1.04(0.97,

COPD:
All year:
Lag 5: 1.06(1.02,
ug/m3):



1.11)

1.05)

1.11)



1.10)
O
HH
H
W

-------
                            Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
to
o
o
X
Reference, Study
Location and Period
Latin America
Hemandez-Garduno
etal. (1997)
Mexico City
May 1992-Jan 1993



Lin etal. (1999)
Sao Paulo, Brazil
May 1991 -Apr 1993



Ilabacaetal. (1999)
Santiago, Chile
Feb 1995-Aug 1996






Asia
Chew etal. (1999)
Singapore
Jan 1990-Dec 1994





Outcomes and Design

Visits to clinics for
respiratory diseases in
persons aged 1 month
to 92 years.



Daily pediatric (age
unspecified) respiratory
emergency department
visits.


The association
between pollutant
levels and emergency
visits for pneumonia
and other respiratory
illnesses among
children.



Emergency department
visits for asthma
in persons aged
3-21 years.



Mean O3
Levels

% time
exceeding 1-h
max O3 of 120
ppb:
6.1-13.2% by
location

l-hmaxO3:
34 ppb
SD22



O3 1-h max:

Warm season:
66.6 ug/m3
SD25.2

Cold season:
27.6 ug/m3
SD 20.2

l-hmaxO3:
23 ppb
SD15




Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

SO2, NO2, CO 0,1,2,3,4,5 GLM with pre-adjustment.
Ozone at lags 0 and 5 days
significantly associated with
daily visits for all ages,
age < 14 years, and 15+ years.
Neither O3 nor NO2 significant
in 2-pollutant model.
SO2, CO, 0,1,0-1,0-2, Seasonal control using month
PM10,NO2 0-3,0-4,0-5 dummy variables. Also
controlled day of week,
temperature. Both O3 and
PM10 associated with outcome
alone and together.
PM10, PM2 5, 1,2, 3, 1-7 Poisson regression analysis.
SO2,NO2








TSP, PM10, 0,1,2 Simplistic but probably
SO2, NO2 adequate control for time by
including 1-day lagged
outcome as covariate.
In adjusted models that
included covariates, O3
had no significant effect.
Effects
(Relative Risk and 95% CI)

1-h max O3 (per maximum
less average, value not given)

LagO: 1.19(SE0.08),
p<0.05
Lag 5: 1.19 (SE 0.08),
p<0.05
1-h max O3 (per 5 ppb):

O3 only model:
Lag 0-4: 1.022(1.016,1.028)
O3 with PM10 model:
Lag 0-4: 1.015(1.009,1.021)
Warm season:
1-h max O3 (per 30 ug/m3):
Lag 2: 1.019(1.003,1.035)

Cold season:
1-h max O3 (per 24 ug/m3):
Lag 2: 0.995(0.978,1.011)



No quantitative results
presented for O3.






-------
Table AX7-4. Effects of O3 on Daily Hospital Admissions
t—t
1
to
o
o
^













X

OJ
ON


O
|>

6
0
2;
o
H
O
0
w
o

o
HH
H
W



Reference, Study
Location and Period

United States
Niedell (2004)
California
1992-1998



















Mann et al. (2002)
South Coast air basin,
CA
1988-1995



Linn et al. (2000)
Los Angeles, CA
1992-1995







Outcomes and Design


Emergency department
visits for asthma within
five age strata (0-1, 1-3,
3-6, 6-12, and 12-1 8
years).

















Ischemic heart disease
admissions for age
40+ years.




Total respiratory and
total cardiovascular
admissions for age
30+ years.






Copollutants
Mean O3 Levels Considered


O3 (index not CO, NO2, PM10;
specified): multipollutant
38.9 ppb models
SD 17.8

Low SES:
40.1 ppb

High SES:
38.3 ppb












8-h max O3: PM10, CO, NO2
50.3 ppb
SD30.1
IQR 39.6



24-havg03: PM10,CO,NO2

Winter: 14 ppb
SD7
Spring: 32 ppb
SD10
Summer: 36 ppb
SD8
Fall: 15 ppb
SD9
Lag Structure Method, Findings,
Examined Interpretation


Not specified. Statistical analysis using
naturally occurring seasonal
variations in pollutant
concentrations within zip
codes. Methodology not
clearly stated. Consistent
significant positive effects
only observed for CO.
Negative O3 effect observed
in all age groups. Number of
smog alerts was negatively
associated with asthma
hospitalizations, indicating
avoidance behavior on high
O3 days. Interaction term
with indicator variable for
low SES was positive in all
age groups and statistically
significant in age 3-6 years
and 12-18 years, after
adjusting for number of smog
alerts.
0, 1 , 2, 3, 4, 5, Poisson GAM with cubic
0-1, 0-2, 0-3, B-splines; co-adjustment.
0-4 No significant O3 effects
observed overall or in warm
season. CO and NO2
significant in full-year
analyses.
0 Poisson GLM;
co-adjustment. Only
significant O3 effects
observed were inverse
associations with total
cardiac admission in full-year
and winter season, suggesting
residual confounding.
No significant effects of O3
on respiratory admissions.
Effects
(Relative Risk and 95% CI)


Slope estimate (adjusting for
number of smog alerts):

O3 with CO, NO2, and PM10
models:
Age 3-6 years:
-0.038 (SE 0.014)
Age 6-12 years:
-0.044 (SE 0.013)
Age 12- 18 years:
-0.022 (SE 0.011)

O3 x low SES interaction
term:
Age 3-6 years:
0.092 (SE 0.026)
Age 6-12 years:
0.024 (SE 0.024)
Age 12- 18 years:
0.042 (SE 0.01 9)


O3 coefficients all negative,
but no consistent, significant
effect.




24-h avg O3 (per 10 ppb):

All year:
Respiratory:
1.008(1.000,1.016)
Cardiovascular:
0.993(0.987,0.999)




-------
^**

to
o
o











X

OJ



O
E>
'•Tj
H
6
0
o
H
0
C|
o
w
o
H
w

Reference, Study
Location and Period

United States (cont'd)
Nauenberg and Basu
(1999)
Los Angeles, CA
1991-1994
Sheppard et al.
(1999; reanalysis
Sheppard, 2003)
Seattle, WA
1987-1994


Moolgavkar et al.
(1997)
Minneapolis/St. Paul,
MN and Birmingham,
AL
1986-1991





Schwartz et al. (1996)
Cleveland, OH
Apr-Oct 1988-1990







i  64 years.







Total respiratory
admissions for age
65+ years.







LF1C J"V^V/-t. 1^1 ICV. IS Ul \_/j Ull J
Copollutants
Mean O3 Levels Considered


24-h avg O3: PM10
19. 88 ppb
SDH. 13

8-hmaxO3: PM25,PM10,
30.4 PM10.25, SO2,
IQR 20 CO




24-h avg 03: PM10, SO2,NO2

Minnesota:
26.2 ppb
IQR 15.3

Alabama:
25.1 ppb
IQR 12.7


l-hmaxO3: PM10, SO2
56 ppb
IQR 28







uauy nus|jii
-------
                                          Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
to
o
o
X
oo
Reference, Study
Location and Period Outcomes and Design
United States (cont'd)
Weisel et al. (2002) Asthma admissions for
New Jersey all ages.
May-Aug 1995







Canada
Burnett et al. (1 997a) Total respiratory
16 Canadian cities admissions for all ages,
1981-1991 age <65 years and
65+ years.








Burnett et al. (1995) Respiratory and
168 Hospitals in cardiovascular
Copollutants
Mean O3 Levels Considered

1-h max O3; Pollen, spores
5-h avg O3
(10 a.m.-3 p.m.);
and 8-h avg O3
(2 p.m.-lO p.m.)
analyzed.

Levels not
reported.


1-h max O3: SO2,NO2,
CO, coefficient
All cities: ofhaze
A
All year;
31 ppb
95th % 60

Apr-Dec only:
33 ppb
95th % 64

1-h max O3: SO42~
36.3 ppb
Lag Structure Method, Findings,
Examined Interpretation

0,1,2,3 No control for time, but
authors report no
autocorrelation, which
alleviates concerns about
lack of control. Significant
O3 effects reported after
adjusting for potential
confounding by pollen.



0, 1, 2, 0-1, Poisson GLM with co-
0-2, 1-2 adjustment. Results stratified
by season. Significant O3
effect observed in warm

season only. No O3 effects
on control outcomes. Results
consistent across cities.




1 GLM with pre-adjustment of
outcome variables. Results
Effects
(Relative Risk and 95% CI)

Slope estimate
(admissions/day /ppb):

5-h avg O3 and 8-h avg O3:
O3 only model:
Lag 2: 0.099, p = 0.057

All three O3 indices:
O3 with pollen model:
Lag 2: 0.1 l,p = 0.033

1-h max O3 (per 30 ppb):

All ages:
Jan-Mar:

Lag 1: 0.994(0.964,1.025)
Apr-Jun:
Lagl: 1.042(1.012,1.073)
Jul-Sep:
Lag 1: 1.050(1.026,1.074)
Oct-Dec:
Lagl: 1.028(0.998,1.059)
No quantitative results
presented for O3.
        Ontario, Canada
        1983-1988
admissions for all
ages and within age
strata. Study focused
mainly on testing for
sulfate effects.
stratified by season. Authors
report that O3 associated with
respiratory admission in
warm season only.

-------
                                            Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
to
o
o
          Reference, Study
        Location and Period
                       Outcomes and Design   Mean O3 Levels
                                         Copollutants
                                         Considered
                  Lag Structure
                    Examined
                                      Method, Findings,
                                        Interpretation
                                                                                          Effects
                                                                                 (Relative Risk and 95% CI)
X
VO
Canada (cont'd)

Burnett etal. (1997b)
Toronto, Ontario,
Canada
Summers 1992-1994
Burnett etal. (1999)
Toronto, Ontario,
Canada
1980-1994
                               Unscheduled
                               respiratory and
                               cardiovascular
                               admissions for all ages.
Cause-specific
respiratory and
cardiovascular
admissions for all ages.
Cause categories
included asthma,
COPD, respiratory
infections, heart failure,
ischemic heart disease,
and cerebrovascular
disease.
                                              l-hmaxO3
                                              41.2ppb
                                              IQR22
24-h avg O3 :
19. 5 ppb
IQR19
                                         PM25,PM10,
                                         S042-, S02,
                                         N02, CO,
                                         coefficient
                                         of haze
           H+
Estimated
PM25,PM10,
PM10.25, CO,
NO2, SO2
                   0,1,2,3,4,
                  2 to 5 multiday
                  periods lagged
                    1 to 4 days
                                                                                           0,1,2,0-1,
                                                                                          0-2,1-2,1-3,
                                                                                            2-3,2-4
                                                   Poisson GLM with co-
                                                   adjustment. Results stratified
                                                   by season. Ozone and
                                                   coefficient of haze strongest
                                                   predictors of outcomes.
                                                   Ozone effects on both
                                                   outcomes were robust to PM.
                                                   PM effects were not robust to
                                  Poisson GAM with LOESS
                                  pre-filter applied to pollution
                                  and hospitalization data.
                                  Ozone effects seen for
                                  respiratory outcomes only.
                                  Ozone effect robust to PM;
                                  not vice versa. No seasonal
                                  stratification.
12-h avg O3 (8 a.m.-8 p.m.)
(per 11.5 ppb):

Models adjusted for
temperature and dewpoint:
Respiratory :
Lag 1-3: 1.064(1.039,1.090)
Cardiovascular:
Lag 2-4: 1.074(1.035,1.115)

24-h avg O3 (per 19.5 ppb):
                                                                                         1.063(1.036,1.091)
                                                                                                                              Asthma:
                                                                                                                              Lag 1-3:
                                                                                                                              COPD:
                                                                                                                              Lag 2-4:  1.073(1.038,1.107)
                                                                                                                              Respiratory infection:
                                                                                                                              Lag 1-2:  1.044(1.024,1.065)
        Burnett etal. (2001)
        Toronto, Ontario,
        Canada
        1980-1994
                       Acute respiratory
                       disease admissions for
                       age <2 years.
                       l-hmaxO3
                       45.2 ppb
                       IQR25
Estimated
PM25,PM10,
PM10.25,CO,
NO2, SO2
                     1,2,3,4,     Poisson GAM with LOESS
                      5, 0-4       pre-filter applied to pollution
                                  and hospitalization data.
                                  Sensitivity analyses using co-
                                  adjustment. Results stratified
                                  by season. Ozone effects
                                  significant only in summer.
                                  Ozone effect robust to PM;
                                  not vice versa.
                                                                                1-h max O3 (per 45.2 ppb):

                                                                                Summer:
                                                                                O3 only model:
                                                                                Lag 0-4:  1.348(1.193,1.524)
                                                                                O3 with PM2 5 model:
                                                                                Lag 0-4:  1.330(1.131,1.565)

-------
                                            Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
1
to
o
o
Reference, Study
Location and Period
Canada (cont'd)
Copollutants
Outcomes and Design Mean O3 Levels Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

X

-------
                                          Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
to
o
o
          Reference, Study
        Location and Period
                      Outcomes and Design  Mean O3 Levels
                 Copollutants
                 Considered
                 Lag Structure
                   Examined
Method, Findings,
  Interpretation
         Effects
(Relative Risk and 95% CI)
X
Canada (cont'd)

Yang et al. (2003)
Vancouver, British
Columbia, Canada
Jan 1986-Dec 1998
                              Daily respiratory
                              admissions in children
                              aged < 3 years and
                              adults aged 65+ years.
24-havgO3:
13.41 ppb
SD 66.61
IQR 9.74
CO, NO2, SO2,       1,2,3,4,5     Used bidirectional case-
coefficient                        crossover analysis,
of haze                           comparing air pollution on
                                 day of admission to levels
                                 one week prior and after.
                                 SES evaluated. O3 was
                                 positively associated with
                                 respiratory hospital
                                 admissions among young
                                 children and the elderly.
                       24-h avg O3 (per 9.74 ppb):

                       Odds ratios:
                       Age < 3 years:
                       Lag 4:  1.22(1.15,1.30)
                       Age 65+ years:
                       Lag 4:  1.13(1.09,1.18)
Europe
Anderson etal. (1997)
Five European cities:
London, Paris,
Amsterdam,
Rotterdam, Barcelona
Study periods vary by
city, ranging from
1977-1992




Anderson et al. (1998)
London, England
1987-1992










Emergency COPD
admissions for all ages.
Each city analyzed
previously by
individual teams.
Results combined here
via meta-analysis.





Admissions for asthma
in all ages and age
0-14, 15-64, and
65+ years.









Range across
five cities:

l-hmaxO3
(median):

All year:
36-77 ug/m3
Warm season:
48-91 ug/m3
Cool season:
20-64 ug/m3
8-h max O3:
15. 5 ppb
IQR 13

l-hmaxO3:
20.6 ppb
IQR 16






TSP, SO2, NO2, 0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
BS methodology. Results
stratified by season.
Ozone most consistent and
significant predictor of
admissions. Warm season
effect larger.





SO2,NO2, 0,1,2,0-1, Poisson GLM using APHEA
BS, pollens 0-2 method; co-adjustment.
Ozone significantly
associated with asthma
admissions in the warm
season for all ages and for
age 15-64 years. Warm
season O3 effect robust in
2-pollutant models. Inverse
associations observed in the
cool season for some age
groups.

1-h max O3 (per 50 ug/m3):

Weighted mean effect across
five cities (best lag selected
for each city):

All year:
1.03(1.01,1.05)
Warm season:
1.03(1.01,1.05)
Cool season:
1.01(0.98,1.05)
8-h max O3 (per 10 ppb):

All ages:
Warm season:
Lag 1: 1.022(1.006,1.038)
Cool season:
Lagl: 0.968(0.946,0.992)






-------
1
to
o
o

Reference, Study
Location and Period

1 ilUIC .M.yv/-t ^lUlll U^. ijllCtlS Ul VJ
Copollutants
Outcomes and Design Mean O3 Levels Considered

'3 Ull LWliy AAUS|JI
Lag Structure
Examined

Hill j-VUIIIISSIUIIS
Method, Findings,
Interpretation


Effects
(Relative Risk and 95% CI)

        Atkinson et al.
        (1999b)
        London, England
        1992-1994
                       Total and cause-
                       specific respiratory and
                       cardiovascular
                       admissions in all ages
                       and in all ages and
                       age 0-14, 15-64, and
                       65+ years.
                        17.5 ppb
                        SD11.5
                  NO2, SO2, CO,
                  PM10, BS
0,1,2,3,0-1,    Poisson GLM using APHEA
   0-2, 0-3       methodology. No significant
                associations seen between O3
                and respiratory admissions.
                Ozone was positively
                associated with total
                cardiovascular admissions in
                age 65+ years. Seasonal
                analyses were not conducted.
                             8-h max O3 (per 25.7 ppb):

                             All ages:
                             Total respiratory:
                             Lag 1:   1.012(0.990,1.035)
                             Total cardiovascular:
                             Lag 2:   1.023(1.002,1.046)
X

-------
Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
t—t
1
o
o













X
T]
OJ


o

H
O'
O

O
H
0

0
w
o

o
HH
W

Reference, Study
Location and Period

Europe (cont'd)
Schouten et al. (1996)
Amsterdam and
Rotterdam,
the Netherlands
1977-1989








Hagen et al. (2000)
Drammen, Norway
Nov 1994-Dec 1997




Oftedal et al. (2003)
Drammen, Norway
1995-2000

Ponka and Virtanen
(1996)
Helsinki, Finland
1987-1989




Ballester et al. (2001)
Valencia, Spain
1994-1996
Outcomes and Design


Unscheduled total
respiratory, asthma, and
COPD admissions in
all ages.









Total respiratory
admissions for all ages.





Admissions for
respiratory disease.


Asthma admissions for
age 0-14 years and
15-64 years.





Emergency total
cardiovascular
admissions for all ages.
Mean O3 Levels


l-hmaxO3:

Amsterdam:
Summer:
97 ng/m3
Winter:
62 ng/m3

Rotterdam:
Summer:
96 |ig/m3
Winter:
54 |ig/m3
24-h avg O3:
44.48 ng/m3
SD 18.40
IQR 26.29



24-h avg O3:
44.6 |ig/m3
SD 19.2
IQR 26. 9
O3 (index not
specified):
22 |ig/m3





8-h max O3:
23ppb
Range 5-64
Copollutants
Considered


SO2, NO2, BS












PM10, NO2, SO2,
benzene,
toluene,
formaldehyde



Benzene,
formaldehyde,
toluene, PM10,
NO2, SO2
TSP, S02, N02







SO2, NO2,
CO,BS

Lag Structure Method, Findings,
Examined Interpretation


0,1,2,0-1, Poisson GLM using APHEA
0-2, 0-3, 0-4, methodology; co-adjustment.
0-5 No consistent O3 effects.
Concern regarding multiple
comparisons.








0 Poisson GAM with partial
splines; co-adjustment.
Single and multipollutant
models evaluated. No O3
effects. Ozone levels low
and cycles may not have
been adequately controlled.
0 Benzene had the strongest
association.


0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
methodology. Reported
significant O3 effect for age
0-14 years, but also for
control (digestive disease)
conditions. Ozone levels
very low.

0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
methodology. Results
stratified by season. No O3
Effects
(Relative Risk and 95% CI)


1-h max O3 (per 100 ng/m3):

Amsterdam and Rotterdam:

Total respiratory, all ages:
Summer:
Lag 2: 1.051(1.029,1.073)
Winter:
Lag 2: 0.976(0.951,1.002)




24-h avg O3
(per 26.29 |ig/m3):

LagO: 0.964(0.899-1.033)



24-h avg O3 (per 26.9 |ig/m3):

0.996(0.942,1.053)

Not quantitatively useful.







8-h max O3 (per 5 ppb):

Lag 2: 0.99(0.97-1.01)
                                               effects.

-------
to
o
o
                                            Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
Reference, Study
Location and Period
Outcomes and Design
Mean O3
Levels
Copollutants
Considered
Lag Structure
Examined
Method, Findings,
Interpretation
Effects
(Relative Risk and 95% CI)
        Latin America

        Gouveia and Fletcher
        (2000a)
        Sao Paulo, Brazil
        Nov 1992-Sep 1994
                       Total respiratory,        l-hmaxO3:
                       pneumonia, and asthma   63.4 |ig/m3
                       admissions for age       SD 38.1
                       < 5 years.               IQR 50.3
                                         PM10,N02,
                                         S02, CO
                                     0,1,2       Poisson GLM with co-
                                                 adjustment using sine/cosine
                                                 waves.  Significant O3 effects
                                                 on total respiratory and
                                                 pneumonia admissions. Ozone
                                                 effects fairky robust to NO2
                                                 andPM,n.
                                             l-hmaxO3 (per
                                             119.6|ig/m3):

                                             Total respiratory:
                                             LagO:  1.054(1.003,1.107)
                                             Pneumonia:
                                             LagO:  1.076(1.014,1.142)
                                             Asthma:
                                             Lag 2:  1.011(0.899,1.136)
X
Australia

Morgan etal. (1998a)
Sydney, Australia
1990-1994
        Petroeschevsky et al.
        (2001)
        Brisbane, Australia
        1987-1994
Admissions for asthma
(age 1-14 years, 15-64
years), COPD (age
65+ years), and heart
disease (all ages, 0-64
years, 65+ years).
                       Unscheduled asthma,
                       total respiratory and
                       total cardiovascular
                       admissions in several
                       age strata: all ages, 0-4,
                       5-14,15-64, 65+ years.
l-hmaxO3:
25 ppb
SD13
IQR 11
                                                                       Bscatter, NO2
0,1,2,0-1,
   0-2
                        l-hmaxO3:
                        25.3 ppb
                        Range 2.5-107.3

                        8-h avg O3
                        (10a.m.-
                        6 p.m.):
                        19.0 ppb
                        Range 1.7-64.7
                 Better, S°2,
                 NO,
,1,2,3,0-2,
   0-4
Poisson with GEE.
No significant effects of O3
in single or multipollutant
models.
Poisson GLM using APHEA
co-adjustment methodology.
Results stratified by season.
Ozone significantly related to
asthma and total respiratory
admissions, not for cardiac
admissions. Effects varied by
age group. Ozone effects
robust to copollutants.
1-h max O3 (per 28 ppb):

Asthma, age 1-14 years:
Lag 1:  0.975(0.932,1.019)
Asthma, age 15-64 years:
LagO:  1.025(0.975,1.078)
COPD, age 65+ years:
LagO:  1.010(0.960,1.062)
Heart disease, all ages:
LagO:  1.012(0.990,1.035)

8-h avg O3 (per 10 ppb):

All ages:
Total respiratory:
Lag 2:  1.023(1.003,1.043)
Asthma:
Lag 0-4:  1.090(1.042,1.141)
Total cardiovascular:
Lag 3:  0.987(0.971,1.002)

-------
1
to
o
o
1 ilLMC ±\-t
Reference, Study
Location and Period Outcomes and Design

-V/-t ^lUlll U
Mean O3
Levels

If. ijllCtlS Ul V
Copollutants
Considered

J3 Ull ifail^ AAUS|
Lag Structure
Examined

Illtll j-VUIIIISSIUIIS
Method, Findings,
Interpretation


Effects
(Relative Risk and 95% CI)

        Wongetal. (1999a)
        Hong Kong
        1994-1995
                       Total and cause-
                       specific respiratory and
                       cardiovascular
                       admissions in several
                       age strata: all ages, 0-4,
                       5-64, 65+ years.
                        8-h max O3:
                        20.2 ng/m3
                        Median 24.15
                        IQR31.63
                  N02, S02,
                  PM,n
0,1,2,3,4,5,
 0-1,0-2,0-3,
   0-4, 0-5
Poisson GLM using APHEA
methodology. Ozone
significantly associated with
total and cause specific
respiratory and cardiac
outcomes. Ozone results
robust to adjustment for high
PM10, but not high NO2.
Effects of O3 persisted in
cold season.
8-hmaxO3(perlO|ig/m3):

All ages:
Total respiratory:
Lag 0-3: 1.022(1.015,1.029)
Total cardiovascular:
Lag 0-5: 1.013(1.005,1.021)
X
Wongetal. (1999b)
Hong Kong
Jan 1995-Jun 1997
Total and cause-
specific cardiovascular
admissions in all ages.
O3 (index not      NO2, SO2,        0, 1, 2, 3, 4, 5,    GLM with sinusoids;
specified):         respirablePM     0-1,0-2,0-3,    co-adjustment. Ozone
                                      0-4, 0-5       significantly associated with
Warm season:                                       total and cause-specific
31.2 |ig/m3                                         cardiovascular admissions in
Cool season:                                        cool season only, when O3
34.8 |ig/m3                                         levels are higher in Hong
                                                   Kong. Details missing in brief
                                                   report.
                                               O3 (per 50 ng/m3):

                                               O3 with NO2 models:
                                               Total cardiovascular:
                                               All year:
                                               Lag 0-1:  1.03(1.00,1.07)
                                               Warm season:
                                               Lag 0-1:  1.01 (0.95,1.07)
                                               Cool season:
                                               Lag 0-1:  1.08(1.02,1.14)

-------
1
to
o
o

Reference, Study
Location and Period

i 



O
E>
Tj
H
6
o
21
O
H
O
0
H
W
O
O
H
W
Samet et al. (2000; All cause; 24-h avg O3: PM10, NO2, SO2,
reanalysis Dominici et al., cardiopulmonary Mean range: CO; 2-pollutant
2003) Approximately 12 ppb models
90 U.S. cities (Des Moines, IA) to 36
1 987-1 994 ppb (San Bernardino,
CA)















0,1,2 Poisson GAM 24-h avg O3 (per 1 0 ppb):
(reanalyzed with
stringent Posterior means:
convergence
criteria); Poisson All cause (results given in
GLM graphic format):
All year:
Lag 0: approximately 0.4%
Lag 1: approximately 0.2%
Summer:
Lag 1: approximately 0.5%
Winter:
Lag 1: approximately -0.5%









-------
                                           Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
to
o
o
        Reference, Study
        Location and Period
    Outcome
    Measure
    Mean O, Levels
   Copollutants
    Considered
Lag Structure
  Reported
    Method
     Effect Estimates
X
        United States (cont'd)

        Huang et al. (2004)
        19 U.S. cities
        Jun-Sept 1987-1994
        Schwartz (2004)
        14 U.S. cities
        1986-1993
Cardiopulmonary
24-havgO3:
26ppb
PM10, PM2.5; 2-
pollutant models
  0,1,2,0-6
Poisson GLM;
Bayesian
hierarchical
model
All cause
l-hmaxO3:
Median range:
35.1 ppb (Chicago, IL)
to 60.0 ppb (Provo, UT)
PM10; 2-pollutant
models
                 Case-crossover
                 analysis;
                 controlled for
                 temperature
                 using nonlinear
                 regression splines
                 and matching
24-h avg O3 (per 10 ppb):

Posterior means:

O3 only model:
LagO: 0.73% (0.27, 1.19)
O3 with PM10 model:
LagO: 0.74% (-0.33, 1.72)

O3 only model:
Lag 0-6: 1.25% (0.47, 2.03)
Model adjusted for heat
waves:
Lag 0-6: 1.11%(0.38, 1.86)

l-hmaxO3(10ppb):

Analysis with temperature
regression splines:
All year:
0.19% (0.03, 0.35)
Warm season:
H
6
o
o
H
O

0
H
W
o

O
l— '
H
W
0.26% (0.07, 0.44)
Cold season:
0% (-0.27, 0.27)
Analysis with temperature
matched controls:
All year:
0.23% (0.01, 0.44)
Warm season:
0.37% (0.1 1,0.62)
Cold season:
-0.13% (-0.53, 0.28)



-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
e
to
o
o











>
£j
4^
oo

O
^
'•Tj
H
6


H
0
0
H
W
O
O
H
W
Reference, Study Outcome
Location and Period Measure

United States (cont'd)
Kinney and Ozkaynak All cause;
(1991) respiratory;
Los Angeles County, CA circulatory
1970-1979

Kinney et al. (1995) All cause
Los Angeles County, CA
1985-1990



Ostro(1995) All cause
San Bernardino County and
Riverside County, CA
1980-1986
Fairley (1999; reanalysis All cause;
Fairley, 2003) respiratory;
Santa Clara County, CA cardiovascular
1989-1996














Mean O3 Levels


1-h max total oxidants
(Ox):
75 ppb


l-hmaxO3:
70 ppb




l-hmaxO3:
140 ppb


8-h max O3:
29 ppb

24-havgO3:
16 ppb

O3 ppb-hours > 60 ppb:
Levels not reported.









Copollutants
Considered


KM (particle
optical reflectance),
N02, S02, CO;
multipollutant
models
PM10, N02, S02,
CO; 2-pollutant
models



PM25



PM10,PM25,
PM10.2,, S042-,
coefficient of haze,
NO3", NO2, SO2; 2-
pollutant models












Lag Structure
Reported Method


1 OLS (ordinary
least squares) on
high-pass filtered
variables

1 Linear, log-
linear, and
Poisson



0 Autoregressive
linear; Poisson


0 Poisson GAM
(reanalyzed with
stringent
convergence
criteria);
Poisson GLM












Effect Estimates


All cause:
Multipollutant model:
Slope estimate:
0.03 deaths/ppb (SE 0.009),
p = 0.0005
1-h max O3 (per 143 ppb):

O3 only model:
2% (0, 5)
O3 with PM10 model:
0%(-6, 6)
1 -h max O3 (per 100 ppb):

Warm season:
2.0% (0.0, 5.0)
GLM:

All cause:

8-h max O3 (per 33 ppb):
3.3% (-0.3, 7.0)

O3 ppb-hours > 60 ppb
(increment not reported):
4.1% (1.5, 6.7)








-------
r**
1
to
O
o









Reference, Study
Location and Period

United States (cont'd)
Gamble (1998)
Dallas, TX
1990-1994





1 ilLMC
Outcome
Measure


All cause;
respiratory;
cardiovascular;
cancer; other




J"V^V/-J ^lUlll U^. Hill
Mean O3 Levels


24-h avg O3:

All year:
22 ppb
Summer:
30 ppb
Winter:
12 ppb
CtlS Ul .fY^UlC V^3 1
Copollutants
Considered


PM10, NO2, SO2,
CO; 2-pollutant
models





^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method Effect


1-2 Poisson GLM All cause:

All year:
24-h avg O3
2.7% (0.6, 4

Summer:
24-h avg O3

Estimates





(per 14.7 ppb):
.8)


(per 13.1 ppb):
X
                                                                                                                                           3.5% (p< 0.05)

                                                                                                                                           Winter:
                                                                                                                                           24-h avg O3 (per 7.7 ppb):
                                                                                                                                           2.4% (p > 0.05)
k
4^
VO


M
^^
fe

o
2!
•£—\
O
H
0
C
0
H
W
O
O
H
W
Dockery etal. (1992)
St. Louis, MO and Eastern
Tennessee
1985-1986


Ito and Thurston (1996)
Cook County, IL
1985-1990












All cause





All cause;
respiratory;
circulatory;
cancer;
race/gender
subcategories









24-h avg O3 : PM10, PM2 5, SO42',
H+, NO2, SO2
St. Louis, MO:
22.5 ppb
Eastern Tennessee:
23.0 ppb
l-hmax03: PM10, NO2, SO2,
3 8.1 ppb CO; 2-pollutant
models












1 Poisson with 24-h avg O3 (per 20
GEE
St. Louis, MO:
0.6% (t = 0.38)
Eastern Tennessee:
-1.3%(t=-0.37)
0-1 Poisson GLM 1-h max O3 (per 100

All cause:
O3 only model:
10% (6, 15)
O3 with PM10 model:
7% (1,12)








|ig/m3):





ppb):















-------
3
                                           Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
to
o
o
        Reference, Study
        Location and Period
Outcome
Measure
Mean O3 Levels
Copollutants
 Considered
Lag Structure
  Reported
Method
                                                                                                                                 Effect Estimates
United States (cont'd)

Lippmann et al. (2000;
reanalysis Ito, 2003)
Detroit, MI
1985-1990
1992-1994
                                  All cause;
                                  respiratory;
                                  cardiovascular;
                                  cause-specific
               24-havgO3:
               25ppb
                    PM10,PM25,
                    PM10.25,S042-,H+,
                    N02, S02, CO;
                    2-pollutant models
                   0,1,2,3,0-1,
                     0-2, 0-3
                Poisson GAM
                (reanalyzed with
                stringent
                convergence
                criteria); Poisson
                GLM
              24-h avg O3 (per 5th to
              95th % increment):

              GAM with stringent
              convergence criteria:

              For all lags and outcomes
              during both study periods
              (n= 140):
              Median 1.6%
              Range-1.8-2.6
X
Lipfert et al. (2000a)
Seven counties in
Philadelphia, PA area
1991-1995




Moolgavkar et al. (1995)
Philadelphia, PA
1973-1988









All cause;
respiratory;
cardiovascular;
all ages; age 65+
years; age < 65
years; various
subregional
boundaries
All cause











l-hmaxO3:
44.76 ppb

24-h avg O3:
23. 44 ppb



24-h avg O3:

Spring:
25. 9 ppb
Summer:
35.5 ppb
Fall:
16.2 ppb
Winter:
11. 9 ppb


PM10,PM25,
PM10.2.5, S042-,
NCV, other PM
indices, NO2, SO2,
CO; 2-pollutant
models


TSP, SO2;
multipollutant
models









0-1 Linear with 19-
day weighted
average
Shumway filters




1 Poisson; GEE
and
nonparametric
bootstrap
methods







1-h max O3 (per 45 ppb less
background, not reported):

All cause, all ages:
O3 only model:
3.19%, p< 0.055
O3 with PM2 5 model:
3. 34%, p< 0.055
24-h avg O3 (per 100 ppb):

O3 with TSP and SO2
models:
Spring:
2.0% (-6.7, 11.5)
Summer:
14.9% (6.8, 23.6)
Fall:
-4. 5% (-13. 9, 5.9)
Winter:
0.4% (-15. 6, 19.4)

-------
1
to
o
o

Reference, Study
Location and Period

1 ilLMC t
Outcome
Measure

l.yv/-j ^lUlll U). ijllCtlS Ul .fY^UlC V^3 1
Copollutants
Mean O3 Levels Considered

^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method


Effect Estimates

        Chock et al. (2000)
        Pittsburgh, PA
        1989-1991
All cause; age
< 74 years;
age 75+ years
l-hmaxO3:
Levels not reported.
PM10, NO2, SO2,
CO; 2-, 5-, and
6-pollutant models
             Poisson GLM
                   1-h max O3 (per 40 ppb):

                   Age <74 years:
                   O3 only model:
                   -1.5% (t=-0.68)
                   O3 with PM10 model:
                   -2.0% (t=-0.93)

                   Age 75+ years:
                   O3 only model:
                   -1.8% (t=-0.82)
                   O3 with PM10 model:
                   -2.2% (t=-0.98)
X
        De Leon et al. (2003)
        New York City
        1985-1994
Circulatory and
cancer with and
without
contributing
respiratory causes
24-havgO3:
21.59ppb
PM10, NO2, SO2,
CO; 2-pollutant
models
Oorl
Poisson GAM;
Poisson GLM
Quanititative results not
given.

Circulatory deaths:
Larger O3 effect estimates
with contributing respiratory
causes than without (RS.
non-significant).

Cancer deaths:
Smaller O3 effect estimates
with contributing respiratory
causes than without (PJl
non-significant).

-------
                                  Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
X
to
H

6
o


o
H

O


O
H
W

O


O
HH
H
W
Reference, Study
Location and Period
United States (cont'd)
Klemm and Mason (2000);
Klemm et al. (2004)
Atlanta, GA
Aug 1998- July 2000


Canada
Vedal et al. (2003)
Vancouver, British
Columbia, Canada
1994-1996


Goldberg et al. (2003)
Montreal, Quebec, Canada
1984-1993


Outcome
Measure Mean O3 Levels

All cause; 8-h max O3:
respiratory; 47.03 ppb
cardiovascular;
cancer; other; age
< 65 years; age
65+ years



All cause; 1 -h max O3 :
respiratory; 27.3 ppb
cardiovascular



Congestive heart 24-havgO3:
failure as 29 |ig/m3
underlying cause
of death versus
those classified as
having congestive
heart failure one
year prior to death

Copollutants Lag Structure
Considered Reported Method Effect Estimates

PM25,PM10.25,EC, 0-1 PoissonGLM All cause, age 65+ years:
OC, NO2, SO42~, using quarterly,
NO3', SO2, CO monthly, or Quarterly knots:
biweekly knots Slope estimate:
for temporal 0.00079 (SE 0.00099),
smoothing t = 0.80
Monthly knots:
Slope estimate:
0.00136 (SE 0.001 11),
t=1.22

PM10, NO2, SO2, 0,1,2 PoissonGAM 1-h max O3 (per 10.2 ppb):
CO
All cause:
Summer:
LagO: 4.2% (1.4, 7.0)
Winter:
LagO: 0.5% (-1.9, 3.0)
TSP, coefficient of 0-2 PoissonGLM 24-h avg O3 (per 2 1.3
haze, PM10, SO42~, Hg/m3):
S02,N02,CO
Congestive heart failure as
underlying cause of death:
4.54% (-5.64, 15.81)
Having congestive heart
failure one year prior to
death:
2.34% (-1.78, 6.63)

-------
1
to
o
o

Reference, Study
Location and Period

1 ilLMC t
Outcome
Measure

l.yv/-j ^lUlll U). ijllCtlS Ul .fY^UlC V^3 1
Copollutants
Mean O3 Levels Considered

^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method


Effect Estimates

X
        Gryparis et al. (2004)
        23 European cities
        Study periods vary by city,
        ranging from 1990-1997
All cause;
respiratory;
cardiovascular
l-hmaxO3:
Median range:

Summer:
44 ppb (Tel Aviv, Israel)
to 117 ppb (Torino,
Italy)

Winter:
11 ppb (Milan, Italy) to
57 ppb (Athens, Greece)

8-h max O3:
Median range:

Summer:
30 ppb (Rome, Italy) to
99 ppb (Torino, Italy)

Winter:
8 ppb (Milan, Italy) to
49 ppb (Budapest,
Hungary)
PM10, NO2, SO2,
CO; 2-pollutant
models
0-1
Poisson GAM;
Bayesian
hierarchical
model
8-h max O3 (per 10 ug/m3):

Weighted mean effect
across 21 cities with 8-h
max O3 concentrations:

Random effects model:

All cause:

All year:
0.03% (-0.18, 0.21)

Summer:
O3 only model:
0.31% (0.17, 0.52)
O3 with PM10 model:
0.27% (0.08, 0.49)

Winter:
O3 only model:
0.12% (-0.12, 0.37)
O3 with PM10 model:
0.22% (-0.08, 0.51)
Touloumi et al. ( 1 997) All cause
Four European cities:
London, Paris
Barcelona, Athens
Study periods vary by city,
ranging from 1986-1992






l-hmaxO3:

London:
4 1.2 ug/m3
Paris:
46.1 ug/m3
Barcelona:
72.4 ug/m3
Athens:
93.8 ug/m3


BS,NO2; 0,1,2,3,0-1, Poisson l-hmaxO3 (50 ug/m3):
2-pollutant models 0-2, 0-3 autoregressive
Weighted mean effect
across four cities (best lag
selected for each city):

Random effects model:

O3 only model:
2. 9% (1.0, 4.9)
O3 with BS model:
2. 8% (0.5, 5.0)

-------
^**

to
o
o













X
r N
L/l
[\

O

£5
H
6
0
2;
o
H
0
0
w
o
o
H
w
1 ilLMC
Reference, Study Outcome
Location and Period Measure

Europe (cont'd)
Zmirouetal. (1998) Respiratory;
Four European cities: cardiovascular
London, Paris, Lyon,
Barcelona
Study periods vary by city,
ranging from 1985-1992









Anderson et al. ( 1 996) All cause;
London, England respiratory;
1987-1992 cardiovascular















J"V^V/-J ^lUlll U^. Hill

Mean O3 Levels


8-h avg O3 (9 a.m.-
5 p.m.):

London:
Cold: 2 1.0 ug/m3
Warm: 40.8 ug/m3
Paris:
Cold: 11.5 ug/m3
Warm: 42.7 ug/m3
Lyon:
Cold: 2 1.0 ug/m3
Warm: 40.8 ug/m3
Barcelona:
Cold: 51.5 ug/m3
Warm: 89.7 ug/m3
l-hmaxO3:
20.6 ppb

8-h avg O3 (9 a.m.-
5p.m.): 15. 5 ppb













ctis ui .rvtuic v^3 tiA|ju»uic uii IVAUI liiiuy
Copollutants Lag Structure
Considered Reported Method Effect Estimates


BS, TSP, SO2, NO2 0,1,2,3,0-1, Poisson GLM 8-h avg O3 (per 50 ug/m3):
0-2, 0-3
Weighted mean effect
across four cities (best lag
selected for each city):

Random effects model:

Respiratory:
5% (2, 8)
Cardiovascular:
2% (0, 3)



BS,NO2, SO2; 0 Poisson GLM All cause:
2-pollutant models
All year:
1-h max O3 (per 31 ppb):
2.59% (1.30, 3.89)

Warm season:
1-h max O3 (per 34 ppb):
3.49% (1.81, 5.20)

Cool season:
1-h max O3 (per 26 ppb):
0.99% (-0.80, 2.81)






-------
^**

to
o
o











^
X
^
£

O

>
H
1
O
0
2|
O
H

O
H
W
O
O
H
W

Reference, Study
Location and Period

Europe (cont'd)
Bremneretal. (1999)
London, England
1992-1994






Prescott etal. (1998)
Edinburgh, Scotland
1992-1995



Dab etal. (1996)
Paris, France
1987-1992




Zmirou etal. (1996)
Lyon, France
1985-1990








1 ilLMC /
Outcome
Measure


All cause;
respiratory;
cardiovascular; all
cancer; all others;
all ages; age
specific (0-64,
65+, 65-74,
75+ years)

All cause;
respiratory;
cardiovascular; all
ages, age < 65
years, age 65+
years
Respiratory






All cause;
respiratory;
cardiovascular;
digestive







^^V/-J ^lUlll U^. Hill

Mean O3 Levels


l-hmaxO3:
22.6 ppb

8-h max O3:
17.5 ppb




24-havgO3:
14.5 ppb




l-hmaxO3:
23.2 ng/m3

8-h max O3:
11.5 |ig/m3


l-hmaxO3:
15.23 ng/m3

8-h avg O3
(9 a.m.-5 p.m.):
9.94 ng/m3





CtlS Ul .fY^UlC V^3 HiA|JU»UIC Ull 1VAUI lillllV
Copollutants Lag Structure
Considered Reported Method Effect Estimates


BS, PM10,NO2, Selected best Poisson GLM 8-h max O3 (per 26 ppb):
S02,CO; fromO, 1,2, 3,
2-pollutant models (all cause); All ages:
0,1,2,3,0-1, All cause:
0-2,0-3 Lag 2: -0.7% (-2. 3, 0.9)
(respiratory, Respiratory:
cardiovascular) Lag 2: -3. 6% (-7.7, 0.8)
Cardiovascular:
Lag 2: 3. 5% (0.5, 6.7)
BS,PM10,NO2, 0 Poisson GLM 24-h avg O3 (per 10 ppb):
S02, CO; 2-
pollutant models All cause, all ages:
-4.2% (-8.1, -0.1)


BS,PM13,NO2, 0 Poisson l-hmaxO3 (per 100 |ig/m3):
SO2, CO autoregressive 1 .074 (0.934, 1 .235)

8-h max O3 (per 100 |ig/m3):
1.040(0.934,1.157)


PM13, SO2,NO2 Selected best Poisson GLM 8-h avg O3 (per 50 |ig/m3):
fromO, 1,2, 3
All cause:
LagO: 3% (-5, 12)
Respiratory:
Lagl: l%(-8, 10)
Cardiovascular:
Lagl: 0%(-11,12)




-------
                                            Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
to
o
o
        Reference, Study
        Location and Period
                               Outcome
                               Measure
                       Mean O, Levels
                           Copollutants
                            Considered
                     Lag Structure
                       Reported
    Method
                                    Effect Estimates
X
ON
o
o
2
o
H
O
o
H
W
O
O
HH
H
W
        Europe (cont'd)

        Sartor etal. (1995)
        Belgium
        Summer 1994
Hoek et al, (2000;
reanalysis Hoek, 2003)
The Netherlands: entire
country, four urban areas
1986-1994
                          All cause; age
                          < 65 years;
                          age 65+ years
All cause; COPD;
pneumonia;
cardiovascular
                   24-havgO3:

                   During heat wave (42
                   day period):  72.4 |ig/m3

                   Before heat wave (43
                   day period):  52.4 |ig/m3

                   After heat wave (39 day
                   period): 38.6 |ig/m3
8-h avg O3 (12 p.m.-
8 p.m.):
Median: 47 |ig/m3
                        TSP, NO, NO2, SO2
PM10, BS, SO42%
N03-, N02, S02,
CO; 2-pollutant
models
0,1,2
                                                                                                        1,0-6
Log-linear
regression
Poisson GAM
(reanalyzed
with stringent
convergence
criteria);
Poisson GLM
                                                        No individual regression
                                                        coefficient for O3 alone;
                                                        interaction with temperature
                               24-havgO3(froml8.8to
                               111.5 |ig/m3)and
                               temperature (from 10.0 to
                               27.5°C):

                               Age < 65 years:
                               Lag 1: 16% increase in
                               mortality (5.3% expected)
                               Age 65+ years:
                               Lag 1:  36.5% increase in
                               mortality (4% expected)

                               GLM:

                               All cause:

                               8-h avg O3 (per 150 |ig/m3):
                               Lagl:  4.3% (2.4, 6.2)

                               8-h avg O3 (per 120 |ig/m3):
                               Lag 0-6: 5.9% (3.1, 8.7)

-------
r**
1
to
O
o









Reference, Study
Location and Period

Europe (cont'd)
Hoeketal. (2001;
reanalysis Hoek, 2003)
The Netherlands
1986-1994




1 ilLMC t
Outcome
Measure


Total
cardiovascular;
myocardial
infarction;
arrhythmia; heart
failure;
cerebro vascular;
thrombosis-related
Y^V/-J ^Ulll U^. ijllCtlS Ul .fY^UlC \J^ I
Copollutants
Mean O3 Levels Considered


8-havgO3(12p.m.- PM10, NO2, SO2,
8 p.m.): CO
Median: 47 |ig/m3





^A|JUSUI C Ull 1TAUI lillll V
Lag Structure
Reported Method


1 Poisson GAM
(reanalyzed with
stringent
convergence
criteria); Poisson
GLM



Effect Estimates


8-h avg O3 (per 150 |ig/m3):

GLM:

Total cardiovascular:
6.2% (3.3, 9.2)
Myocardial infarction:
4. 3% (0.1, 8.6)
X
Arrhythmia:
11.4% (-1.2, 25.5)
Heart failure:
10.2% (1.2, 19.9)
Cerebrovascular:
9.1% (2.9, 15.7)
Thrombosis-related:
16.6% (2.8, 32.2)


L
>
H
O'
O

O
H
0
"

O
x^
H
W
O
O
H
W
Roemer and van Wijinen
(2001)
Amsterdam, the
Netherlands
1987-1998




Verhoeffetal. (1996)
Amsterdam, the
Netherlands

1986-1992





All cause 8-h max O3:

Background sites:
43 lig/m3
Traffic sites:
36 |ig/m3



All cause; all ages; 1-h max O3:
age 65+ years 43 |ig/m3








BS, PM10, NO2, 1 , 2, 0-6 Poisson GAM
SO2, CO (default
convergence
criteria but with
only one
smoother)



PM10, NO2, SO2, 0,1,2 Poisson
CO; multipollutant
models







8-h max O3 (per 100 |ig/m3):

Total population using
background sites:
Lagl: -0.3% (-4. 1,3.7)
Total population using
traffic sites:
Lagl: 0.2% (-3.6, 4.2)

1-h max O3 (per 100 |ig/m3)

All ages:

LagO: 1.8% (-3. 8, 7.8)
Lagl: 0.1% (-4.7, 5.1)
Lag 2: 4.9% (0.1, 10.0)




-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
O
C2
v\













X
(!/i
00

O

>
H
M
^^
0

O
H
0
crj
O
H
W
O
O
HH
H
W



Reference, Study
Location and Period
Europe (cont'd)


Peters et al. (2000b)
NE Bavaria, Germany and
coal basin in Czech
Republic
1982-1994



Ponkaetal. (1998)
Helsinki, Finland
1987-1993






Garcia-Aymerich et al.
(2000)
Barcelona, Spain
1985-1989




Saezetal. (1999)
Barcelona, Spain
1986-1989


Sunyeretal. (1996)
Barcelona, Spain
1985-1991






Outcome
Measure



All cause for
Czech Republic;
all cause and
cardiovascular for
Bavaria, Germany



All cause;
cardiovascular;
age < 65 years,
age 65+ years





All cause;
respiratory;
cardiovascular;
general
population;
patients with
COPD

Asthma mortality;
age 2-45 years



All cause;
respiratory;
cardiovascular;
all ages;
age 70+ years




Mean O3 Levels



24-h avg O3:

Czech Republic:
40.3 ng/m3
Bavaria, Germany:
38.2 ng/m3


24-h avg O3:
Median: 18 |ig/m3







l-hmaxO3:
Levels not reported.






l-hmaxO3:
Levels not reported.



l-hmaxO3:
Summer:
86.5 ng/m3
Winter:
55.2 |ig/m3



Copollutants Lag Structure
Considered Reported Method Effect Estimates



TSP, PM10, NO2, 0, 1, 2, 3 Poisson GLM 24-h avg O3 (per
SO2,CO 100ng/m3):

All cause:
Czech Republic:
Lag 2: 7.8% (-1.8, 18.4)
Bavaria, Germany:
LagO: 8.2% (0.4, 16.7)
TSP, PM10, NO2, 0, 1 , 2, 3, 4, 5, Poisson GLM 24-h avg O3 (per 20 |ig/m3):
S02 6, 7
All cause, age < 65 years:
Not significant, values not
reported.
Cardiovascular, age < 65
years:
Lag 5: -11.7% (-18.9, -3.9)
Lag 6: 9.9% (1.1, 19.5)
BS,NO2, SO2, 5 (general Poisson GLM l-hmaxO3 (per 50 |ig/m3):
population);
3 (COPD cohort) General population:
2.4% (0.6, 4.2)
COPD patients:
4.0% (-4.7, 13.4)


BS,NO2, SO2, 0-2 Poisson with Slope estimate:
GEE 0.021 (SE 0.01 l),p = 0.054



BS, SO2, NO2 0, 1, 5 Autoregressive 1-h max O3 (per 100 |ig/m3):
Poisson
All cause, all ages:
All year:
LagO: 4.8% (1.2, 8.6)
Summer:
LagO: 5. 8% (1.7, 10.1)
Winter:
LagO: 2.6% (-3. 5, 9.1)

-------
X
VO
Diaz etal. (1999)
Madrid, Spain
1990-1992
                                           Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o








Reference, Study
Location and Period

Europe (cont'd)
Sunyer and Basagana
(2001)
Barcelona, Spain
1990-1995
Sunyer et al. (2002)
Barcelona, Spain
1986-1995

Outcome
Measure


Mortality in a
cohort of patients
with COPD

Mortality in a
cohort of patients
with severe
asthma
Mean O3 Levels


l-hmaxO3:
Mean not reported
IQR21 ng/m3

l-hmaxO3:
Median: 69.3 |ig/m3

8-h max O3:
Copollutants
Considered


PM10, NO2, CO



PM10, BS, SO2,
N02, CO, pollen


Lag Structure
Reported Method


0-2 Conditional
logistic (case-
crossover)

0-2 Conditional
logistic (case-
crossover)

Effect Estimates


1-hmax O3(per21 jig/m3):

Odds ratio:
0.979(0.919,1.065)
1-hmax O3 (per 48 jig/m3):

Odds ratios:
Patients with only one
All cause;
respiratory;
cardiovascular
                                                     Median: 54.4 |ig/m3
                                                     24-havgO3:
                                                     Levels not reported.
                                           TSP, NO,, SO,, CO
                                                 1,4,10
                                      Autoregressive
                                      linear
                               admission:
                               1.096(0.820,1.466)
                               Patients with more than
                               one admission:
                               1.688(0.978,2.643)

                               24-h avg O3 (per 25 |ig/m3):
                                                                                                                                      For O3 levels higher than 35
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
        Latin America

        Borja-Aburto et al. (1997)
        Mexico City
        1990-1992
All cause; all ages;
age < 5 years; age
> 65  years
l-hmaxO3:
Median 155 ppb

8-h max O3:
Median 94 ppb

10-h avg O3 (8a.m.-
6 p.m.):
Median 87 ppb

24-h avg O3:
Median 54 ppb
TSP, SO2, CO;
2-pollutant models
0,1,2
                                                                                                            Poisson
                                                                                                            iteratively
                                                                                                            weighted and
                                                                                                            filtered least-
                                                                                                            squares method
                                                                                                                                      All cause:
                                                                                                                                      Lag 4:  12% (p < 0.01)

                                                                                                                                      U-shaped (quadratic) O3-
                                                                                                                                      mortality relationship with a
                                                                                                                                      minimum of 35 |ig/m3.
                                                                                                    1-h max O3 (per 100 ppb):

                                                                                                    All ages:
                                                                                                    O3 only model:
                                                                                                    LagO: 2.4% (1.1, 3.9)
                                                                                                    O3 with TSP model:
                                                                                                    LagO: -1.8% (-10.0, 6.4)

-------
         Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o












Reference, Study
Location and Period
Latin America (cont'd)
Borja-Aburto et al. (1998)
SW Mexico City
1993-1995









Outcome
Measure

All cause;
respiratory;
cardiovascular;
other; all ages;
age > 65 years







Mean O3 Levels

l-hmaxO3:
163 ppb

24-h avg O3:
44 ppb







Copollutants Lag Structure
Considered Reported Method

PM25, NO2, SO2; 0, 1, 2, 3, 4, 5, Poisson GAM
2-pollutant models 1-2 (default
convergence
criteria but with
only one
smoother)






Effect Estimates

24-h avg O3 (per 10 ppb):

All cause, all ages:
Lag 1-2: 0.6% (-0.3, 1.5)
All cause, age > 65 years:
Lag 1-2: 0.8% (-0.4, 2.0)
Respiratory, all ages:
Lag 1-2: -0.7% (-3.6, 2.1)
Cardiovascular, all ages:
Lag 1-2: 1.8% (0.1, 3.5)
Other noninjury, all ages:
Lag 1-2: 0.3% (-0.9, 1.4)
^      O'Neill et al. (2004)
Jx;      Mexico City
T-"      1996-1998
All cause; all ages;
age 65+ years;
SES gradient
24-h avg O3:
35.3 ppb
PM,,
0-1
Poisson GAM
24-h avg O3 (per 10 ppb):

All ages:
0.65% (0.02, 1.28)
Age 65+ years:
1.39% (0.51, 2.28)

SES gradient did not show
any consistent pattern.
H
o
^^
o
H
0
c|
o
H
W
O
^
O
H
W


Tellez-Rojo et al. (2000)
Mexico City
1994






Gouveia and Fletcher
(2000b)
Sao Paulo, Brazil
1991-1993




Respiratory ; 1 -h max O3 :
COPD mortality; 1 34.5 ppb
age 65+ years;
within medical
unit; outside of
medical unit



All ages (all l-hmaxO3:
cause); age 67.9 |ig/m3
< 5 years (all
cause, respiratory,
pneumonia); age
65+ years (all
cause, respiratory,
cardiovascular)
PM10, N02, S02 1, 2, 3, 4, 5, 1-3, Poisson,
1-5, 1-7 iteratively
weighted and
filtered least-
squares method




PM10, NO2, SO2, 0, 1, 2 Poisson GLM
CO






1-h max O3 (per 40 ppb):

Outside medical unit:
Respiratory:
Lag 1-5: 14.0% (4.1, 24.9)
COPD mortality:
Lag 1-5: 15. 6% (4.0, 28.4)


1-h max O3 (per 106 |ig/m3):

All cause, all ages:
LagO: 0.8%(-1. 1,2.7)
All cause, age 65+ years:
LagO: 2.3% (0,4.6)



-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o











X
7"1
^

o

>
H
o
0

o
H
0

H
W
O
^
O

w
Reference, Study
Location and Period
Latin America (cont'd)
Pereiraetal. (1998)
Sao Paulo, Brazil
1991-1992
Saldivaetal. (1994)
Sao Paulo, Brazil
1990-1991
Saldivaetal. (1995)
Sao Paulo, Brazil
1990-1991





Cifuentes et al. (2000)
Santiago, Chile
1988-1966







Ostroetal. (1996)
Santiago, Chile
1989-1991






Outcome
Measure Mean O3 Levels

Intrauterine 1 -h max O3 :
mortality 67.5 \igfm3

Respiratory; age 24-havgO3:
< 5 years 12. 14 ppb

All cause; age l-hmaxO3:
65+ years 38.3 ppb

24-havgO3:
12.5 ppb



All cause l-hmaxO3:

Summer:
108.2 ppb






All cause l-hmaxO3:
52.8 ppb







Copollutants
Considered

PM10, N02, S02,
CO

PM10, N02, S02,
CO; multipollutant
models
PM10, NO2, SO2,
CO; 2-pollutant
models





PM25,PM10.25, CO,
SO2,NO2








PM10, NO2, SO2; 2-
pollutant models







Lag Structure
Reported Method

0 Poisson,
linear with
M-estimation
0-2 OLS of
transformed data

0-1 OLS; Poisson
with GEE






0, 1, 2, 3, 4, 5, Poisson GAM
1-2, 1-3, 1-4, 1-5 (default
convergence
criteira);
Poisson GLM





1 OLS, several
other methods







Effect Estimates

Slope estimate:
0.0000 (SE 0.0004)

Slope estimate:
0.01048 deaths/day/ppb
(SE 0.02481 ),p = 0.673
Slope estimate:

l-hmaxO3:
0.0280 deaths/day/ppb
(SE 0.0213), p> 0.05
24-havgO3:
0.0093 deaths/day/ppb
(SE 0.0813), p> 0.05
1-h max O3 per (108.2 ppb):

GLM:

Summer:
O3 only model:
Lag 1-2: 0.3%(t=0.3)
Multipollutant model:
Lag 1-2: -0.1% (t= -0.1)

All year:
1-h max O3 (per 52.8 ppb):
-3% (-4, -2)

Summer:
1 -h max O3 (per 100 ppb):
4% (0, 9)



-------
                                 Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
to
o
o
X
to
Reference, Study
Location and Period
Australia
Morgan etal. (1998b)
Sydney, Australia
1989-1993





Simpson etal. (1997)
Brisbane, Australia
1987-1993






Asia
Kim et al. (2004)
Seoul, Korea
1995-1999










Outcome
Measure Mean O3 Levels

All cause; 1 -h max O3 :
respiratory; 24 ppb
cardiovascular





All cause; 8-h avg O3
respiratory; (10 a.m.-6 p.m.):
cardiovascular;
all ages; age All year:
< 65 years; 18.1 ppb
age 65+ years Summer:
20.2 ppb
Winter:
16.1 ppb

All cause l-hmaxO3:

All year:
35. 16 ppb
Summer:
46.87 ppb
Winter:
2 1.26 ppb





Copollutants
Considered

PMby
nephelometer, NO2;
multipollutant
models




PM10,PMby
nephelometer, NO2,
SO2, CO







PM10, NO2, SO2,
CO; 2-pollutant
models










Lag Structure
Reported Method

0 Poisson with
GEE






0 Autoregressive
Poisson with
GEE







1 Poisson GAM
(linear model);
GLM with cubic
natural spline;
GLM with B-
mode spline
(threshold model)







Effect Estimates

l-hmaxO3(per28ppb):

All cause:
2.04% (0.37, 3.73)
Respiratory:
-0.84% (-7.16, 5.91)
Cardiovascular:
2.52% (-0.25, 5.38)
8-h avg O3 (per 10 ppb):

All cause, all ages:
All year:
2.4% (0.8, 4.0)
Summer:
3.0% (1.0, 5.0)
Winter:
1.3% (-1.4, 4.1)

1-h max O3 (per 21 .5 ppb):

All year:
Linear model:
2.6% (1.7, 3.5)
Threshold model:
3.4% (2.3, 4.4)

Summer:
Linear model:
1.9% (0.5, 3.3)
Threshold model:
3. 8% (2.0, 5.7)

-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
t—t
1
to
o
o











>
X
a\
OJ


O
!>
-n
H
6
0
2|
O
H
O

0
H
W
o

O
HH
H
W
Reference, Study Outcome
Location and Period Measure

Asia (cont'd)
Lee et al. (1999) All cause
Seoul and Ulsan, Korea
1991-1995



Lee and Schwartz (1999) All cause
Seoul, Korea
1991-1995






Tsai et al. (2003) All cause;
Kaohsiung, Taiwan respiratory;
1994-2000 cardiovascular;
tropical area







Yang et al. (2004) All cause;
Taipei, Taiwan respiratory;
1994-1998 cardiovascular;
subtropical area





Copollutants
Mean O3 Levels Considered


l-hmaxO3: TSP, SO2

Seoul:
32.4 ppb
Ulsan:
26.0 ppb
l-hmaxO3: TSP, SO2

Seoul:
32.4 ppb





24-h avg O3: PM10, SO2, NO2,
23. 6 ppb CO









24-h avg O3: PM10, SO2, NO2,
17.18 ppb CO







Lag Structure
Reported Method Effect Estimates


0 Poissonwith 1-h max O3 (per 50 ppb):
GEE
Seoul:
1.5% (0.5, 2.5)
Ulsan:
2.0% (-11. 1,17.0)
0 Conditional 1-h max O3 (per 50 ppb):
logistic (case-
crossover with Two controls, plus and
bidirectional minus one week:
control sampling) 1.5% (-1.2,4.2)
Four controls, plus and
minus two weeks:
2. 3% (-0.1, 4. 8)

0-2 Case-crossover 24-h avg O3 (per 19.2 ppb):
analysis
Odds ratios:
All cause:
0.994(0.995,1.035)
Respiratory:
0.996(0.848,1.169)
Cardiovascular:
1.005(0.919,1.098)


0-2 Case-crossover 24-h avg O3 (per 9.34 ppb):
analysis
Odds ratios:
All cause:
0.999(0.972-1.026)
Respiratory:
0.991(0.897-1.094)
Cardiovascular:
1.004(0.952-1.058)

-------
                                           Table AX7-6.  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
   Mean O3 Levels
               Study Description
                                                                                            Results and Comments
X
        United States

        Galizia and Kinney
        (1999; expo sure data
        Kinney et al., 1998)
        U.S. nationwide
        1995
l-hmaxO3:
10-year mean Jun-Aug:
61.2 ppb
SD15.5
Range 13-185
        Goss et al. (2004)
        U.S. nationwide
        1999-2000
l-hmaxO3
Sl.Oppb
SD7.3
                        Nationwide sample of 520 young adults.  Subjects
                        were nonsmokers, aged 17-21 years, 50% males.
                        Each subject provided one spirometric lung function
                        measurement in the spring of their 1st year at Yale
                        College in New Haven, CT, and completed a
                        questionnaire addressing residential history,
                        respiratory diseases, and activity patterns. Long-
                        term O3 exposure was treated as a high/low
                        dichotomous variable, with subjects assigned to the
                        high O3 category if they lived for 4+ years in
                        counties with 10-year summer mean O3 levels
                        greater than 80 ppb. Four lung function variables
                        (FVC, FEY^ FEF25.75, FEF75) were regressed on
                        O3 exposure, controlling for age, height, height
                        squared, sex, race, parental education, and maternal
                        smoking history.  Respiratory symptom histories
                        (cough, phlegm, wheeze apart from colds,
                        and composite index for any  of the three symptoms)
                        were logistically regressed on O3 exposure,
                        controlling for sex, race, parental education, and
                        maternal smoking.
11,484 cystic fibrosis patients over the age of
6 years. Exposure to O3, PM2 5, PM10, NO2, SO2,
and CO assessed by linking Aerometric Information
Retrieval System with patients' home zip code.
Studied exacerbation and lung function.  Mortality
was also of interest, but study was underpowered to
examine this outcome.   Logistic regression models
were used to analyze the exacerbations and multiple
linear regression was used to study lung function.
O3 monitoring season and regional effects also were
examined.
                                                 Significant decrements in FEV[ and FEF25.75 in relation to O3
                                                 exposure were observed for all subjects and for males alone, but
                                                 not for females alone. Similar patterns observed for FVC and
                                                 FEF75, but not with statistical significance.

                                                 % difference in lung function for high versus low O3 exposure
                                                 groups:
                                                                         All subjects: -3.07% (-0.22, -5.92)
                                                                         Females: -0.26% (3.79, -4.31)
                                                                         Males: -4.71% (-0.66, -8.76)

                                                                         FEF25.75:
                                                                         All subjects: -8.11% (-2.32, -13.90)
                                                                         Females: -1.96% (6. 39, -10.30)
                                                                         Males: -13.02% (-4.87, -21.17)

                                                                         Wheeze and respiratory symptom index were significantly elevated
                                                                         for high O3 exposure group.

                                                                         Odds ratios for symptoms:

                                                                         Wheeze: 1.97(1.06,3.66)
                                                                         Respiratory symptom index: 2.00 (1.15, 3.46)

                                                                         Ozone may increase the risk for pulmonary exacerbations in cystic
                                                                         fibrosis patients.

                                                                         Odds ratios for two or more exacerbations (per 10 ppb increase in
                                                                         1-hmax O3):

                                                                         O3 only model: 1.10 (1.03, 1.17)
                                                                         O3 withPM25 model: 1.08 (1.01, 1.15)

                                                                         PM2 5, but not O3, was significantly associated with declines in
                                                                         lung function in these patients.

-------
                                     Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                          Mean O3 Levels
Study Description
                                                                                           Results and Comments
X
United States (cont'd)

Kinney and Lippmann
(2000)
Fort Sill, OK;
Fort Leonard Wood,
MO; Fort Dix, NJ;
Fort Benning, GA;
West Point, NY
Apr-Sep 1990
        Greeretal. (1993)
        California
        1973-1987
l-hmaxO3:
Mean during 5-week
summer training
period:

Fort Benning, GA:
55.6 ppb
(0 hours O3 > 100 ppb)

Fort Dix, NJ:
71.3 ppb
(23 hours O3> 100
ppb)

Fort Leonard Wood,
MO:
55.4 ppb
(1 hours O3 > 100 ppb)

Fort Sill, OK:
61.7 ppb
(1 hours O3 > 100 ppb)

Annual mean O3:
Levels not reported.
                                                       72 nonsmoking students (mean age 20.25 years) at
                                                       the U.S. Military Academy at West Point, NY were
                                                       measured for lung function and respiratory
                                                       symptoms before (Apr) and after (Aug-Sep) taking
                                                       part in an intensive, largely outdoor, summer
                                                       training over five weeks (Jul 11-Aug 15) at four
                                                       U.S. military bases. Ozone levels in the Fort Dix,
                                                       NJ area were consistently higher than at the three
                                                       remaining three locations. Analysis assessed
                                                       change in lung function and respiratory symptoms
                                                       measured before and soon after the summer
                                                       training, and examined whether adverse trends
                                                       would be  more pronounced in students exposed to
                                                       higher O3 levels during summer training.
                                               3,914 nonsmoking adults aged 25+ years at
                                               enrollment in 1977 completed questionnaires at two
                                               time points, 1977 and 1987. To be eligible, subjects
                                               had to have lived 10 or more years within 5 miles of
                                               current residence. Residential histories used to
                                               interpolate air pollution levels to zip centroids  over
                                               a 20-year period (1966-1987). New asthma cases
                                               defined as answering yes to doctor diagnosed
                                               asthma at 1987 followup among those answering no
                                               at enrollment in 1977.  Multiple logistic regression
                                               used to test for associations between new-onset
                                               asthma and long-term exposures to air pollution,
                                               controlling for age, education, pneumonia or
                                               bronchitis before age 16 years, and years worked
                                               with a smoker through 1987. All models stratified
                                               by gender.
                                  Mean FEV! declined significantly over the two measurement
                                  points for all subjects combined, which may reflect combined
                                  effects of O3 with exposures to dust, vehicle exhaust, and
                                  environmental tobacco smoke as reported by subjects from all four
                                  locations in the post-summer questionnaire. However, a larger
                                  mean decline was seen at the higher O3 site, Fort Dix, than at the
                                  remaining three sites, suggesting an influence of O3 exposures.

                                  A larger decline was observed in subjects with post-summer
                                  measurements in the 1 st two weeks after returning from training
                                  compared to those measured in the 3rd and 4th weeks, which is
                                  consistent with the lung function effects being somewhat transient.

                                  Change in lung function over the summer:
                                                                                                       All locations: -44 mL (SE 21), p = 0.035
                                                                                                       Fort Dix: -78 mL (SE 41), p = 0.07
                                                                                                       Forts Sill, Leonard Wood, and Benning combined:
                                                                                                       -31mL(SE24),p = 0.21
                                  There were 27 incident cases of asthma among 1,305 males and
                                  51 incident cases among 2,272 females.  In logistic regression
                                  analyses, long-term O3 exposures were associated with increased
                                  risk of incident asthma among males but not females.

                                  Relative risks for incident cases of asthma (per 10 ppb increase
                                  in annual mean O3):

                                  Males: 3.12 (1.61,  5.85)
                                  Females: 0.94 (0.65, 1.34)

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                          Mean O3 Levels
                                       Study Description
                                                                    Results and Comments
X
Oi
        United States (cont'd)

        McDonnell et al.
        (1999)
        California
        1973-1992
Peters etal. (1999a,b)
12 Southern
California
communities
1993-1994
                       8-h avg O3
                       (9 a.m.-5 p.m):
                       20-year mean:
                       46.5 ppb
                       SD15.3
l-hmaxO3:
Mean range:

1986-1990:
30.2 ppb (Santa Maria)
to 109.2 ppb
(San Dimas)

1994:
35.5 ppb (Santa Maria)
to 97.5 ppb
(Lake Gregory)
This study continued the work of Greer et al.
(1993).  3,091 nonsmoking adults completed
questionnaires at one additional time point, 1992.
Residential histories used to interpolate air pollution
levels to zip centroids over the period 1973-1992,
yielding annual mean exposure estimates for O3,
PM10, SO2, and NO2. New asthma cases defined as
answering yes to doctor diagnosed asthma at either
1987 or 1992. Multiple logistic regression used to
test for associations between new-onset asthma and
long-term exposures to air pollution, controlling for
age, education, pneumonia or bronchitis before age
16, and ever smoking. All models run separately
for males and females.

3,676 children aged 9-16 years enrolled into
the 1st cohort of the Children's Health Study in
1993. Subjects provided questionnaire data on
respiratory  disease histories and symptoms.
3,293 subjects also underwent pulmonary function
testing, of which 2,781 were used in air pollution
regressions. Air pollution data for O3, PM10, PM2 5,
NO2, and inorganic acid vapors analyzed from
1986-1990  and 1994. For cross-sectional analysis
of respiratory diseases, individual pollutants were
tested for associations with ever asthma, current
asthma, bronchitis, cough, and wheeze after
controlling for covariates.  For analysis of lung
function, individual pollutants and pairs of
pollutants were regressed with FVC, FEV1; FEF25.75,
and PEF, controlling for usual demographic and
anthropometric covariates.
                                                                          There were 32 incident cases of asthma among 972 males and
                                                                          79 incident cases among 1,786 females.  In logistic regression
                                                                          analyses, long-term O3 exposures were associated with increased
                                                                          risk of incident asthma among males but not females.  Other
                                                                          pollutants were neither associated with asthma incidence nor
                                                                          were  confounders of the O3 association in males.

                                                                          Relative risks for incident cases of asthma (per 27 ppb increase
                                                                          in annual mean 8-h avg O3):

                                                                          Males: 2.09 (1.03, 4.16)
                                                                          Females: 0.86 (0.58, 1.26)
Acids and NO2, but not O3, were associated significantly with
prevalence of wheeze.  No associations of O3 with any of the
respiratory diseases or symptoms.

Decreased lung function was associated with multiple pollutants
among females but not males. For O3 exposure in females, all four
lung function variables declined with increasing exposure.
Associations were stronger for current (1994) exposure compared
to previous (1986-1990) exposure. In males who spent more time
outdoors, FVC and FEV[ declined significantly with higher current
exposure to O3.

Change in lung function (per 40 ppb 1-h max O3 from 1986-1990):

Females:
PEF:  -187.2mL/s(SE50.1),p<0.005
FEF25.75:  -102.2 mL/s (SE 28.8), p<0.0\

Males:
PEF: 31.1 mL/s (SE 48.8), p> 0.05
FEF25.75: 11.7 mL/s (SE 26.7), p> 0.05

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
   Mean O3 Levels
               Study Description
                   Results and Comments
X
        United States (cont'd)

        Gauderman et al.
        (2000, 2004a,b)
        12 Southern
        California
        communities
        1993-2001
        Gauderman et al.
        (2002)
        12 Southern
        California
        communities
        1996-1999
        McConnell et al.
        (1999)
        12 Southern
        California
        communities
        1993
8-h avg O3
(10 a.m.-6 p.m.):
Mean range:
Approximately 28 ppb
(Long Beach) to 65 ppb
(Lake Arrowhead)
8-h avg O3
(10 a.m.-6 p.m.):
Mean range:
Approximately 27 ppb
(Long Beach) to 67 ppb
(Lake Gregory)
l-hmaxO3:
Estimated annual daily
mean:
65.5 ppb
Range 35.5-97.5
Analysis of longitudinal lung function change
in relation to long-term air pollution levels in
the Children's Health Study. Children from 4th
(n = 1,498), 7th (n = 802), and 10th (n = 735) grade
enrolled in 1993. Children enrolled in 7th and
10th grade were followed until 1997; 4th graders
were followed until 2001.  Baseline questionnaires
completed and annual pulmonary function tests
(FVC, FEVj, FEF25.75, FEF75) performed. Air
pollution monitoring stations established in the
12 study communities beginning in 1994 to measure
O3, NO2, PM10, PM2 5, and inorganic acid. Analysis
using  adjusted linear regression models.

Second cohort of the Children's Health Study.
2,081  4th graders (mean age 9.9 years) enrolled in
1996.  Baseline questionnaires were completed and
annual pulmonary function tests (FVC, FEV1; FEF25.
75, FEF25.75/FVC, PEF) were performed.  1,672
children had at least two pulmonary function test
data.  Air pollutants examined include O3, NO2,
PM10, PM25, inorganic acid, elemental carbon, and
organic carbon.  Adjusted linear regression model
was used.

First cohort of the Children's Health Study.
Association of O3 with prevalence of chronic lower
respiratory tract symptoms among children with a
history of asthma was examined in a cross-sectional
study  in 12 communities. Questionnaires were
completed by parents of 3,676 4th, 7th, and 10th
graders, of which 493 had asthma. Exposure data
(O3 NO2, PM10, PM2 5, and inorganic acid vapors)
collected in 1994 used to estimate exposure.
Analysis using logistic regression method.
In the 7th and 10th grade cohorts, difference in lung function
growth from the least to the most polluted community was not
associated with any of the air pollutants, including O3. Among the
4th graders, decreased lung growth was associated with exposures
to PM and NO2, but not with O3.
In this cohort, a significant association between O3 and PEF and
FVC was noted in children spending more time outdoors.

% difference in lung function growth from least to most polluted
community (per 36.6 ppb increase in annual mean 8-h avg O3):

PEF:
All children: -1.21% (-2.06, -0.36)
Children more outdoors: -1.62% (-2.93, -0.29)
Children less outdoors: -0.87% (-2.09, 0.37)

Children with asthma were much more likely to have bronchitis or
related symptoms than children without such history. Among the
asthmatic children, significant relationship were observed between
phlegm and all pollutants studied, with the exception of O3.

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
   Mean O3 Levels
               Study Description
                   Results and Comments
X
OO
        United States (cont'd)

        McConnell et al.
        (2002)
        12 Southern
        California
        communities
        1993-1998
        Rite et al. (2000)
        Southern California
        1989-1993
l-hmaxO3:
Four-year mean
(1994-1997):

Low pollution
communities (n = 6):
50.1 ppb
Range 37.7-67.9

High pollution
communities (n = 6):
75.4 ppb
Range 69.3-87.2
8-h avg O3
(9 a.m.-5 p.m.):
Six weeks before birth:
36.9 ppb
SD 19.4
Range 3.3-117 ppb
3,535 children (age 9-16 years) without a history of
asthma recruited in 1993 and 1996, and followed
with annual surveys through 1998 to determine
incidence of new onset asthma. Participation
in sports assessed at baseline. Copollutants included
PM10, PM2 5, NO2, and inorganic acid vapors.
Asthma incidence was examined as a function of
number of sports played in high and low pollution
communities, controlling for age, sex, and ethnic
origin.
Data on 97,158 singleton births over period
1989-1993 linked to air monitoring data during
different periods of pregnancy to determine
associations between pollution exposures and
preterm birth. Besides O3, pollutants of interest
included PM10, NO2, and CO. Multiple regression
analysis used, controlling for maternal age, race,
education, parity, and other factors.
Asthma incidence was not higher in the high pollution
communities as compared with the low pollution communities,
regardless of the pollutant used to define high/low. In fact,
the high O3 communities had generally lower asthma incidence.
However, in high O3 communities, there was an increased risk of
asthma in children playing three or more sports compared to those
playing no sports; no such increase was observed in the low O3
communities.  No other pollutant showed this association. These
results suggest that high levels of physical activity is associated
with risk of new asthma development for children living
in communities with high O3 levels.

Relative risk of developing asthma in children playing three or
more sports compared to those playing no sports:

Low pollution communities: 0.8 (0.4, 1.6)
High pollution communities: 3.3 (1.9, 5.9)

Both PM10 and CO during early or late pregnancy were  associated
with increased risk for preterm birth.  No associations observed
withO3.

-------
                                      Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
   Mean O3 Levels
Study Description
                                                                     Results and Comments
X
vo
        United States (cont'd)

        Kilnzlietal. (1997);
        Tageretal. (1998)
        Los Angeles and San
        Francisco, CA;
        Berkeley, CA
        1995
        Sherwin et al. (2000)
        Los Angeles, CA and
        Miami, FL
        1995-1997
8-h avg O3
(10 a.m.-6 p.m.):

Range of lifetime
mean:

Los Angeles:
25-74 ppb

San Francisco:
16-33 ppb
Levels not reported.
In a pilot study, 130 freshman students (age 17-21
years) at the University of California at Berkeley
measured for lung function and histories of
residential locations and indoor/outdoor activity
patterns and levels. By design, students had
previously resided in one of two metropolitan areas
that differed greatly in O3 concentrations,
San Francisco or Los Angeles. A key goal was to
test whether measures of small airways function
(e.g., nitrogen washout, FEF25.75, FEF75) were
sensitive measures of long-term O3 impacts.
Lifetime exposures to O3, PM10 and NO2 assigned
by interpolation to sequence of residence locations
from available monitoring stations. Multiple
exposure measures were derived with varying
degrees of incorporation of time-activity
information, going from ecological concentration
to individual time-activity weighted exposure.
Performed linear regression analysis of lung
function on O3 exposures, controlling for height,
ethnicity, gender, and region.

Lungs obtained from autopsies of young residents
(age 11-30 years) of Miami (n = 20) and Los
Angeles (n = 18) who died suddenly from homicide,
vehicular accident, or other violence.
Semiquantitative measurements of centriacinar
region alterations were compared between the
two cities.
                                   Decreased FEF25.75 and FEF75 were associated with long-term O3
                                   exposures. Results were similar whether O3 exposure was purely
                                   ecologic or incorporated time-activity information. FVC, FEV^
                                   and nitrogen washout were generally not associated with O3 levels.
                                   No evidence for PM10 or NO2 main effects or confounding of O3.
                                   Similar patterns results using O3 hours > 60 ppb as exposure metric
                                   instead of daily 8-h avg O3 (10 a.m.-6 p.m.).  Surprisingly, region
                                   of residence was a major negative confounder as lung function was
                                   lower on average among students from the low O3 city, San
                                   Francisco, than among those who had lived in Los Angeles.
                                   Ozone exposures were significant predictors only after controlling
                                   the regional effect.

                                   Change in lung function (per 20 ppb increase in lifetime mean 8-h
                                   avg O3):

                                   FEF25.75: -420 mL/s (-886, 46); 7.2% of population mean
                                   FEF75: -334 mL/s (-657, -11);  14% decline of population mean
                                   A greater extent (p < 0.02) and severity (p < 0.02) of centriacinar
                                   region alterations were observed in lungs of the Los Angeles
                                   residents than the Miami residents. These differences could not be
                                   attributed to smoking history. The higher O3 levels in Los Angeles
                                   might be responsible for the greater centriacinar region alterations,
                                   however correlations could not be performed due to the relatively
                                   small number of cases available.

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                          Mean O3 Levels
                                       Study Description
                                                                    Results and Comments
X
        United States (cont'd)

        Gongetal. (1998b)
        Glendora, CA
        1977-1987
Chen et al. (2002)
Washoe County, NV
1991-1999
        Kinneyetal. (1996)
        New York City
        1992-1993
l-hmaxO3:
Annual means range
(1983-1987):
109 ppb to 134 ppb
8-hmaxO3:
27.23 ppb
SD 10.62
Range 2.76-62.44

l-hmaxO3:

Summer (Jul-Sep
1992):
58 ppb
Maximum 100

Winter (Jan-Mar 1992):
32 ppb
Maximum 64

Summer (Jul-Sep
1993):
69 ppb
Maximum 142
164 adults (mean age 45 years; 34% males) from a
high O3 community underwent lung function testing
in 1986-1987 (T3).  Subjects were recruited from a
cohort of 208 nonsmoking adults who had been
tested on two previous occasions: 1977-1978 (Tl)
and 1982-1983 (T2). Analyzed changes in lung
function at three time points. Subjects were also
asked to undergo controlled exposures to 0.40 ppm
O3 over 2 hours with intermittent exercise.
45 subjects agreed to participate.  Investigators
hypothesized that acutely responsive subjects would
show more rapid declines in function over the study
period.

Birth weight for 36,305 single births analyzed
in relation to mean PM10, O3, and CO levels in
trimesters  1, 2, and 3.


19 healthy adult joggers (age 23-38 years; 18 males)
from the Governors Island U.S. Coast Guard facility
in New York harbor underwent a series of two
bronchoalveolar lavages, first in the summer of
1992 and then again in the winter of 1992. Because
the summer of 1992 had lower than average O3
levels, six  subjects underwent a third
bronchoalveolar lavage in the summer of 1993.
Study tested whether inflammatory markers in
bronchoalveolar lavage fluid were elevated during
the summer O3 season among adults who regularly
exercised outdoors.  Outcomes included cell
differentials, release of interleukin-8 (IL-8) and
tumor necrosis factor-alpha (TNF-a) in
bronchoalveolar lavage cells supernatants, release of
reactive oxygen species by macrophages, and
concentrations of protein, lactate dehydrogenase,
IL-8, fibronectin, al-antitrypsin (al-AT),
complement fragments (C3a), and prostaglandin E2
(PGE2) in  bronchoalveolar lavage fluids.
                                                                                                Mean FVC and FEV! showed nonsignificant increase from T2 to
                                                                                                T3, whereas an earlier analysis of the Tl to T2 change had found a
                                                                                                significant decline in function (Betels et al., 1987). There was
                                                                                                evidence for 'regression to the mean,' in that subject with larger
                                                                                                declines in function from Tl to T2 tended to have larger increases
                                                                                                in function from T2 to T3.  A consistent decline in FEVj/FVC ratio
                                                                                                was observed at all three time points (p < 0.0001 by ANOVA).

                                                                                                Acute changes in lung function, determined using controlled O3
                                                                                                exposures, were not associated with chronic lung function changes.
PM10 was the only air pollutant associated with decreased birth
weights. Ozone levels quite low throughout study.
                                                                                                There was no evidence of acute inflammation in the summer of
                                                                                                1992 compared to the winter; i.e., neutrophil differentials, IL-8 and
                                                                                                TNF-a showed no significant differences. However, a measure of
                                                                                                cell damage, lactate dehydrogenase, was elevated in the summer,
                                                                                                suggesting possible O3-mediated damage to the lung epithelium
                                                                                                with repeated exposures to O3 while exercising. O3 levels during
                                                                                                the summer of 1992 were atypically low for New York City.
                                                                                                Among six subjects who agreed to undergo a third bronchoalveolar
                                                                                                lavage test in the summer of 1993, lactate dehydrogenase was
                                                                                                again elevated compared to winter. In addition, IL-8 was elevated
                                                                                                in the summer of 1993, suggesting acute inflammation.

-------
                                      Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                                       Study Description
                                                                                                                            Results and Comments
X
        Europe

        Charpinetal. (1999)
        Seven towns in
        SE France
        Jan-Feb 1993
Ramadour et al.
(2000)
Seven towns in
SE France
Jan-Feb 1993
8-h max O3:              2,073 children (age 10-11 years) from 7 towns
Range of means:         tested for atopy based on skin prick testing
30.2-52.1 ug/m3          (house dust mite, cat dander, grass pollen,
                        cypress pollen, and Alternaria).  Towns represented
24-h avg O3:             a range of ambient O3 and other pollutant (NO2 and
Range of means:         SO2) levels.  Tested hypothesis that atopy is greater
20.1-42.1 ug/m3          in towns  with higher photochemical pollution levels.
                        To be eligible, subjects must have resided in current
                        town for  at least 3 years. Authors stated that Jan to
                        Feb pollution levels correlated with levels observed
                        throughout the year, though no data was given to
                        support this.

8-h max O3:              2,445 children (age 13-14 years) who had lived at
Range of means:         their current residence for at least three years were
30.2-52.1 ug/m3          recruited from schools in seven towns in SE France.
                        This region has highest O3 levels in France.
                        Subjects  completed ISAAC  survey of asthma and
                        respiratory symptoms. In addition to O3 also
                        collected data on SO2 and NO2. Analyzed
                        relationships between asthma and other respiratory
                        conditions with mean air pollution levels across the
                        seven towns using logistic regression, controlling
                        for family history of asthma, personal history of
                        early-life respiratory diseases, and SES.  Also did
                        simple univariate linear regressions.
                                                                                                         In this cross-sectional analysis, no differences in atopy levels were
                                                                                                         seen across the seven towns. Authors concluded that long-term
                                                                                                         exposures to oxidant pollution do not favor increased allergy to
                                                                                                         common allergens. The very low winter O3 levels monitored and
                                                                                                         lack of long-term exposure data make it impossible to reach this
                                                                                                         conclusion in a definitive manner.
                                                                                                         In logistic regressions, no significant associations seen between O3
                                                                                                         and 12-month history of wheezing, history of asthma attack,
                                                                                                         exercise induced asthma and/or dry cough in last 12 months.

                                                                                                         In simple bivariate scatterplots of respiratory outcomes versus
                                                                                                         mean O3 levels in the seven towns, there appeared to be strong
                                                                                                         positive relationships (r = 0.71 for wheezing in last 12 months
                                                                                                         and r = 0.96 for asthma attacks). No data on slope estimates given.
                                                                                                         Concerns about potential confounding across towns
                                                                                                         limits the interpretation of this study.

-------
                                    Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                                      Study Description
                                                                  Results and Comments
X
to
        Europe (cont'd)

        Ihorst et al. (2004)
        Nine communities in
        Lower Austria
        Apr 1994-Oct 1997
        Six communities in
        Germany
        Febl996-0ctl999
        Kopp et al. (2000)
        Ten communities in
        Austria and SW
        Germany
        Mar 1994-Nov 1995
!/2-h avg O3:
Quartile ranges:

Summer:
1st quartile: 22-30 ppb
2nd quartile: 30-38 ppb
3rd quartile: 38-46 ppb
4th quartile: 46-54 ppb

Winter:
1st quartile: 4-12 ppb
2nd quartile: 12-20 ppb
3rd quartile: 20-28 ppb
4th quartile: 28-36 ppb
!/2-h avg O3:
Stratified by low,
medium, high
exposure:

Low: 24-33 ppb
Medium: 35-38 ppb
High: 44-52 ppb
2,153 children (median age 7.6 years) were studied
for the effects of semi-annual and 3'/2-year mean O3
concentrations on FVC and FEVj. As a measure of
lung growth, the difference between two
consecutive values for each child was divided by the
number of days between tests. The effect of O3
exposure on lung growth was analyzed by linear
regression models, after adjusting for sex, age,
height at start of the time period, and passive
smoking exposure.
797 children with a mean age of 8.2 years.  Four
pulmonary function tests (FVC, FEV[) performed
on each child over two summers. Examined
association between average daily lung function
growth and exposure to O3, PM10, NO2, and SO2.
Analysis using linear regression models.
Higher semi-annual mean O3 levels were associated with
diminished lung function growth during the summer, but
increased lung function growth in the winter.

Change in lung function (4th quartile compared to 1 st quartile
semi-annual O3 mean):

Summer:
FEVj (mL/100 days): -18.5 (-27.1,-9.8)
FVC (mL/100 days):  -19.2 (-27.8, -10.6)

Winter:
FEV; (mL/100 days): 10.9 (2.1, 19.7)
FVC (mL/100 days):  16.4 (8.3, 24.6)

No associations between longer term O3 exposure (mean summer
O3 over a 3!/2-year period) and lung function growth was found.

Lower FVC and FEV[ increases were observed in children exposed
to high ambient O3 levels compared to those exposed to lower O3
levels during the summer.  During the winter, children in higher O3
areas showed a slightly  greater increase in FVC and FEV; than
those in the low O3 areas, which might reflect that children catch
up in lung function deficits during the winter season.

Change in lung function for high versus low O3 exposure groups
(per ppb O3):

FEVI:
Summer of 1994:  -0.303 mL/day, p = 0.007
Winter of 1994/1995: 0.158 mL/day, p = 0.006
Summer of 1995:  -0.322 mL/day, p = 0.001

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                                       Study Description
                                                                    Results and Comments
X
        Europe (cont'd)

        Frischeretal. (1999)
        Nine communities in
        Austria
        1994-1996
        Frischeretal. (2001)
        Nine communities in
        Austria
        Sep-Octl997
24-h avg O3:

Summer:
34.8ppb
SD8.7

Winter:
23.1ppb
SD7.7
!/2-h avg O3:
30-day mean:
31.57ppb
IQR 20.61
Communities from two counties chosen to represent
a broad range of O3 concentrations; a two-fold range
in mean levels was observed.  1,150 children (mean
age 7.8 years; 52% males) from grades 1 and 2
performed spirometry in spring and fall over three
years (total of six measurements per child) to
determine if seasonal exposure to O3 would be
associated with diminished lung function growth,
especially over the summer seasons. Ozone levels
were low during lung function testing periods.
Participation rates were high.  At baseline,
respiratory histories were collected and subjects
were tested for allergy by skin prick. Examined
association between O3 levels and change in lung
function (FVC, FEVj, and MEF50 [maximal
expiratory flow at 50% of vital capacity]) over each
season, controlling for baseline function, atopy,
gender, site, environmental tobacco
smoke exposure, season, and change in height.
Other pollutants studied included PM10, SO2, and
NO,.
A cross-sectional study of 877 school children
(mean age 11.2 years). Analyzed for urinary
eosinophil protein as a marker of eosinophil
activation determined from a single spot urine
sample using linear regression models.
Seasonal mean O3 exposures were associated with reductions in
growth in all three lung function measures.  Inconsistent results
seen for other pollutants.  Summer season lung function growth
decrements per unit O3 were larger when data restricted to children
who spent whole summer in their community. No evidence for
nonlinear O3 effect. No confounding of O3 effect by temperature,
ETS, or acute respiratory illnesses.

Change in lung function (per ppb O3):

FEV; (mL/day):
All subjects:
Summer: -0.029 (SE 0.005), p< 0.001
Winter:  -0.024 (SE 0.006), p< 0.001
Restricted to subjects who stayed in community:
Summer: -0.034 (SE 0.009), p < 0.001

FVC (mL/day):
All subjects:
Summer: -0.018 (SE 0.005), p< 0.001
Winter:  -0.010 (SE 0.006), p = 0.08
Restricted to subjects who stayed in community:
Summer: -0.033 (SE 0.007), p < 0.001

Log-transformed urinary eosinophil protein-X concentrations were
found to be significantly associated with O3 levels, after adjusting
for gender,  site, and atopy.

Change in log urinary eosinophil protein-X (per ppb O3):
0.007 ug/mmol creatinine (SE 0.02), p < 0.001

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                                      Study Description
                                                                   Results and Comments
X
        Europe (cont'd)

        Horak et al. (2002a,b)
        Eight communities in
        Austria
        1994-1997
        Palli et al. (2004)
        Florence, Italy
        1993-1998
Seasonal mean O3:

Summer:
31.8ppb
Range 18.7-49.3

Winter:
19.8ppb
Range 12.7-35.9
24-h avg O3:
Range of monthly
means from 1993-1998:
Approximately
25-125 ppb
This study continued the work of Frischer et al.,
1999 by including one additional year of data, 1997.
The major hypothesis considered PM10. For this
study, 80.6% of the 975 children (mean age 8.11
years) performed all six lung function tests. A total
of 860 children were included in the GEE analysis.
Multipollutant analysis for PM10, SO2, and NO2.
320 residents (age 35-64 years) in the metropolitan
area of Florence enrolled in a study investigating the
correlation between levels of DNA bulky adducts
and cumulative O3 exposure. One blood sample was
collected for each subject.  Various time windows
of exposure were examined, ranging from 0-15 days
to 0-90 days prior to the blood draw.  Simple
Spearman correlations between DNA adduct levels
and different O3 exposure time windows were
calculated after stratifying by smoking history,
area of residence, and population type (random
sample or exposed workers).
Seasonal mean O3 showed a negative effect on lung function
growth, confirming the previous shorter study.  Ozone effects were
robust to inclusion of PM10 into the model. However, for FEV; in
winter, the O3 effect slightly diminished after including PM10.
Taking into account only children who stayed at home the whole
summer period did not affect the results.

Change in lung function (per ppb O3):

FEV; (mL/day):
O3 only models:
Summer: -0.021, p<0.00\
Winter: -0.020, p<0.00\
O3 with PM10 models:
Summer: -0.020,p<0.00\
Winter: -0.012, p = 0.04

Consistent relationships between O3 exposure and DNA adduct
levels were  observed only among never smokers.  Correlations
were highest among never smokers who resided in the urban area
and were not occupationally exposed to vehicle traffic pollution.
Associations were significant up to a time window of 0-60 days
prior to the blood draw in the subgroup of never smokers, with
strongest relationships observed between 30-45 days prior.

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                                      Study Description
                                                                   Results and Comments
X
        Latin America

        Calderon-Garciduenas
        etal. (1995)
        SW Mexico City
        Nov 1993
        Manzanillo, Mexico
        Jan 1994
        Calderon-Garciduenas
        etal. (1997)
        SW Mexico City
        Sep-Nov 1995
        Manzanillo, Mexico
        Jan 1995
SW Mexico City
(urban):
l-havgO3>120ppb:
4.4 hours/day
Maximum 307 ppb

Manzanillo, Pacific
port (control):
No detectable air
pollutants.
SW Mexico City
(urban):
l-havgO3> 120 ppb:
82 hours/month
Maximum 286 ppb

Manzanillo, Pacific
port (control):
No detectable air
pollutants.
Nasal lavage samples collected from 38 urban
(mean age 12.2 years) and 28 control (mean age
11.7 years) children. Samples analyzed for
polymorphonuclear leukocyte counts, expression
of human complement receptor type 3 (GDI Ib) on
nasal polymorphonuclear leukocytes, and nasal
cytologies.
129 urban and 19 control children aged 6-12 years
old with no history of smoking or environmental
tobacco smoke exposure and no current medication
use for atopy or asthma. Three nasal biopsies
obtained at 4-week intervals and analyzed for DNA
damage based on the presence of DNA fragments.
Nasal cytologies revealed that children from Mexico City had
abnormal nasal mucosae, including mucosal atrophy, marked
decreases in the numbers of ciliated-type cells and goblet cells,
and squamous metaplasia.

Exposed children had significantly higher nasal
polymorphonuclear leukocyte counts (p < 0.001) and nasal GDI Ib
expression (p < 0.001) compared to controls. However, the
inflammatory response did not seem to correlate with the previous
day's O3 exposure in a dose-dependent manner, suggesting that
there might be a competing inflammatory mechanism at the
brochioalveolar level. Overall, these results suggest that ambient
O3 produces an inflammatory response in chronically exposed
children.

Urban children had significantly more DNA fragments than did
control children (p < 0.0001).  Percentage of damaged cells was
82.2% (SE 6.4) in urban children and 17.0% (SE 6.1) in control
children. Among urban children, more upper respiratory
symptoms and DNA damage was seen with increasing age.
Older children spent more time outdoors and engaged in physical
activities (p<0.001).

Urban children were exposed to a complex pollution mix, making
it difficult to attribute effects to O3 specifically. However, authors
noted that O3 was the pollutant with most exceedences of air
quality standard.

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period
                                       Study Description
                                                                    Results and Comments
X
        Latin America (cont'd)

        Calderon-Garciduenas   SW Mexico City
        etal. (1999)
        SW Mexico City
        May-June 1996
        Manzanillo, Mexico
        May 1996
        Fortoul et al. (2003)
        Mexico City
        May 1997
(urban):
l-havgO3>80ppb:
May:
161 hours/month
Maximum 232 ppb
June:
98 hours/month
Maximum 261 ppb

Manzanillo, Pacific
port (control):
Mean< 10 ppb
9-h avg O3
(9 a.m.-6 p.m.):

South:
121 ppb
North:
89 ppb
86 urban and 12 control children aged 6-13 years
old with no history of smoking or environmental
tobacco smoke exposure and no use of medication
for atopy or asthma.  Urban children stratified into
five groups by school grade level (1st through 5th).
Nasal epithelial biopsies obtained from inferior
nasal turbinates, and analyzed for single strand
DNA breaks and for 8-OHdG (8-hydroxy-2'-
deoxyguanosine), a mutagenic lesion produced by
G^T mutations. These outcomes relate to possible
carcinogenic effects of air pollution exposures.
Multiple air pollutants monitored in SW Mexico
City within 3 miles of urban subject residences.
Estimated DNA strand breaks on nasal epithelial
cells and leucocytes sampled from asthmatic
(n = 15) and nonasthmatic (n = 224) medical
students aged 18-28 years using a single-cell
gel electrophoresis assay.
No respiratory symptoms reported by control children; urban
children reported multiple nasal and lung symptoms, including
cough and chest discomfort among 46% of urban children, with
higher rates for 5th versus 1 st graders. 8-OHdG was
approximately 3-fold higher in biopsies from urban children
(p < 0.05), however, no differences by school grade. Single strand
DNA breaks were more common in urban versus control children,
with an age-dependent increase in the urban children (p < 0.05).
These results suggest that DNA damage is present in the nasal
epithelial cells of children living in highly polluted SW Mexico
City  and may reflect enhanced risk of cancer later in life.

Though O3 represents an important component of the pollution
mix, it is not possible to attribute effects solely to O3.

Greater genotoxic damage in asthmatics' nasal epithelial cells
(p < 0.05) may reflect their higher vulnerability for DNA damage,
or a decreased ability to repair it, compared with nonasthmatic
subjects.
        Gouveia et al. (2004)
        Sao Paulo, Brazil
        1997
l-hmaxO3:
63.0 ppb
SD33.5
Birth weight for 179,460 single births analyzed
in relation to PM10, SO2, CO, NO2, and O3 levels in
trimester 1, 2, and 3. GAM and logistic regression
models used for analysis.
Exposures to PM10 and CO during 1 st trimester were found to
have significant negative associations with birth weight.
No associations observed for the other air pollutants, including O3.

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
to
o
o
          Reference, Study
        Location, and Period      Mean O3 Levels
                                       Study Description
                                                                    Results and Comments
X
        Asia

        Kuo et al. (2002)
        Central Taiwan
        1996
l-hmaxO3:
Annual mean range
across 7 of 8 schools:
18.6-27.3 ppb
Respiratory questionaire administered to
12,926 children aged 13-16 years at eight junior
high schools in central Taiwan, to determine asthma
prevalence. The association between asthma
prevalence and air pollution exposure analyzed by
simple Pearson correlations of prevalence with
annual mean air pollution levels (O3, SO2, PM10,
and NO2), and by multiple logistic regression.
The 775 asthmatics who were identified then
provided follow-up data on symptoms and
exacerbations over a one-year period. Simple
Pearson correlations were computed between
monthly hospital admissions and air pollution
levels, not controlling for covariates such as season
or weather.
Asthma prevalence ranged from 5.5% to 14.5% across the 8
schools. Based on simple Pearson's correlations, mean O3
(r = 0.51) and NO2 (r = 0.63) levels were correlated with variations
in asthma prevalence. However, only NO2 remained significant
in multiple logistic regression analyses after adjusting for various
potential confounding factors.

Longitudinal hospital admissions results are inconclusive due
to analytical limitations.  Monthly  correlations of hospital
admissions for asthmatics yielded variable results, all of which
would be confounded by temporal factors.

-------
                                                 Table AX7-7.  Effects of Chronic O3 Exposure on Mortality
to
o
o
Reference, Location,
    Study Period
Mean O3 Levels
                                                                              Study Description
Results and Comments
        United States

        Pope et al. (2002)
        U.S. nationwide
        1982-1998
                         l-hmaxO3:
                         59.7 ppb
                         SD 12.8

                         24-h avg O3:
                         45.5 ppb
                         SD7.3
                      Approximately 500,000 members of American Cancer
                      Society cohort enrolled in 1982 and followed through
                      1998 for all cause, cardiopulmonary, lung cancer, and
                      all other cause mortality. Age at enrollment was
                      30+ years. Air pollution concentrations in urban area of
                      residence at time of enrollment assessed from 1982
                      through 1998. Other pollutants considered include TSP,
                                                   PM15 , PM10, PM2
                                                                              , PM15.2 5, S042-, S02, N02, and CO.
                                                                                                                  No significant effect of O3 on mortality risk, though the
                                                                                                                  association of Jul-Sep O3 concentrations with all cause
                                                                                                                  and cardiopulmonary mortality were positive and nearly
                                                                                                                  significant.

                                                                                                                  Residential location was known only at enrollment to
                                                                                                                  study in 1982. Thus, exposure misclassification is
                                                                                                                  likely to be high.
 0.05

                                                                           Analyses were robust to the deletion of diastolic blood
                                                                           pressure in the models, indicating that the association
                                                                           between mortality  and O3 was not mediated through
                                                                           blood pressure.

-------
                                           Table AX7-7 (cont'd).  Effects of Chronic O3 Exposure on Mortality
to
o
o
Reference, Location,
    Study Period
Mean O3 Levels
Study Description
Results and Comments
 lOOppb:
                        330 h/year
                        SD295
                        IQR 551
                      Prospective cohort study of 6,338 nonsmoking non-
                      Hispanic white adult members of the Adventist Health
                      Study followed for all cause, cardiopulmonary,
                      nonmalignant respiratory, and lung cancer mortality.
                      Participants were aged 27-95 years at enrollment in
                      1977. 1,628 (989 females, 639 males) mortality events
                      followed through 1992. All results were stratified by
                      gender.  Used Cox proportional hazards analysis,
                      adjusting for age at enrollment, past smoking,
                      environmental tobacco smoke exposure, alcohol use,
                      education, occupation, and body mass index. Analyzed
                      mortality from all natural causes, cardiopulmonary,
                      nonmalignant respiratory, and lung cancer. Ozone
                      results were presented for both metrics.
                                    Of 16 regressions involving O3 exposures (two genders;
                                    four mortality causes; two O3 metrics), 11 were positive
                                    and one was statistically significant, for lung cancer in
                                    males for O3 h/year > 100 ppb.

                                    Relative risks for lung cancer mortality in males:

                                    24-h avg O3 (per 12.0 ppb):
                                    2.10(0.99,4.44)

                                    O3 h/year > 100 ppb (per 551 hours/year):
                                    4.19(1.81,9.69)

                                    Inconsistency across outcomes and genders raises
                                    possibility of spurious finding. The lack of
                                    cardiopulmonary effects raises plausibility concerns.
        Beesonetal. (1998)
        Three California air
        basins: San Francisco,
        South Coast (Los Angeles
        and eastward), San Diego
        1977-1992
                        Annual mean O3
                        26.2 ppb
                        SD7.7
                        O3h/year> 100 ppb:
                        333 h/year
                        SD297
                      6,338 nonsmoking non-Hispanic white adult members
                      of the Adventist Health Study aged 27-95 years at time
                      of enrollment.  36 (20 females, 16 males) histologically
                      confirmed lung cancers were diagnosed through 1992.
                      Extensive exposure assessment, with assignment of
                      individual long-term exposures to O3, PM10, SO42", and
                      SO2, was a unique strength of this study.  All results
                      were stratified by gender.  Used Cox proportional
                      hazards analysis, adjusting for age at enrollment, past
                      smoking, education, and alcohol use.
                                    Males, but not females, showed moderate association
                                    for O3 and incident lung cancer risk.

                                    Relative risks for lung cancer incident in males:

                                    O3 h/year > 100 ppb (per 556 hours/year):
                                    All males: 3.56(1.35,9.42)
                                    Never smokers: 4.48 (1.25, 16.04)
                                    Past smokers: 2.15 (0.42, 10.89)

-------
 i                         8.  INTEGRATIVE SYNTHESIS
 2
 3
 4      8.1  INTRODUCTION
 5           This integrative synthesis (Chapter 8) aims to provide a coherent framework for the
 6      assessment of health risks associated with human exposures to ambient ozone (O3) in the United
 7      States.  The main goal of this chapter is to integrate newly available scientific information with
 8      that discussed in the 1996 O3 AQCD, to address issues central to the EPA's assessment of
 9      scientific information needed to support the current review of the primary O3 NAAQS.  Other
10      scientific information concerning ambient O3 welfare effects (i.e., effects on vegetation
11      /ecosystems, surface-level solar UV flux/climate changes, and man-made materials) and
12      pertinent to review of secondary O3 standards is assessed in ensuing Chapters 9, 10, and 11. The
13      integrated assessment of scientific findings provided here and elsewhere in this document will be
14      used and their policy implications considered in an Ozone Staff Paper to be prepared by EPA's
15      Office of Air Quality Planning and Standards (OAQPS). The scientific and technical
16      assessments provided in that Staff Paper will "bridge the gap" between scientific assessments in
17      this criteria document and judgments required of the EPA administrator in evaluating whether to
18      retain or, possibly, to revise the  current O3 NAAQS.
19           Ozone found in the earth's atmosphere generally originates from photochemical reactions
20      that are predominantly catalyzed by the interaction of sunlight with other pollutants, especially
21      nitrogen oxides (NOX) and hydrocarbons such as volatile organic compounds (VOCs). Other
22      photochemical oxidants, such as peroxyacetyl nitrate (PAN) and hydrogen peroxide (H2O2), are
23      also generated along with O3 by such atmospheric interactions. In addition to the tropospheric
24      O3 generated by these interactions, some O3 is found near the earth's surface as the result of its
25      downward transport from the stratosphere, even in the absence of photochemical reactions in the
26      troposphere.  However, in contrast to stratospheric O3, which plays an important role in
27      maintaining the habitability of the planet by shielding the surface from harmful solar ultraviolet
28      (UV) radiation, tropospheric O3 at the surface can exert adverse effects on humans, animals, and
29      vegetation.  This criteria document is mainly focused on assessment of health and welfare effects
30      resulting from exposures to surface level concentrations of tropospheric O3, with only relatively
31      limited attention begin accorded to other photochemical oxidants such as PAN or H2O2.

        January 2005                              8-1        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.1.1  Chapter Organization
 2           This integrative synthesis chapter is divided into several major sections.  This first section
 3      (Introduction) not only aims to orient the reader to the organization and content of the chapter,
 4      but also provides background information on the current O3 NAAQS and important types of
 5      human responses to O3 exposure that were considered as key bases for the 1997 EPA revision of
 6      the O3 NAAQS.  The next section (Section 8.2) focuses on air quality trends and current ambient
 7      O3 concentrations to provide context for ensuing discussions of ambient O3 exposures and its
 8      effects on human health and welfare.
 9           The subsequent sections (8.3, 8.4, and 8.5) then build upon the integrative synthesis
10      presented in Chapter 9 of the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996) to
11      integrate newly available key scientific information assessed in Chapters 4 through 7 of this
12      document. This includes integration of information on dosimetry, as well as lexicological,
13      human clinical, and epidemiological studies.
14           These sections collectively address the following key issues: (1) ambient exposures,
15      personal exposures, and dosimetric considerations; (2) experimental studies on toxicological
16      responses to acute O3 exposures in humans (clinical studies) and both acute and chronic effects
17      in animals; (3) assessment of epidemiological evidence for associations between O3 exposure in
18      human populations and health effects and the robustness of these associations; (4) integration of
19      the experimental data with epidemiological assessments; (5) biological mechanisms and other
20      evidence useful in judging the plausibility of adverse health effects being associated with human
21      exposures to ambient O3 levels encountered in the  United States; and (6) identification of
22      susceptible and vulnerable populations potentially  at increased risk for O3-related health effects
23      and potential public health impacts of human exposure to ambient O3 in the United States.
24           The present chapter mainly focuses on discussion of new scientific information that has
25      become available since the 1996 O3 criteria review that supported EPA's revision of the O3
26      NAAQS in 1987. However, it also highlights important data gaps and uncertainties that still
27      exist with regard to various key issues and notes important research needs in a number of key
28      areas. Detailed evaluation of such research needs is beyond the scope  of this document, but will
29      be undertaken as part of later EPA efforts focused  on identification of O3 research needs and
30      development of research planning documents.
31

        January 2005                               8-2        DRAFT-DO NOT QUOTE OR CITE

-------
 1     8.1.2   Current Standards
 2          The NAAQS for ambient O3 were revised in 1997 by adding an 8-h standard (Table 8-1) in
 3     addition to the 1979 1-h standard, which is met if the fourth highest daily maximum 1-h O3 over
 4     a 3-year period is < 0.12 ppm.  The 8-h standard is met when the 3-year average of the annual
 5     fourth highest daily maximum 8-h average concentration is < 0.08 ppm. The 1997  standards
 6     were based on various scientific supportive data from human exposure and epidemiological
 7     studies as assessed in the 1996 O3 AQCD. The gradations of individual responses observed with
 8     short-term  exposure to O3 in healthy persons (Table 8-2) and in persons with impaired
 9     respiratory systems (Table 8-3) are representative of the critical information used in these
10     evaluations, as summarized in Tables 9-1 and 9-2 of the 1996 O3 AQCD and reproduced herein
11     Tables 8-2  and 8-3, respectively. Detailed assessments of the scientific information and
12     supportive  data used in generating these tables can be found in the 1996 O3 AQCD  (U.S.
13     Environmental Protection Agency,  1996). Key findings from health studies that have become
14     newly available since the 1996 criteria review are discussed below in later sections of this
15     chapter and any important consequent reaffirmations or modifications of findings of the types
16     summarized in Tables 8-2 and 8-3 are highlighted.
17
18
              Table 8-1. Current National Ambient Air Quality Standards (NAAQS) in the
                                             United States
Pollutant
Ozone

Date of
Promulgation
7/18/97
(62FR38856)
3/9/94
(58FR52852)
Primary NAAQS
0.08 ppm
(157 Mg/m3)
0.12 ppm
(235 Mg/m3)
Averaging Time
8-ha
l-hb
Secondary
NAAQS
Same as primary
Same as primary
         "Based on the 3-year average of the annual fourth-highest daily maximum 8-h average O3 concentration
         measured at each monitor within an area.
         b The standard is attained when the expected number of days per calendar year with maximum hourly average
         concentrations above 0.12 ppm is < 1.
       January 2005                               8-3         DRAFT-DO NOT QUOTE OR CITE

-------
            Table 8-2. Gradation of Individual Responses to Short-Term Ozone Exposure in
                                         Healthy Personsa*
Functional Response
FEVj
Nonspecific
bronchial responsiveness'5
Duration of
response
Symptomatic Response
Cough
Chest pain
Duration of response
Impact of Responses
Interference with normal
activity
None
Within normal
range (±3%)
Within normal
range
None
Normal
Infrequent
cough
None
None
Normal
None
Small
Decrements of
3 to < 10%
Increases of
< 100%
< 4 hours
Mild
Cough with deep
breath
Discomfort just
noticeable on
exercise or deep
breath
< 4 hours
Normal
None
Moderate
Decrements of
> 10 but < 20%
Increases of
< 300%
> 4 hours but
< 24 hours
Moderate
Frequent
spontaneous cough
Marked discomfort
on exercise or deep
breath
> 4 hours but
< 24 hours
Mild
A few sensitive
individuals choose
to limit activity
Large
Decrements of
> 20%
Increases of
> 300%
> 24 hours
Severe
Persistent
uncontrollable
cough
Severe discomfort
on exercise or
deep breath
> 24 hours
Moderate
Many sensitive
individuals choose
to limit activity
       a See text for discussion; see Appendix A for abbreviations and acronyms.
       b An increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD20 or PD100
       (see Chapter 7, Section 7.2.3).
       This table is reproduced from the 1996 O3 AQCD (Table 9-1, page 9-24) (U.S. Environmental Protection
       Agency, 1996).
1     8.2  TRENDS IN UNITED STATES OZONE AIR QUALITY
2     8.2.1  Ozone Concentrations, Patterns
3           Ozone is monitored in the United States during "O3 seasons," which vary in length from
4     geographic region to region.  The O3 season extends all year in the Southwest, but in most other
5     areas of the country, O3 is typically monitored from April to October. However,  O3 is present
6     year-round, not only in polluted areas, but in clean areas as well. The median O3 concentration
7     in the United States from 1996 to 2000, averaged over the appropriate O3 season,  was 33 ppb for
8     "urban" monitors located in Metropolitan Statistical Areas (MSAs); and it was 37 ppb for
      January 2005
8-4
DRAFT-DO NOT QUOTE OR CITE

-------
      Table 8-3. Gradation of Individual Responses to Short-Term Ozone Exposure in
                       Persons with Impaired Respiratory Systemsa*
Functional Response
FEVj change
Nonspecific
bronchial responsiveness'5
Airway resistance
(SRaw)
Duration of response
Symptomatic Response
Wheeze

Cough
Chest pain

Duration of response
Impact of Responses
Interference with normal
activity

None
Decrements of
<3%
Within normal
range
Within normal
range (±20%)
None
Normal
None

Infrequent
cough
None

None
Normal
None

Small
Decrements of
3 to < 10%
Increases of < 100%
SRaw increased
< 100%
< 4 hours
Mild
With otherwise
normal breathing
Cough with deep
breath
Discomfort just
noticeable on
exercise or deep
breath
< 4 hours
Mild
Few individuals
choose to limit
activity
Moderate
Decrements of
> 10 but < 20%
Increases of
< 300%
SRaw increased up
to 200% or up to
15cmH2O/s
> 4 hours but
< 24 hours
Moderate
With shortness of
breath
Frequent
spontaneous
cough
Marked
discomfort on
exercise or deep
breath
> 4 hours, but
< 24 hours
Moderate
Many individuals
choose to limit
activity
Large
Decrements of
> 20%
Increases of
> 300%
SRaw increased
> 200% or more
thanl5cmH2O/s
> 24 hours
Severe
Persistent with
shortness of
breath
Persistent
uncontrollable
cough
Severe discomfort
on exercise or
deep breath
> 24 hours
Severe
Most individuals
choose to limit
activity
 Medical treatment
No change
Normal medication
as needed
Increased         Physician or
frequency of       emergency room
medication use or   visit
additional
medication
 a See text for discussion; see Appendix A for abbreviations and acronyms.
 b An increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD20 or PD100
 (see Chapter 7, Section 7.2.3).
 This table is reproduced from the 1996 O3 AQCD (Table 9-2, page 9-25) (U.S. Environmental Protection Agency,
 1996).
January 2005
                  $-5
               DRAFT-DO NOT QUOTE OR CITE

-------
 1      monitors located outside MSAs. Median daily maximum 8-h concentrations between 10:00 a.m.
 2      and 6:00 p.m. were 46 and 47 ppb for monitors located in and outside MSAs, respectively.
 3      Median daily maximum 1-h concentrations were 56 ppb for monitors located in MSAs and 55
 4      ppb for monitors located outside of them. The daily maximum 1-h concentrations tended to be
 5      much higher in some large urban areas or in areas downwind of them, e.g., they were 202 ppb in
 6      Houston, TX in 1999 and 161 ppb in 2000. Daily 1-h maximum O3 concentrations were lower
 7      in nonurban areas of the country but still above 120 ppb in many locations.  Eight-hour daily
 8      maximum concentrations were not as high as 1-h daily maxima, but they also tended to be highly
 9      correlated with the 1-h maxima.
10          Within individual MSAs,  O3 concentrations tend to be well correlated across monitoring
11      sites, although variations in concentrations can be substantial. In many city centers, O3
12      concentrations tend to be lower than in either upwind  or downwind areas, largely due to NO
13      emitted by motor vehicles. Thus, although emissions  of nitrogen oxides and VOCs from motor
14      vehicles contribute to O3 formation, the relationship to O3 concentrations is not straightforward
15      in terms of proximity to mobile sources. In urban areas with high traffic density or near
16      highways, emissions of NO from traffic react with ozone, thereby reducing its concentration.
17      For example, much lower ozone concentrations overall are found in downtown Los Angeles
18      (e.g., in Lynwood) than at sites located further downwind (e.g., in San Bernadino).  The much
19      higher levels are formed from photochemical reactions involving the urban emissions, including
20      products produced as the result of reactions titrating ozone in the urban core. Thus, ozone
21      concentrations tend to be higher downwind of urban centers, and they decrease again in going to
22      areas that are remote from precursor sources.
23
24      8.2.2   Seasonal Variations
25          Ozone concentrations tend to peak in early to mid-afternoon in areas where there is strong
26      photochemical activity and to peak later in the afternoon or early evening in areas where
27      transport is more important in determining the O3 abundance. Summertime maxima in O3
28      concentrations occur in U.S. areas where substantial photochemical activity acts on O3
29      precursors emitted as the result of human activities. Monthly maxima can occur anytime from
30      June through August. However, springtime maxima are observed in National Parks, mainly in
31      the western United States and at a number of other relatively unpolluted monitoring sites

        January 2005                              8-6        DRAFT-DO NOT QUOTE OR CITE

-------
 1      throughout the Northern Hemisphere. For example, the highest O3 concentrations at
 2      Yellowstone National Park tend to occur during April and May.  Typically, monthly minima
 3      tend to occur from November through February at polluted sites and during the fall at relatively
 4      remote sites.
 5
 6      8.2.3 Long-Term Trends
 7           Nationwide, 1-h O3 concentrations decreased -29% from 1980 to 2003 and by -16% from
 8      1990 to 2003; and, for the 8-h standard, O3 levels decreased -21% since 1980 and -9% from
 9      1990 to 2003. Note that 1-h and 8-h O3 levels continue to decrease nationwide, but the rate of
10      decrease has slowed since  1990. These trends have not been uniform across the United States.
11      In general, O3 reductions have been largest in New England and in states along the West Coast
12      and smallest in the Midwest. Downward trends in O3 in California have been driven mainly by
13      reductions in Southern California, with reductions in other areas not being as large.
14
15      8.2.4 Ozone Interactions with Other Ambient  Pollutants
16           Data for other oxidants (e.g., H2O2, PAN) and oxidation products (e.g., HNO3, H2SO4)
17      in the atmosphere are not as abundant as they are for O3. Because data for these species are
18      usually obtained only as part of specialized field studies, it is difficult to relate O3 concentrations
19      to ambient levels of other species.  In general, these secondary species are expected  to be at least
20      moderately positively correlated with O3. On the other hand, primary species are expected to be
21      more highly correlated with each other than with secondary species, provided that the primary
22      species originate from common source areas. Relationships between ambient O3 and PM2 5
23      concentrations are complex, because particulate matter (PM) is not a single distinct chemical
24      species, but rather a mix of primary and secondary species. As an example of the subject
25      complexity, PM2 5 concentrations were positively correlated with O3 during the summer, but
26      negatively correlated with  O3 during the winter at Ft. Meade, MD.  More data are needed before
27      this result can be applied to other areas; and the degree of positive or negative correlation
28      between O3 and PM or other pollutants may vary to a greater or lesser extent by season.
29
30
        January 2005                               8-7         DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.3  AMBIENT OZONE EXPOSURE ASSESSMENTS
 2           Exposure to O3 and related photochemical oxidants varies with time due to changes in
 3      ambient concentration and because people move between locations with different O3
 4      concentrations. The amount of O3 delivered to the lung is not only influenced by the ambient
 5      concentration but also by the individual's breathing rate. Thus, activity level is an important
 6      consideration in determining the potential exposure and dose received.
 7           The use of ambient air monitoring stations is still the most common surrogate for assigning
 8      exposure estimates in epidemiological studies.  Since the primary source of O3 exposure is the
 9      ambient air, monitoring concentration data should provide a relative assignment of exposure
10      with time if: concentrations were uniform across the region; time-activitys pattern were the
11      same across the population; and housing characteristics, such as ventilation rates and O3 sinks
12      contributing to its indoor decay rates, were constant for the study area. Because these factors
13      vary by population and location, there tend to be errors not only in estimating the magnitude of
14      the exposure but also in relative exposure assignments based solely on ambient monitoring data.
15      Still,  such data can be used to evaluate health outcomes associated with chronic O3 exposure.
16
17      8.3.1  Personal Exposure
18           Personal O3 concentrations have been measured for children, outdoor workers, and
19      individuals with COPD,  populations potentially susceptible to respiratory irritants.  Children and
20      outdoor workers have somewhat higher exposures than other individuals, because they spend
21      more time outdoors engaged in moderate and heavy exertion.  Children are also more active
22      outside and, therefore, have a higher breathing rate than most adults. However, the available
23      exposure studies are not  sufficient to allow for confident generalization of differences in
24      exposure between the general population and potentially susceptible subpopulations.
25
26      8.3.2  Indoor Concentrations
27           There are few indoor sources of O3. Generally, O3 enters indoor environments through
28      infiltration from outdoors and through building components such as windows, doors, and
29      ventilation systems. The concentration of O3 in indoor environments is primarily dependent on
30      the outdoor O3 concentration and the air exchange rate (AER) or outdoor infiltration. Ozone
31      concentrations indoors are higher during the outdoor O3 season.

        January 2005                              8-8         DRAFT-DO NOT QUOTE OR CITE

-------
 1           Ozone reacts indoors with other contaminants, possibly producing compounds with greater
 2     toxicity.  Ozone concentrations are typically lower indoors than outdoors, in part due to gas
 3     phase reactions that produce other oxidants analogous to the production of photochemical smog.
 4     The production of these species indoors is a function of the indoor O3 concentration and the
 5     presence of the other necessary precursors, volatile organic compounds (VOCs) and nitrogen
 6     dioxide (NO2 ), along with an optimal AER.
 7           Several studies have measured O3 concentrations in residences, schools, office buildings
 8     and museums, and concentrations varied at all locations. However, indoor concentrations were
 9     generally associated with the AER in the indoor environment (increasing with higher AER) and
10     generally tend to be notably lower than outdoor ambient O3 levels. For example, one study
11     examining the relationship between O3 concentrations indoors and outside of a  school in
12     New England reported averaged O3 concentrations of 20 ppb indoors and 40 ppb outdoors.  With
13     regard to mobile source microenvironments, as is the case for other enclosed environments,
14     ozone exposures depend on the extend of mixing of outdoor air into the vehicle cabin.
15     If windows are kept open, ozone in the vehicle may be expected to approach outdoor values;
16     however, if windows are kept closed  and there is air conditioning, then interior  values could be
17     much lower than those outside, especially if recirculated air is used.  For example, in one N.C.
18     study involving police cars with air conditioning and recirculated air, O3 concentrations in the
19     vehicle cabin (11.7 ppb average) were less than half those outside (28.3 ppb average at outdoor
20     monitoring sites in the area).
21
22
23     8.4  SYNTHESIS OF AVAILABLE  INFORMATION ON
24           OZONE-RELATED HEALTH EFFECTS
25           The integrated synthesis of the latest available information on O3-related health effects
26     poses large challenges, especially in view of the emergence of important new information
27     generated since the 1996 O3 AQCD, which adds greatly to the complexity of any integrative
28     assessment. Such information includes new findings from:
29       •  Epidemiological studies, reflecting progress in addressing many research recommendations
           from the last review as well as raising new issues and reevaluating previously addressed
           issues that remain important in interpreting the body of epidemiological evidence and
           characterization of its strengths and limitations;

       January 2005                              8-9        DRAFT-DO NOT QUOTE OR CITE

-------
 1       •  Experimental toxicological studies using laboratory animals and controlled human
            exposures aimed at understanding the potential biochemical mechanisms underlying toxic
            effects, pathology, and susceptibility.
 2      Thus despite substantial progress, challenges remain in integrating the new scientific information
 3      on O3 health effects, including newly reported epidemiological evidence for associations
 4      between ambient O3 exposures and increased mortality risks among human populations.
 5
 6      8.4.1   Assessment of Epidemiological Evidence
 7           Based on the O3 epidemiological evidence available at the time, the 1996 O3 AQCD
 8      arrived at the following conclusions:
 9
10                 An association between daily mortality and O3 concentration for areas with high O3
11                 levels (e.g., Los Angeles) has been suggested, although the magnitude of such an effect
12                 is unclear. Increased O3 levels are associated with increased hospital admissions and
13                 emergency department visits for respiratory causes. Analyses from data in the
14                 northeastern United States suggest that O3 air pollution is associated with a substantial
15                 portion (on the order of 10 to 20%) of all summertime respiratory hospital visits and
16                 admissions. Pulmonary function in children at summer camps in southern Ontario,
17                 Canada, in the northeastern United States, and in Southern California is associated
18                 with O3 concentration." (U.  S. EPA, 1996, pl-29).
19
20      The 1996 O3 AQCD further stated that only suggestive epidemiologic evidence existed for health
21      effects of chronic ambient O3 exposure in the population, and this was partly due to an inability
22      to isolate potential effects related to O3 from those of other pollutants, especially PM (U.S.
23      Environmental Protection Agency,  1996).
24           The scientific strength and limitations of the growing body  of epidemiological evidence for
25      associations between exposure to O3 and health effects discussed in this  section is based
26      primarily  on Chapter 7 evaluations. The following criteria were considered in assessing the
27      relative scientific quality of the epidemiologic studies:  (1) quality of exposure metrics to
28      evaluate credible exposure indicators; (2) quality and size of the study groups/population to
29      arrive at meaningful analysis of health effects; (3) robustness of reported associations (based on
30      defined health endpoint criteria),  potential confounding by copollutants; (4) the strength of
31      reported associations, in terms of magnitude, statistical significance and statistical power of
32      effects estimates; (5) temporality, in terms of lag periods between exposure and observed effects;
33      and (6) biological plausibility, consistency and coherency of the reported findings.  The body of
34      epidemiological evidence is further considered in terms of its coherence within itself and in

        January 2005                                8-10        DRAFT-DO NOT QUOTE OR CITE

-------
 1      relation to findings derived from controlled human exposure studies which, overall, provide
 2      insights into the plausibility of reported O3 human health effects reflecting causal relationships.
 3           Many newly available epidemiological studies have provided additional evidence for
 4      O3-related health effects beyond that which was known previously.  Significant statistical
 5      associations have been observed by various investigators between acute O3 exposure and several
 6      respiratory health endpoints, including: mortality; hospital admissions; emergency department
 7      visits; respiratory illness and symptoms; and changes in pulmonary function. Similarly, long-
 8      term exposure to O3 has been associated with: increased morbidity; development of respiratory
 9      disease; and declines in lung function and lung function growth.  The epidemiological studies
10      that have been conducted in areas across the United States and Canada, as well as in Europe,
11      Latin America, Australia and Asia, are summarized in Annex 7. Based on evidence extracted
12      from the full body of epidemiologic studies that have been carried out and reviewed since the
13      1996 O3 AQCD (U.S. Environmental Protection Agency, 1996), it has been well demonstrated
14      that deleterious human health outcomes are positively associated with ambient O3 concentrations
15      currently  encountered in the United States and elsewhere.
16
17      8.4.2  Strength of Epidemiological Associations
18           As quoted above, assessments in the 1996 O3 AQCD supported a consistent relationship
19      between O3 concentration and respiratory illness,  hospital visits and reduced lung function.
20      However, due to insufficient evidence examining O3-mortality associations and uncertainties
21      regarding weather model specification, the 1996 O3 AQCD was limited to only a very qualitative
22      assessment  of O3-mortality associations. Since then, generalized Additive Models (GAMs) have
23      become widely utilized for epidemiologic analysis of health effects attributable to air pollution,
24      making quantitative estimation of O3-mortality risks much more meaningful. On the other hand,
25      certain statistical issues raised with regard to use of default convergence criteria in applications
26      of commercially available software employed for GAM analyses in many newly available air
27      pollution  epidemiologic studies led to a reanalyses of previously published studies and revised
28      estimation of reported PM — mortality/morbidity risks. The impacts of the GAM-related
29      statistical issues were thoroughly discussed in the 2004 PM AQCD (U.S. Environmental
30      Protection Agency, 2004).  Of most importance here, the reanalyses of a number of studies,
31      comparing results using default GAM convergence criteria to results from analyses using

        January 2005                              8-11        DRAFT-DO NOT QUOTE OR CITE

-------
 1      stringent GAM convergence criteria and/or from GLM analyses, found little difference among
 2      the O3 effect estimates obtained (as discussed in detail in Chapter 7 of this document).
 3      Furthermore, the magnitude of the effect-size estimates observed from O3-mortality relationships
 4      tend to be relatively consistent across the newly available studies and to compare well with those
 5      obtained for O3-morbidity endpoints.
 6
 7      8.4.3 Acute Exposure Studies
 8           Numerous epidemiological studies carried out over the past decade have added evidence to
 9      the knowledge base that was assessed in the 1996 O3 AQCD, which included both (a) individual-
10      level camp and exercise studies that established a relationship between human lung function
11      decline with ambient O3 exposure and (b) aggregate time-series studies that suggested positive
12      relationships for O3-related respiratory  morbidity. The new studies reviewed in Chapter 7 in this
13      document included numerous field/panel studies and time-series studies from various regions.
14      In field studies on the effects of air pollution exposure, the most common health outcomes
15      measured were lung function and respiratory symptoms.  Time-series studies examined daily
16      hospital admissions, emergency department visits, and mortality data.
17
18      8.4.3.1 Panel Studies
19           Many of the new field/panel  studies reviewed in Chapter 7 and the controlled human
20      exposure studies reviewed in Chapter 6 of this document provide additional data supporting two
21      major findings reported in the 1996 O3  AQCD, i.e.: (1) O3-related lung function decrements and
22      (2) respiratory symptoms in exercising healthy subjects and asthmatic subjects. Pulmonary
23      function was determined by either spirometry (forced expiratory volume in 1  s [FEVJ and
24      forced vital capacity [FVC]) or by peak expiratory flow (PEF) meters. While the spirometric
25      parameter, FEVj is a stronger and more consistent measure of lung function, PEF is more
26      feasibly performed in field studies.
27           In a number of newly available field/panel studies, FEVj was measured in panels of
28      exercising children, outdoor workers, and adult hikers exposed to ambient O3 while experiencing
29      elevated exertion levels. Collectively, the results of the new studies (discussed in Section
30      7.2.3.1) confirm and extend those from analogues field/panel studies assessed in the 1996 O3
31      AQCD and findings from experimental controlled human exposure studies indicating that acute

        January 2005                              8-12        DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3 exposures prolonged over several hours and combined with elevated levels of exertion or
 2      exercise magnify O3 effects on lung function, as evaluated in terms of FEVj.
 3           For example, six field studies by three different research groups of 7- to 17-year-old,
 4      healthy (nonasthmatic) children exposed for several hours to ambient O3 during increased
 5      physical exertion in summer camp activities were assessed in the 1996  O3 AQCD. When
 6      analyzed together by consistent statistical methods, the data from those studies showed an
 7      average relationship between afternoon FEVj and concurrent 1-h O3 concentrations of
 8      -0.50 mL/ppb, with individual slopes ranging from  -0.19 to -1.29 mL/ppb (likely reflecting, in
 9      part, the multi-hour O3 exposures preceding the pulmonary function tests). Four new filed/panel
10      studies assessed in Section 7.2.3.1 of this document that evaluated pulmonary function in healthy
11      school-aged children exposed to mean 1-h O3 concentrations ranging from -20 to 120 ppb found
12      exposure-response functions of approximately -0.23 to -1.42 mL/ppb. Also, two other studies
13      assessed in the 1996 document that measured lung function before  and after well-defined
14      exercise events (0.5-h long) in adults during exposures to ambient O3 across 4 to  135 ppb found
15      exposure-response slopes of -0.4 mL/ppb.  In comparison, four new studies of healthy adult
16      workers (street workers, berry pickers) and hikers engaged in prolonged (> 6 to 8 h) strenuous
17      physical exertion while exposed to mean ambient O3 concentrations of-26 to 70  ppb  (1-h
18      maximum) or 40 ppb (8-h average) reported exposure-response slopes of-1.13 to -3.8 mL/ppb
19      (as assessed in Chapter 7 of this document). The most representative data  is that  of Korrick et al.
20      (1998) from a U.S. study of adult hikers that provided outcome measures stratified by gender,
21      age, smoking-status, and presence of asthma within  a population capable of above-normal
22      exertion.
23
24      8.4.3.2 Asthma Panels
25           Several studies assessed in the 1996 O3 AQCD that evaluated elevated respiratory
26      symptoms and/or pulmonary function decrements in asthmatic children showed greater
27      responses in asthmatic than nonasthmatic subjects, suggesting that asthmatic individuals might
28      constitute a sensitive population group in oxidant epidemiologic studies.
29           Additional panel studies carried out over the past decade to understand the effect of acute
30      exposure to O3 in asthmatics evaluated either (a) lung function by PEF and/or (b) respiratory
31      symptoms (i.e. cough, wheeze, shortness of breath and medication use) ascertained by

        January 2005                              8-13        DRAFT-DO NOT QUOTE OR CITE

-------
 1      questionnaire.  Several regional studies, typically consisting of children with asthma, collectively
 2      tend to confirm O3-induced decrements in pulmonary function both in the United States and in
 3      other countries (see Section 7.2.3.2). One U.S. multicity study (Mortimer et al., 2002), featuring
 4      the largest panel of asthmatic children from eight urban areas, observed a statistically significant
 5      decrement in PEF with a cumulative lag of 1 to 5 days  (-1.18% per 30 ppb increase in 8-h
 6      average O3). Overall, these studies suggest that ambient O3 exposures may be associated with
 7      enhanced decreases in lung function in asthmatics.
 8           Most studies evaluating respiratory symptoms (i.e. cough, shortness of breath, and wheeze)
 9      and the increased use of asthma medications related to  O3 exposure also focused on asthmatic
10      children.  Several U.S. studies observed significant associations between O3 ambient
11      concentrations and increased symptoms or asthma medication use that appeared to be fairly
12      robust to adjustment for copollutants.  Analyses by Mortimer et al. (2002), conducted in eight
13      U.S. urban areas, and Gent et al. (2003), conducted in the New England region, have used data
14      from large sample populations likely to be representative of U.S. data. Odds ratios from six new
15      studies for prevalence of cough among asthmatic children mainly varied from ~1.05 to 1.5
16      standardized per 40 ppb increase in  1-h max O3 (or equivalent) or -1.35 for all symptoms per
17      30 ppb increase in 8-h O3 concentration.
18
19      8.4.3.3 School Absences
20           Two new U.S.  studies (Chen et al., 2000; Gilliland et al., 2001) investigated the
21      relationship between ambient O3 concentrations and school absenteeism.  Both studies were
22      carried out during a period when O3 levels were mostly below the highest levels typically
23      observed in the summer season.  In the Chen et al. study, with a distributed lag of 1 to 14 days, a
24      10.4% excess rate of school absences was found per 40 ppb increase in daily 1-h max O3
25      concentrations. The study by Gilliland et al. (2001), which was able to distinguish specific
26      illness-related absence, found significant O3 effects on  school absences due to respiratory causes
27      with a lag period ranging from 2 to 4 weeks. A notably higher respiratory-related absence rate
28      increase (147% increase per 30 ppb  increase in 8-h O3) was seen versus that seen for
29      non-respiratory causes (61% increase per 30 ppb).
30
        January 2005                              8-14        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.3.4  Field Studies on Cardiovascular Effects
 2           A limited number of air pollution studies have examined cardiac physiologic endpoints,
 3      including heart rate variability, arrhythmia, and risk of myocardial infarction.  One large U.S.
 4      study (Liao et al., 2004) found that ambient O3 concentrations were inversely associated with
 5      ECG high-frequency power readings among whites. However, consistently more pronounced
 6      associations were suggested between PM10 and heart rate variability among persons with a
 7      history of hypertension. While these results may somewhat be supportive of hypothesized air
 8      pollution-heart rate variability-cardiovascular disease pathways at the population level, a lack of
 9      consistency within or across the limited available studies indicates that additional studies are
10      needed before any clear conclusions can be made.
11
12      8.4.4  Emergency Department Visits and  Hospital Admissions
13           Many time-series studies reviewed in the 1996 O3 AQCD indicated positive associations
14      between O3 air pollution and increased hospital admissions. Strong evidence establishing a
15      correlation between O3 exposure and increased exacerbations of preexisting respiratory disease
16      in the general public were reported at 1 h-maximum O3 concentrations < 0.12 ppm.  Several
17      studies have been published over the past decade examining the temporal associations between
18      O3 exposures and emergency department visits for respiratory diseases (see Table AX7-2 in
19      Annex 7). Among studies with adequate controls for seasonal patterns, many reported at least
20      one significant positive association involving O3. Overall, the analyses of data for asthma-
21      related emergency room visits clearly indicate increased respiratory morbidity during warm
22      seasons when ambient O3 concentrations are high. These studies are summarized in Figure 8-1,
23      showing both yearly and seasonal results from U.S. and Canadian studies.
24           Studies reviewed in Chapter 7 (Section 7.3.3) reported a significant O3 effect on respiratory
25      hospital admissions. While some inconsistencies are noted across studies, the evidence is
26      supportive of significant and robust O3 effects on hospitalizations for various respiratory
27      diseases.  Large multeity studies, as well as many studies from individual cities, have reported
28      significant O3 associations with total  respiratory asthma and chronic obstructive pulmonary
29      disease (COPD) hospitalizations, especially in studies analyzing the O3 effects during the
30      summer or warm season (Figure  8-2). The most robust and informative results on the effects of
31      O3 on respiratory hospital admissions are from muticity studies that used a consistent analytical

        January 2005                               8-15         DRAFT-DO NOT QUOTE OR CITE

-------
                             Percentage Change in Emergency Department Visits for Asthma
               Jaffeetal. (2003):
           Cincinnati,OH (age 5-34)
           Cleveland OH (age 5-34)
           Columbus,OH (age 5-34)
               Jaffe etal. (2003):
             Cincinnati, Cleveland,
        and Columbus,OH (age 5-34)
            Friedman etal. (2001):
             Atlanta, GA (age 1-16)
               Peel etal. (2004):
             Atlanta,GA (all ages)
             Tolbert etal. (2000):
             Atlanta, GA (age 0-16)
               Stieb etal. (1996):
          St.John,Canada (all ages)
-20 0 20 40
i i 1 i i i i i

• Inn °
• ho °
• laa 3

• hn 0


^X- lag 0-2
• lag 0-2 (SE not given)
	 X 	 Iag1

60 80 1(
i i i

X All year

• Warm


lag 0 1
ho 0 ">


       Figure 8-1. Ozone-associated percent change (95% CI) in emergency room visits for
                   asthma. Percent change effects are per 40 ppb increase in 1-h maximum O3 or
                   equivalent. Analysis includes all age unless otherwise noted, and only studies
                   conducted in the United States and Canada are presented.
 1     methodology across a broad geographic area, such as the 16 Canadian cities study by Burnett
 2     et al. (1997a). Of the few studies that examined the relationship between O3 and hospital
 3     admissions for cardiovascular diseases, most did not find any consistent positive associations.
 4
 5     8.4.5  Acute Effects of Ozone on Mortality
 6           Due to the limited number of studies and uncertainties regarding weather model
 7     specifications, no meaningful quantitative assessment of O3-mortality associations were possible
 8     in the 1996 O3 AQCD. However, newly available large multicity studies designed specifically to
 9     examine the effect of O3 on mortality have provided much more robust and credible information.
10     The results from two key studies carried out in 95 U.S. communities (U.S. National Morbidity,
11     Mortality Air Pollution Study [NMMAPS]; Bell et al., 2004) and in 23  European cities (Air
12     Pollution on Health: European Approach [APHEA]; Gryparis et al., 2004) showed positive and
13     significant O3 effect estimates for all cause (nonaccidental) mortality (Figure 8-3; see Section 7.4
14     of Chapter 7 for complete discussion).  The influence of season on O3-mortality risk estimates
15     from various U.S. and Canadian time-series studies is also shown in Figure 8-3. In the APEHA
       January 2005
8-16
DRAFT-DO NOT QUOTE OR CITE

-------
                 Linn etal. (2000):
           Los Angeles, CA (age 30+)

             Schwartz etal. (1996):
            Cleveland, OH (age 65+)

              Burnett etal.(1997a):
          16 Canadian Cities (all ages)

             Burnett etal.(1997b): .
           Toronto,Canada (all ages)

               Burnett etal. (2001):
           Toronto, Canada (age< 2)

             Luginaah etal. (2004):
                 Windsor,Canada •
               (all ages, males only)

             Luginaah etal. (2004):
                 Windsor,Canada •
             (all ages,females only)
                                    Percentage Change in Respiratory Hospitalization
                               -20         0         20         40         60         80        100
                                 i	i	i	i	i	i	i	i	i	i	i	i
                                            -x-
-e-
       lagO
          - lag 1-2


          Iag1
                lag 1-3
            Iag1
                             lag 0-4
        Figure 8-2. Ozone associated percent change (95% CI) in total respiratory hospitalizations
                    (95% CI) for all year and for by season. Percent change effects are per
                    40 ppb increase in 1-h maximum O3 or equivalent. Analysis includes all age
                    unless otherwise noted, and only studies conducted in the United States and
                    Canada  are presented.
 1      study (Gryparis et al., 2004), significant O3 effects were observed only during the warm season.

 2      With the exception of one study (Chock et al., 2000), all risk estimates from warm-season-only

 3      analyses were positive, with the majority indicating statistical significance at p < 0.05.

 4           The effect of PM on mortality was thoroughly discussed in the 2004 PM AQCD. Because

 5      PM indices correlate highly with O3 levels in some areas, confounding of the O3-mortality

 6      association by PM is of great concern.  Figure 8-4 shows O3-mortality risk estimates with and

 7      without adjustment for PM indices. Collectively, the results indicate that the O3 risk estimates

 8      were not substantially affected with the addition of PM in the various reported analyses.

 9           The effect estimates presented in Figures 8-3 and 8-4 lead to the following findings:

10      (1) O3-mortality associations from several U.S. and Canadian studies reported fairly consistent

11      and positive combined estimates of 0.4 to 4.8% excess risk of total nonaccidental mortality per

12      40 ppb increase in 1-h maximum O3 (excluding the Vedal et al., 2003 study, which examined the
        January 2005
            8-17
DRAFT-DO NOT QUOTE OR CITE

-------
                 Bell etal. (2004): U.S. 95 communities -

                       Sametetal.(2000;reanalysis _
                    Dominici et al. 2003): U.S. 90 cities

                       Schwartz (2004): U.S. 14 cities -
      Kinney and Ozkaynak (1991): Los Angeles County, CA —
             Kinneyetal.fi 995): Los Angeles County, CA —
                   Ostro (1995): San Bernadino County
                          and Riverside County, CA
                  Fairley (2003): Santa Clara County,CA -

                          Gamble (1998):Dallas,TX -

                    Dockeryetal.fi 992): St. Louis, MO -
               ItoandThurston (1996): Cook County, IL -
      Lippmannetal.(2000;reanalysisIto,2004): Detroit,Ml -
                  Lipfertetal.(2000a): Philadelphia, PA -
                  Lipfert et al. (2000a): 4 Counties in PA -
             Lipfert et al. (2000a): 7 Counties in PA and NJ -
     Lipfert et al. (2000a): 7 Counties in PA and NJ (age 0-65) -
     Lipfert etal.(2000a): 7 Counties in PA and NJ (age 65+) -

               Moolgavkar et al. (1995): Philadelphia, PA -

            Chock etal. (2000): Pittsburgh, PA (age 0-74) -

             Chock etal. (2000): Pittsburgh, PA (age 75+) -

               Dockery et al. (1992): Eastern Tennessee -
                 Klemm and Mason (2000): Atlanta, GA -
              Klemm etal. (2004): Atlanta, GA (age 65+) -

                 Vedaletal.(2003):Vancouver,Canada -
                                                          Percentage Change in Mortality
                                                      -10
              0
              I
                                                                                  10
                         20
                        _l	
 30
_l	
                                                Multiple Cities/Regions
               rX-
                                                Single City/Region
               x  Iag1
               •X-  Iag1
          lag 1-2
          lag 0-1
          lag 0-1
          lag 0-1
          lag 0-1
          lag 0-1
          -e-
       -e-
lagO
                  lag 0-6
                   Iag1
                         lagO
    •  (SE not given; significant at 0.05 level)
  O  (SE not given; not significant at 0.05 level)
                        lag 0-1
                   lag 0-3
     ' Confidence bands not provided but
      noted as significant at p=0.055
                       Iag1
                     lagO
                        Iag1
                               lag 0-1
Figure 8-3.  All cause (nonaccidental ) O3 excess mortality risk estimates (95% CI)
                for all year and for by season per 40 ppb increase in 1-h maximum O3
                or equivalent.  Analysis includes all age unless otherwise noted, and
                only studies conducted in the United States and Canada are presented.
January 2005
 8-18
DRAFT-DO NOT QUOTE OR CITE

-------
                                                       Percentage Change in Mortality
                 Samet et al. (2000, reanalysis
             Dominicietal.,2003): U.S. 90 cities —
                              with PM-|o ~
                 Schwartz (2004): U.S. 14 cities -
                              with PM10 -

Kinney and Ozkaynak (1991): Los Angeles County, CA —
                               with KM -

       Kinney etal.(1995): Los Angeles County, CA —
                              with PM10 —

         Ostro et al. (1995): San Bernadino County
                    and Riverside County, CA —
                              with PM2 5 —

           Fairley (2003): Santa Clara County,CA -
                              withPM25 -

                   Gamble (1998): Dallas,TX -
                              with PM10 -

         ItoandThurston (1996): CookCounty.lL —
                              with PM-|o ~~

           Lipfert etal.(2000a): Philadelphia, PA -
                              with PM2 5 —
                              with PM10 -

           Lipfert et al. (2000a): 4 Counties in PA —
                              with PM2 5 —
                              with PM-|o —

      Lipfert et al. (2000a): 7 Counties in PA and NJ -
                              with PM2 5 —
                              with PM-|o —

      Lipfert et al.(2000a): 7 Counties in PA and NJ
                               (age 0-65) -
                              with PM2 5 -
                              with PM10 —
      Lipfert et al.(2000a): 7 Counties in PA and NJ
                               (age 65+) —
                              with PM2 5 -
            Moolgavkar et al. (1995):Philadelphia, PA —
                                   with TSP -

            Moolgavkar et al. (1995):Philadelphia, PA —
                                   with TSP -

         Chock et al. (2000): Pittsburgh, PA (age 0-74) —
                                 with PM-|o —

          Chock et al. (2000): Pittsburgh, PA (age 0-74) -
                                 withPMjQ —

          Chock et al. (2000): Pittsburgh, PA (age 0-74) -
                                 withPM10 -

          Chock et al. (2000): Pittsburgh, PA (age 75+) -
                                 with PM10 -

          Chock et al. (2000): Pittsburgh, PA (age 75+) -
                                 withPM10 -

          Chock et al. (2000): Pittsburgh, PA (age 75+) -
                                 withPM10 -
                                                     -10
                                                      I
        -5
         I
                                                               lagO
                                                                         0
                                                                          |
10
 I
 15
	I
                                                                          lag 0-1 •
                                                                          lag 0-1
                                                                          lag 0
                                                                          lag 0-1
                                                                          lag 0-
                -e-
                                                                                  lagO

                                                                                  lagO

                                                                                  Iag1

                                                                                  Iag1


                                                                                   lagO
                                           All Year
                                         O  03 only
                                         •  03 with PM
                                        Warm Season
                                         A  03 only
                                         A  03 with PM
                                         Cool Season
                                         D  03 only
                                         •  O3withPM
                                                                                                lag1-2
                                                                                  O
                                                                                  O
                                                                                   O
                                                                              -1   •
                                                                                   0
                         0
                      -1  •
                         •
                                                                                            lag 1-2
                                     Confidence bands not
                                     provided but noted as
                                     significant at p=0.055
                                                                                              Noted as nonsignificant
                                                                                              at p=0.055
                                                                                              Confidence bands not
                                                                                              provided but noted as
                                                                                              significant at p=0.055
                                                                               ^~lag1

                                                                                Iag1

                                                                               ""  lagO
                              lagO

                           1  lagO

                            lag 0

                                 lagO

                                 lagO
Figure 8-4.   All cause (nonaccidental) O3 excess mortality risk estimates (95% CI)
                 with adjustment for PM indices for all analyses. Percent change effects
                 are per 40 ppb increase in 1-h maximum O3 or equivalent, and adjusted
                 for PM indices. Only U.S. and Canada studies are presented.
January 2005
8-19
                                                                     DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3-mortality association in a region with very low O3 concentrations); (2) season-stratified
 2      analyses indicated that the O3-mortality effect estimates were significant and positive in the
 3      warm season, with larger effects observed compared to the year-round and cool-season analyses;
 4      (3) the risk estimates were robust to adjustment for PM indices, indicating that O3 effect on
 5      mortality is independent of PM.
 6           The results from the U.S. studies are generally consistent with those from other regions of
 7      the world.  The results  from all available studies indicate substantial strength in the
 8      epidemiological evidence for the association between exposure to O3 and excess risk of total
 9      nonaccidental mortality.  The overall range of estimates were relatively  narrow with the positive
10      estimates between 0 and 7% per 40 ppb increase in 1-h maximum O3 or equivalent.
11
12      8.4.6  Chronic Ozone Exposure Studies
13           There were a limited number of studies reported in the 1996 O3 AQCD that provided
14      insufficient evidence to consider potential health effects of long-term ambient O3 exposures.
15      Several longitudinal epidemiological studies carried out in the past decade evaluated the
16      potential effects of chronic exposure (several weeks to many years) to O3 on lung function,
17      respiratory symptoms,  lung inflammation, asthma prevalence, and mortality.
18           Evidence from recent long-term morbidity studies indicates that chronic exposure
19      to O3 may have negative effects on inflammation, respiratory symptoms, and development of
20      asthma; however, the evidence is limited and, at times,  lacks consistency.  The strongest
21      evidence for an effect from chronic O3 exposure is derived from studies examining lung function
22      measurements. Seasonal decrements or reduced growth in lung function measures have been
23      reported in several studies; however, the changes appear to be transient. Studies of lung function
24      decrements with longer-term or annual data are not as conclusive.
25           Very few studies have investigated the effect of long-term O3 exposure on mortality.
26      Uncertainties regarding the exposure period of relevance and inconsistencies across mortality
27      outcomes and gender raise concerns regarding plausibility.  The most representative U.S.  study
28      by Pope et al. (2002) observed positive but non-significant associations  between O3 exposure
29      and all cause mortality. Thus, the current evidence is inconclusive for a relationship between
30      chronic O3 exposure and increased mortality risk.
31

        January 2005                              8-20        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.7  Robustness of Epidemiological Associations
 2           In evaluating the strength of the epidemiological evidence, the magnitude of observed O3
 3      effect estimates and their statistical significance is important; however, consideration must be
 4      given to the precision of the effect estimates and the robustness of the effects associations.
 5      Examining the robustness of the associations includes the impact of alternative models, model
 6      specifications for temporal trends and meteorological factors, and potential confounding by
 7      copollutants.  Also of interest are issues related to exposure assessment and measurement error.
 8      A detailed discussion on each of these topics can be found in Chapter 7 (Section 7-6).  The
 9      following sections focus on the extent to which the current epidemiological findings can be
10      considered robust.
11
12      8.4.7.1  Exposure Issues:  Ambient versus Personal
13           In time-series studies, or other large-scale epidemiology studies of long duration, it is often
14      impractical and unfeasible to monitor the personal exposure of each subject. Thus, the ambient
15      concentrations of O3 and other air pollutants at central monitoring sites are often used as a proxy
16      for individual exposure measurements.  The relationship between ambient O3 concentrations and
17      personal O3 exposure levels varies depending on factors such as time spent outdoors, ventilation
18      conditions, personal factors, and air quality indices. Because ambient concentrations often
19      overestimate true personal O3 exposures, the use of ambient data  likely tends to underestimate
20      the effect of the air pollutant on health.
21           Comparisons between ambient concentrations and personal exposures to O3 have indicated
22      that ambient concentrations do not reflect the variability of individual exposures. However,
23      daily ambient O3 concentrations have been shown to be well-correlated to daily-averaged
24      personal exposures obtained by aggregating the personal measurements from all subjects.
25      Therefore, though unresolved issues remain, the  evidence suggests that ambient O3 levels
26      measured at central monitors may serve as valid  surrogate measures for aggregate personal
27      exposures in time-series studies investigating mortality and hospitalization outcomes.
28           Ambient O3 measurements from the three main exposure indices (1-h maximum O3, 8-h
29      maximum O3, and 24-h average O3) used were highly correlated.  As such, the excess health risk
30      estimates and significance of associations appear to be comparable for the same distributional
31      increment. The commonly used 8-h maximum O3 or 8-h average O3 index continue to be

        January 2005                              8-21        DRAFT-DO NOT QUOTE OR CITE

-------
 1      appropriate choices, as no other exposure index has been demonstrated to offer a better
 2      advantage.
 3
 4      8.4.7.2  Confounding by Temporal Trends and Meteorologic Effects
 5           The effect of seasonal differences in the health outcomes and O3 exposure levels were
 6      recognized in the 1996 O3 AQCD; and this issue is discussed in detail in Section 7.6.5 of this
 7      document. Two important factors, i.e., temporal trends and meteorological factors must be
 8      considered in evaluating O3 health effects estimates. In the U.S. 95 communities study (Bell
 9      et al., 2004), sensitivity analyses indicated that the O3 risk estimates were robust to tripling the
10      degrees of freedom for smoothing terms used to control for temporal trends.  In  a case-crossover
11      study by Schwartz (2004), the O3-mortality risk estimates from an analysis using nonlinear
12      regression splines to control for temperature were similar to those from an analysis that matched
13      on temperature, indicating that the effect estimates were not sensitive to methods used to control
14      confounding by temperature.
15           Analysis of O3 health effects is further complicated in view of the fact that the relationship
16      of O3 with temperature and with other pollutants appears to change across seasons.  As shown in
17      Figures 8-4, the O3 effect estimates from warm season data were consistently larger compared to
18      those calculated using all-year data and cool-season data.  In a study of daily hospital admissions
19      (Burnett et al., 2001), season-stratified analyses appeared to effectively control confounding by
20      season.
21           In summary, adjusting for temporal trends and meteorological factors is critical to
22      obtaining meaningful O3-effect estimates. Seasonal analyses indicate that mortality and
23      morbidity data computed using year-round data need to be interpreted with caution.
24      Air pollution epidemiological studies that integrate sensitivity analyses for seasonal
25      stratification, meteorological factors, and multipollutant models may provide a better and
26      more comprehensive understanding of the health effects estimates.
27
28      8.4.7.3  Assessment of Confounding by Copollutants
29           The presence and influence of PM and other gaseous copollutants have to  be considered in
30      assessing O3-health effects associations found by observational studies.  The potential for
31      copollutant confounding in the epidemiological time-series studies was assessed in some detail

        January 2005                               8-22        DRAFT-DO NOT QUOTE OR CITE

-------
 1      in Section 7.6.6.  Multipollutant modeling is the most common method used to test for potential
 2      confounding in epidemiological studies; however, interpretation of the results is often
 3      complicated by the high degree of correlation among air pollutants. The O3 mortality risk
 4      estimates from two-pollutant models adjusted for PM are presented in Figure 8-4 (U.S. and
 5      Canadian studies only). In the two multicity studies analyzed here, the addition of PM10 did not
 6      substantially change the risk estimates (Samet et al., 2000; Dominci et al., 2003; Schwart, 2004).
 7      The O3-mortality effects in single-city studies also were robust after adjusting for PM10 indices,
 8      both in all-year and season-stratified analyses data.
 9           In summary, assessing the health effects attributable to O3 is very challenging, even with
10      well-designed studies. Definitive partitioning out of the individual pollutant-specific health
11      outcomes from among an ambient mixture of multiple components is very difficult due to the
12      dynamic nature of their interactions over time. However, the new limited time-series studies that
13      made an exhaustive survey using populations from multiple U.S.  cities do provide substantial
14      epidemiological evidence indicating that associations for O3 with mortality and morbidity are
15      robust to confounding by copollutants.
16
17      8.4.7.4 Lag Period between Ozone Exposure and Health Response
18           The lag times between causes and  effects depend on underlying biological mechanism
19      involved in the process as well as the hypotheses tested. Different lag periods are appropriate for
20      assessing different health outcomes.  As discussed in Section 7.6.4, examining longer lag periods
21      may be needed to understand more fully the O3-related health outcomes.  The most significant
22      associations between O3 concentrations and mortality and respiratory hospitalization were
23      observed with 0-day and 1-day lags. These associations generally diminished with increased lag
24      days. In the 95 U.S. communities studies (Bell et al., 2004),  the mortality risk estimated over
25      multiple days (cumulative lag of 0 to 6 days) using distributed lag models indicated an effect of
26      O3 that was twice as large as the effect estimated using 1-day lags.  It should be noted that when
27      there is a pattern of effects  across lag periods, selecting the 1-day lag effect estimate is likely to
28      underestimate the overall effect size and does not fully capture the risk distributed over adjacent
29      days. Longer averaging periods may  aid in characterizing cumulative O3-related effects  over
30      several days; however, interpreting these results may not be straightforward.
31

        January 2005                              8-23        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.7.5  Concentration-Response Functions and Threshold
 2           Ozone concentration-response relationships have been explored in several studies with
 3      various health outcomes, including mortality, hospitalizations, emergency department visits,
 4      lung function, and respiratory symptoms. While some studies found no threshold for O3 health
 5      effects, others have found that a low-level threshold may be present. Note that an absence of a
 6      detectable threshold in population studies does not necessarily indicate an absence of individual
 7      thresholds and, conversely, evidence of a threshold for individuals does not necessarily indicate
 8      a population threshold due to variability in response among individuals in the population. With
 9      the current evidence, no definitive conclusion can be made regarding the threshold issue;
10      however, the limited evidence suggests that the possible threshold level may be well below the
11      current O3 standard level. The distribution of potential thresholds, particularly around the
12      NAAQS value of 80 ppb for 8-h maximum O3, needs to be further investigated.
13
14      8.4.7.6  Summary and Conclusions for Epidemiology Findings
15           Discussions presented in the previous sections evaluated the merits of the epidemiological
16      studies to derive judgments about the potential causal relationship between O3 exposures and
17      health outcomes. These evaluations were carried out in the context of the  criteria listed in
18      Section  8.2.1. Information with regard to one of the criteria, i.e., coherence and biological
19      plausibility, is discussed in the section following the next one, which undertakes to provide an
20      integrated analysis of the biological evidence from human and animal toxicology studies with
21      the epidemiological  evidence.
22           The results from the new field/panel studies evaluated in this document provided additional
23      evidence for likely causal relationships being reflected by  significant associations between acute
24      O3 exposure and  (a)  decrements in lung function, (b) respiratory symptoms, and (c) increased use
25      of asthma medication in children and, in some cases, adults. Similarly, significant positive
26      associations can be inferred between acute O3 exposure and respiratory morbidity indexed by
27      hospital admissions  and emergency visits, especially based on season-stratified data.  The results
28      from large multicity studies suggest an elevated risk of mortality for acute exposure to O3;
29      however, the magnitudes of these estimates are small.  Analysis of the data from chronic
30      mortality and morbidity studies indicate some significant associations between O3 and seasonal
        January 2005                              8-24        DRAFT-DO NOT QUOTE OR CITE

-------
 1      changes in lung function; but, overall, the strength of the data does not allow establishment of a
 2      conclusive relationship to O3 as a causal factor for other observed health outcomes.
 3           Issues regarding strengths of models used in air pollution epidemiology were carefully
 4      considered. There have been improvements in the modeling to adjust for potential confounding
 5      variables, including temporal trends, meteorological factors, and copollutants. However, more
 6      sensitivity analyses would still be useful to examine the extent of adequate adjustment for
 7      confounding by these factors. Results from multipollutant models indicate that copollutants,
 8      e.g., PM, generally do not confound the association between O3 and acute health outcomes,
 9      suggesting an independent effect of O3.
10           In conclusion, the epidemiological evidence continues to support likely causal associations
11      between O3 and acute respiratory morbidity and mortality, based on the assessment of strength,
12      robustness, and consistency of results reported from numerous studies reviewed in Chapter 7.
13      Substantial evidence is lacking, however, by which to convincingly establish a positive
14      association between chronic O3 exposure and respiratory morbidity and mortality.  Additional
15      investigations are needed to further understand the health effects resulting from long-term O3
16      exposure.
17           The positive associations for increased morbidity and mortality risk estimates during
18      warmer seasons (when O3 concentrations tend to be high) support a causal role for O3 in
19      affecting human health. Though seasonality in O3-related health effects was observed in both
20      time-series and longitudinal cohort studies, no clear evidence of a threshold for O3 effects has yet
21      been found.
22
23      8.4.8  Integration of Experimental and Epidemiologic Evidence
24           In this section, effects are made to integrate the epidemiological evidence discussed above
25      with results of human and animal experimental studies carried out in vivo and in vitro, to
26      understand O3-induced alterations at the physiological, pathological, and biochemical levels of
27      importance for the assessment of human health effects due to ambient O3  exposure. Also, the
28      influence of O3-induced changes at cellular and molecular levels are integrated to elucidate
29      scientific bases for the observed physiological and pathological alterations.  These research
30      reports will be evaluated to assess (1) the scientific merit pertaining to the biological plausibility
31      of the health outcome associations observed in the epidemiological studies and (2) the coherence

        January 2005                              8-25        DRAFT-DO NOT QUOTE OR CITE

-------
 1      of the overall body of evidence relevant to O3-related health outcomes supporting conclusions
 2      regarding the attribution of observed effects of ambient O3 exposure.
 3           The 1996 O3 AQCD, based on the limited number of controlled human exposure studies
 4      and the animal toxicology data available to that date, arrived at the following conclusions
 5      regarding potential health effects of ambient O3 exposure:
 6        • Human studies have shown decreases in pulmonary function responsiveness to O3
            exposure as a function of increasing age, although symptom rates remain similar across
            age groups.
 7        • Toxicological studies are not easily interpreted, but tend to suggest that young animals are
            not more responsive to O3 than adults.
 8        • Available toxicological and human data have not conclusively demonstrated that males
            and females respond differently to O3. If gender differences exist for lung function
            responsiveness to O3, they are not based on differences in baseline pulmonary function.
 9        • Data are not adequate to determine whether any ethnic or racial group has a different
            distribution of responsiveness to O3. In particular, the responses of nonwhite asthmatics
            have not been investigated.
10        • Information derived from O3 exposure of smokers is limited.  The general trend is that
            smokers are less responsive than nonsmokers, but this reduced responsiveness may wane
            after cessation of smoking.
11        • Nutritional status (e.g., vitamin E deficiency) makes laboratory rats more susceptible to
            O3-induced effects, but it is not clear if vitamin E supplementation has an effect in human
            populations.  Such supplementation has no or minimal effects in animals.  The role of such
            antioxidant vitamins, especially their deficiency, in  O3 responsiveness has not been well
            studied.
12           As mentioned above, many questions remained unanswered,  and the 1996  O3 AQCD
13      neither identified clear biological bases nor provided convincing experimental evidence
14      supporting the biological plausibility of reported O3 effects and/or the mechanisms of action
15      underlying potential O3  toxicity.
16           The available new epidemiological research reports on the health effects of O3 and
17      controlled human exposure studies using novel and refined models indicate certain positive
18      O3-health effect associations.  This section focuses on interpreting the overall meaning of the
19      epidemiological findings and evaluates their bearing in the context of obtaining  evidence for the
20      biological plausibility and possible mechanism(s) of action.  This section also addresses the
21      complexities involved in extrapolating the extent of coherence observed from epidemiological

        January 2005                              8-26        DRAFT-DO NOT QUOTE OR CITE

-------
 1      studies to specific health outcomes and related toxicological and biochemical mechanisms for
 2      observed pathological and physiological changes in controlled human, animal, and in vitro
 3      exposure studies.  Experimental in vitro and in vivo studies using novel molecular technologies
 4      that predict new hypotheses regarding the mechanisms of action are also included and discussed
 5      as appropriate.
 6           Several criteria listed in Section 8.2.1 are used in evaluating the available scientific support
 7      for conclusions regarding potential causal relationships between O3 exposure and specific types
 8      of health outcomes.  In addition to those criteria addressed in the preceding discussion of
 9      epidemiological evidence, certain other critical evaluation measures must be considered to
10      ensure that these observations are biologically relevant and consistent with experimentally
11      demonstrated biological mechanisms of action. For this assessment, the ensuing discussion on
12      biological  plausibility and coherence considers (a) the extent to which available epidemiological
13      evidence shows associations with a range of logically linked health endpoints and (b) whether
14      available toxicological and biochemical evidence provides support for the observed
15      epidemiological associations reflecting causal relationships.
16
17      8.4.8.1  Background on  Cross-Cutting Issues
18           Discussion of several cross-cutting issues that will facilitate a clear understanding of the
19      ensuing assessment is provided here to enhance an integrated and comprehensive understanding
20      of the experimental and epidemiological studies on O3 health effects.  An important issue to
21      be considered is the extrapolation of observed effects from the perspective of dosimetry and
22      animal-to-human extrapolation models used and their strength. The most challenging issue is
23      the interpretation of the epidemiological observations in light of physiological and toxicological
24      endpoints derived from the experimental studies, wherein the data derived for O3-specific effects
25      have to be used to interpret and evaluate possible causative roles for O3 in contributing to health
26      outcomes observed in the air pollution epidemiology studies.
27
28      8.4.8.2  Approaches to Experimental Evaluation of Ozone Health Effects
29           Three chapters in the current document provide detailed discussion of various experimental
30      approaches utilized to evaluate O3-related health effects. Chapter 4 discusses dosimetry issues in
31      both animal and human exposure scenarios.  Chapter 5 discusses the experimental studies of

        January 2005                              8-27        DRAFT-DO NOT QUOTE OR CITE

-------
 1      physiological, biochemical (cellular and molecular changes) and pathological observations in
 2      laboratory animals (including nonhuman primates, dogs, and rodent species) and in vitro studies
 3      using cell culture systems (in certain cases, on humans cells recovered from BALF postexposure
 4      to O3).  Chapter 6 evaluated controlled human experimental studies that investigated various
 5      physiological and biochemical endpoints. Many of the experimental animal toxicology studies
 6      have been carried out using relatively high O3 exposure concentrations/doses that do not reflect
 7      "real-world" exposure scenarios.  These approaches have been used mainly to test hypotheses to
 8      understand potential mechanism(s) of action implicated with the health outcomes identified in
 9      epidemiological studies.  In interpreting the results from the experimental approaches, one must
10      consider the following three issues:  (1) controlled animal exposures studies use high
11      concentration to elicit biochemical/ physiological changes in healthy animals; (2) the roles of
12      other confounding pollutants that commonly occur with ambient exposures cannot be fully
13      reflected in the controlled exposure studies, and (3) the differences between human  and rodents
14      with regard to O3 inhalability, deposition, clearance, and retention profiles (see Chapter 4 and 5
15      for details). Note that most of the in vitro toxicological studies were aimed at hypothesis
16      generation to predict mechanism(s) of action based on cellular and molecular endpoints and,
17      therefore, also generally used high O3 concentrations. The following discussion attempts to
18      integrate the experimental and epidemiological evidence to develop  a holistic understanding of
19      O3 health effects in keeping with the points discussed above.
20
21      8.4.8.3 Interspecies Comparison of Experimental  Effects — Dosimetry Considerations
22           As discussed in the previous section, the most important factor to consider while
23      attempting to integrate experimental studies across the species is the exposure-dose relationship.
24      Animal studies, particularly rats, have been valuable in developing mathematical dosimetry
25      models and useful for dosimetric extrapolation to humans in the strict sense of dose-response
26      basis.  For example, dose-dependent increases in breathing frequency and decrease in tidal
27      volume were observed in both animals and humans with little effect on uptake.  The O3-uptake
28      efficiency data from human and animals were consistent without much bearing on the mode
29      of breathing.
30
        January 2005                              8-28        DRAFT-DO NOT QUOTE OR CITE

-------
 1           Experimental O3 dosimetry studies include response to bolus dose and general uptake.
 2      Although the former approach is of limited relevance to environmental exposures, it has
 3      indicated that prior exposure to O3 limits uptake of bolus dose.  New uptake studies (Ultman
 4      et al., 2004) carried out in controlled human clinical studies observed gender-specific differences
 5      in the uptake of O3, but these differences do not correlate well with spirometric responses.  Bolus
 6      dose studies demonstrated that the uptake and regional respiratory tract distribution of O3 is
 7      sensitive to the mode of breathing (nasal or oral) and to air flow rate. The change in breathing
 8      due to exercise can cause a shift in distribution, allowing deeper penetration with resultant
 9      damage to bronchiolar and alveolar tissue. This observation of an inverse relationship between
10      uptake and air flow is in agreement with animal studies.  Additional uptake studies carried out in
11      humans using environmentally relevant O3 concentrations demonstrated the significance of
12      incorporating intersubject variability in dose-response relationship predictions and extrapolation.
13           The high degree of consistency observed in O3 uptake in animal and human experimental
14      exposure studies provided increased confidence in the use of theoretical dosimetry modeling
15      (see Chapter 4 for detailed discussion). The early models computed dose-response relationships
16      based on the assumption of O3 as the only active toxicant responsible for the observed
17      respiratory injury. Newer models have taken into consideration various factors such as age, as
18      well as anatomical, physiological, and biochemical alterations.  The identification of conducting
19      airways as the primary site of acute cell injury, the site of O3 reaction/diffusion in the epithelial
20      lining fluid, the roles of intermediate reactive oxygen species (ROS)  and lipid-ozonation
21      products in oxidative injury, and the roles of metabolic enzyme profiles in developing lung
22      tissue, when incorporated, will lead to refined novel models. The PBPK models were developed
23      using some of the refinements listed above and indicated the difference in dose metrics between
24      adult and infant with no difference projected after the age of five.
25
26      8.4.8.4  Integrated Critical Analysis  of Physiological, Biochemical Effects
27           In the following subsections, research results generated from experimental studies on
28      humans and animals during the past decade are assessed (keeping in view the interspecies
29      differences discussed in the preceding section) in evaluating experimental evidence for
30      epidemiological studies to lead to the discussion of the biological plausibility and coherence for
31      O3 health effects in the later sections.

        January 2005                              8-29        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.8.4.1  Pulmonary Function
 2           The 1996 O3 AQCD reported decreases in FVC, FEVl3 decreased tidal volume, increased
 3      breathing frequency, and increased resistance on short-term exposure to O3, based on research
 4      reviewed from epidemiological, controlled human exposure, and animal studies. Inhalation of
 5      O3 for several hours while physically active elicits both subjective respiratory tract symptoms
 6      and acute pathophysiologic changes.  The typical symptomatic response consistently reported is
 7      that of tracheobronchial airway irritation. This is accompanied by decrements in lung capacities
 8      and volumes, bronchoconstriction, airway hyperresponsiveness, airway inflammation, immune
 9      system activation, and epithelial injury. The severity of symptoms and the magnitude of
10      response depend on inhaled dose, individual O3 sensitivity, and the extent of tolerance resulting
11      from previous exposures.
12           The development of effects is time-dependent during both exposure and recovery periods,
13      with considerable overlap of evolving and receding effects.  In healthy human subjects exposed
14      to typical ambient concentrations (i.e., < 0.2 ppm O3), spirometric responses largely resolve
15      within a few hours (4 to 6 h) postexposure, however cellular effects persist for longer periods
16      (-24 h).  Persisting small residual lung function effects are almost completely resolved within
17      24 hours. In hyperresponsive individuals, the recovery takes longer, as much as 48 h, to return
18      to baseline values.
19           It has also been observed that there is a large amount of intersubject variability and,
20      furthermore, the majority of these symptoms are attenuated after repeated exposure, but such
21      tolerance to O3 is lost within a week postexposure. Recent controlled human exposure studies
22      on prolonged O3 exposure using healthy subjects (reviewed in Chapter 6) found statistically
23      significant changes in pulmonary function during or after exposure to O3 concentrations
24      > 0.08 ppm.
25           New controlled human exposure studies (reviewed in Chapter 6) clearly indicate that FEVj
26      decrements and symptom responses decrease with age beyond young adulthood (18 to 20 years).
27      Hazucha et al. (2003) also examined gender differences along with age in O3 responsiveness and
28      observed that young females lose O3 sensitivity faster than young males,  but that the rate is about
29      the same for both genders by middle age.
30           Human studies consistently report that inhalation of O3 alters the breathing pattern without
31      significantly affecting minute ventilation. A progressive decrease in tidal volume and a

        January 2005                              8-30        DRAFT-DO NOT QUOTE OR CITE

-------
 1      "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
 2      exposure suggests a direct modulation of ventilatory control.  These changes parallel a response
 3      of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
 4      Earlier human studies (Coleridge et al., 1993; Hazucha and Sant'Ambrogio, 1993) reported a
 5      role for bronchial C-fibers and rapidly adapting receptors are the primary vagal afferents
 6      responsible for O3-induced changes in ventilatory rate and depth. A new study by Passannante
 7      et al. (1998) observed that the primary mechanism of O3-induced reduction in inspiratory lung
 8      function is an inhibition of inspiration elicited by stimulation of the C-fibers and suggest a role
 9      for nociceptive mechanisms in modulating O3-induced inhibition of inspiration.  This neurogenic
10      mechanism also has an effect on airway responsiveness and lung inflammation.
11           Lung function changes evaluated in patients with preexisting respiratory diseases, under
12      controlled  experimental exposure regimens with or without physical exertion in the form of
13      intermittent exercise, indicated minimal O3-induced effects in COPD patients. However, newer
14      studies (see Chapter 6) indicate that pulmonary function deficiencies detected by spirometric
15      analyses in asthmatics augment the observations made in the  1996 O3 AQCD. More specifically,
16      Gong et al. (1997a) exposed nine COPD patients (0.24 ppm O3 for 4 h with intermittent exercise)
17      and observed a nonsignificant FEVj decrement of - 8% in COPD patients, which was not
18      statistically different from the decrement of -3% in healthy subjects. In contract, studies of
19      between 4- and 8-h duration with O3 concentrations of < 0.2 ppm, suggest a tendency for
20      increased O3-induced pulmonary function responses in asthmatics relative to healthy subjects
21      (Scannell et al., 1996). Similarly, Alexis et  al. (2000) observed statistically significant
22      O3-induced decreases in FEVj in mild atopic asthmatics compared to healthy controls. Though
23      controlled  human exposure studies may not  provide the required statistical power (due to the
24      limited number of subjects compared to panel or field studies), they do suggest that asthmatics
25      are at least, if not more, sensitive than healthy subjects.
26
27      8.4.8.4.2  Airway Responsiveness
28           Increased responsiveness of the pulmonary airways, or "airway hyperresponsiveness"
29      (AHR), is usually analyzed in response to a  bronchoconstrictor challenge. An extensive animal
30      studies database (using rats, mice, guinea pigs, and rabbits) exploring the effects of acute, long-
31      term, and repeated exposures  to O3, indicates that induction of AHR occurs at high O3

        January 2005                               8-31         DRAFT-DO NOT QUOTE OR CITE

-------
 1      concentrations. These studies provide clues to the roles of physiological and biochemical
 2      components involved in this process. Experimental human exposure studies also reported that
 3      acute O3-induced AHR,  independent of pulmonary function changes or lung inflammation,
 4      resolves to normalcy within 24 h.  Airway hyperresponsiveness, as well as O3-induced airway-
 5      antigen reactivity, were  observed in asthmatics in several controlled exposure studies and found
 6      to persist several hours postexposure (Chapter 6). Gong et al. (1997b) found that subjects with
 7      asthma developed tolerance to repeated O3 exposures in a manner similar to normal subjects;
 8      however, subjects with asthma had more persistent effects of O3 on airway responsiveness,
 9      which was only partially attenuated when compared to filtered-air (FA) control subjects. These
10      observations suggest that O3 may act as a cofactor in response to airborne allergens or other
11      bronchoconstrictor agents in people with allergic asthma. Ozone-mediated modulation of airway
12      responsiveness may be a plausible link between ambient O3-exposure- related increased use of
13      asthma medication and the increased hospital admissions and emergency department visits
14      observed in epidemiological studies. Biochemical alterations observed in humans and animals
15      with exposure to O3 and discussed in the following sections may provide additional insights into
16      their roles in mechanistic aspects underlying the observed AHR.
17
18      8.2.8.4.3 Morphological and Biochemical Abnormalities
19           Most of the research results alluded to the ensuing discussion come from toxicology
20      studies using various laboratory animal species that were usually exposed to higher, non-ambient
21      concentrations of O3.  However, these  exploratory and mechanistic studies may provide
22      important and useful hypotheses to consider in integrating various health outcomes observed or
23      predicted by epidemiological studies.  Controlled human exposure studies evaluated a few
24      cellular and biochemical parameters, mostly from BALF analyses.  These studies have yielded
25      some limited evidence supporting the observations made in animal toxicology  studies. Again,
26      caution should be exercised in extrapolating these observations to humans, due to species-
27      specific differences, as outlined earlier (see Section 8.2.7.3).
28
29      Lung Injury and Morphological Changes
30           Most animal species tested, including primates, exhibit similar morphological alterations
31      dependent on the exposure dose and the regional specificity. Differences in the distribution of

        January 2005                              8-32        DRAFT-DO NOT  QUOTE OR CITE

-------
 1      antioxidants in the centriacinar region (CAR) of the lung were responsible for the differences in
 2      injury and morphological changes observed between nonhuman primates and rodents. Cells in
 3      the CAR are the primary targets of O3, but ciliated cells in the nasal cavity, airways, and Type I
 4      epithelial cells in the gas-exchange region are also targets. Though acute O3 exposure induces
 5      structural changes such as fibrosis in CAR, these structural alterations appear to be transient with
 6      recovery shortly postexposure, but the time for recovery is dependent on species and the dose of
 7      O3.  The remodeling of lung tissue indicated in a simulated seasonal-exposure scenario in
 8      primates suggests the development of possible stable structural alterations as compared to
 9      continuous-exposure scenarios.  In an autopsy pathologic study, a significantly greater extent
10      and severity of centriacinar region alterations were observed in lungs of Los Angeles residents
11      than Miami residents, independent of a smoking effect (Sherwin et al., 2000). The results
12      suggest that the greater extent and severity of centriacinar region alterations may be related to
13      the higher O3 levels in Los Angeles.  Similar observations of CAR thickening and deposition of
14      collagen in the rat suggests that long-term O3 exposure may cause a progressive structural lung
15      injury that can evolve into a more chronic form, such as fibrosis. Ozone-induced mucous
16      membrane cell metaplasia observed in rodents appears to be  mediated by inflammation.  Again,
17      one must be cautious in extrapolating these observations in animals to humans, given the
18      exposure regimens and doses used.
19
20      Lung Inflammation and Permeability
21           Ozone has long been recognized to cause lung inflammation and increased permeability in
22      the rat lung. These distinct, independent biological events have been observed in all species
23      studied,  including humans, in response to acute exposure to O3.  Increased lung inflammation
24      and permeability have been observed at levels as low as 0.12 ppm exposure for 6 h in rats and
25      24 h in mice.  Both the inflammatory response and increased lung permeability have been
26      observed as early as 1 h and found to persist for at least  18 h in humans  on exposure to O3 at
27      0.2 to 0.6 ppm.  Subchronic exposures in  animals suggest that permeability changes are transient
28      (and species-dependent) and return to control levels even with continuing exposure. Repeated
29      exposures in humans also indicate ongoing cellular damage irrespective of attenuation of the
30      inflammatory  responses and lung function. Several studies have analyzed bronchioalveolar
        January 2005                              8-33        DRAFT-DO NOT QUOTE OR CITE

-------
 1      lavage (BAL) and nasal lavage (NL) fluid and cells from O3-exposed humans for markers of
 2      inflammation and lung damage (see Tables AX6-12 and AX6-13 in the Annex to Chapter 6).
 3           The presence of neutrophils (PMNs) in the lung has long been accepted as a hallmark of
 4      inflammation and is an important indicator that O3 causes inflammation in the lungs. It is
 5      apparent, however, that inflammation within airway tissues may persist beyond the point that
 6      inflammatory cells are found in BAL fluid.  Soluble mediators of inflammation such as the
 7      cytokines IL-6 and IL-8 as well as arachidonic acid metabolites (e.g., PGE2, PGF2a,
 8      thromboxane, and leukotrienes [LTs] such as LTB4) have been measured in the BAL fluid of
 9      humans exposed to O3. In addition to their role in inflammation, many of these compounds have
10      bronchoconstrictive properties and may be involved in increased airway responsiveness
11      following O3 exposure. Inflammation and cellular responses associated with acute O3 exposure
12      were also attenuated after 5 consecutive days of O3 exposure (compared to historical data for
13      responses after a single-day exposure).  Even though indicators of epithelial cell damage were
14      not seen immediately after acute exposure, they were present in BALF after the fifth day of
15      exposure. When reexposed 2 weeks later, changes in BALF indicated that epithelial cells
16      appeared to be fully repaired (Devlin et al., 1997).  Similar adaptive response was also observed
17      in an epidemiological  study by Kopp et al. (1999).  The analysis of BALF in human subjects
18      after first O3 peak in summer indicated increased levels of protein and leukocytes and no such
19      increase was observed later in summer even after exposure to higher levels of O3.
20           Interaction of O3 with the constituents of the extracellular lining fluid and the induction of
21      oxidative stress is implicated in injury and inflammation.  Animal toxicological and a few
22      in vitro studies analyzed cells recovered from BALF for many biochemical mediators implicated
23      in injury and inflammation and found alterations in the expression of cytokines, chemokines,  and
24      adhesion molecules, indicative of an active stress response as well as injury repair and
25      regeneration processes. Both animal and human studies indicate cellular and biochemical
26      changes associated with inflammation and increased permeability, but the relationship between
27      these changes and their role in lung function and airway responses is not known.
28
29      Host Defense
30           A number of closely integrated defense mechanisms exist that offer protection to
31      respiratory tract cells from the adverse effects of inhaled pollutants and microbes.  Acute O3

        January 2005                              8-34        DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposure had been found to impair host defense capabilities both in humans and animals by
 2      depressing alveolar macrophage functions and by decreasing the mucocilliary clearance of
 3      inhaled particles and microbes.  Interference by O3 exposure with the clearance process has been
 4      found to be dose-dependent, whereby low doses accelerate clearance, but high doses slow
 5      clearance. Some respiratory tract regional- and species-specific differences have also been
 6      observed. Though acute O3 exposures were found to suppress alveolar phagocytosis and
 7      immune functions, these alterations appeared to be transient and were attenuated with continuous
 8      or repeated  exposures. Continuous exposures to O3 impairs immune responses, followed by
 9      adaptation and recovery to normalcy.  Studies using various genetically sensitive or susceptible
10      strains indicated a possible interaction between innate and acquired immune system components,
11      possibly through Toll-like receptor-4 and downstream pathways; but these studies were carried
12      out using high O3 doses that may not be relevant to  ambient exposures.
13
14      Biochemical A Iterations
15           An extensive experimental database, including the research presented in 1996 O3 AQCD,
16      suggests that potential biochemical alterations in various intermediary metabolic pathways are
17      involved in lung injury, inflammation, and functional alterations. Recent experimental evidence
18      still points to the importance of initial interactions of O3 with the lipid constituents of the ELF
19      and to the generation of ozonation products and secondary redox mediators in the initiation of
20      site-specific cell-injury response cascades. One such ozonation product, 4-hydroxynonenal, has
21      been found  to bind to proteins and increase protein adducts in human alveolar macrophages,
22      suggesting a possible role in acute cell toxicity. Species- and region-specific increases in lung
23      xenobiotic metabolism has been observed in response to both short- and long-term O3 exposure.
24      Antioxidants in the ELF react with O3  and confer protection from toxicity, and even with
25      environmentally relevant exposures, the reactivity of O3 was not quenched.  Species-specific and
26      age-dependent changes in the antioxidant metabolism add another dimension to their role in this
27      process. Carefully controlled studies of dietary antioxidant supplementation (Samet et al., 2001;
28      Trenga et al., 2001) have found some protective effects of a-tocopherol and ascorbate for
29      O3-induced spirometric lung function decrements but not for the intensity of subjective
30      symptoms and inflammatory responses (including cell recruitment, activation, and a release of
31      mediators).  Dietary antioxidants have also afforded partial protection to asthmatics by

        January 2005                             8-35         DRAFT-DO NOT QUOTE OR CITE

-------
 1      attenuating postexposure bronchial hyperresponsiveness (Trenga et al., 2001).  Also, two
 2      epidemiologic studies of street workers and asthmatic children in Mexico City found that
 3      subjects taking antioxidant supplements containing vitamins E and C were protected from
 4      O3-induced changes in lung function (Romieu et al., 1998, 2002).
 5           Based on the above, it is evident that the extensive experimental database accumulated
 6      from animal toxicology studies (including nonhuman primate studies) and limited controlled
 7      human exposure studies, has provided insights into various biochemical, cellular, and molecular
 8      alterations in lung tissue exposed to O3.  The majority of these studies used acute exposure
 9      regimens and high concentrations, and provide hypotheses regarding potential  molecular
10      mechanisms implicated in O3 toxicity. Utilizing this information in relevant rodent-to-human
11      extrapolation  models with appropriate species-specific adjustments may well provide useful
12      information on initial biochemical alterations that may aid in the development  of suitable
13      biomarkers for O3 exposures/effects.
14
15      Systemic Effects
16           A number of rodent toxicology studies that investigated the effects of acute O3 exposure on
17      extrapulmonary systems have reported neurobehavioral, neuroendocrine, developmental, and
18      skin effects, albeit typically at much higher than ambient O3 concentrations. Ultrastructural,
19      cellular, and biochemical parameters evaluated in these studies indicate a role for O3 in
20      mediating biochemical and functional alterations through interactions with the redox systems.
21      An increasing body of animal toxicology evidence suggests that thermoregulatory and
22      hematological alterations (in heart rate variability and/or core body temperature) may mediate
23      acute cardiovascular effects.  Limited human exposure studies have also explored O3-induced
24      cardiovascular effects, but did not observe acute cardiovascular effects in normal and
25      hypertensive subjects.
26
27      Susceptibility Factors
28           Many factors such as age, gender,  disease, nutritional status,  smoking, and genetic
29      variability may contribute to the differential effects of environmental pollutants, including O3.
30      Genetic factors, such as single nucleotide polymorphisms (SNPs) and developmental defects,
31      can contribute to innate susceptibility, while acquired susceptibility may develop due to personal

        January 2005                               8-36        DRAFT-DO NOT  QUOTE OR CITE

-------
 1      habits (smoking, diet, exercise) and other risk factors such as age, gender, pregnancy, and
 2      copollutants. However, the available information from animal toxicologic and epidemiologic
 3      studies does provide clear scientific evidence by which to identify and/or associate any specific
 4      factor as contributing to adverse health effects of O3 (U.S. Environmental Protection Agency,
 5      1996).
 6           New animal toxicology studies using various strains of mice and rat have identified O3-
 7      sensitive and resistant strains and illustrate the importance of genetic background in determining
 8      O3 susceptibility. Biochemical and molecular parameters extensively evaluated in these
 9      experiments were used to identify specific loci on the chromosomes and, in some cases, to relate
10      the differential expression of specific genes to biochemical and physiological differences
11      observed among these species.  Utilizing O3-sensitive and O3-resistant species, it has been
12      possible to identify the involvement of AHR and inflammation processes in O3 susceptibility.
13      However, most of these studies were carried out using high doses of O3, making the relevance of
14      these studies questionable in human health effects assessment. No doubt, the molecular
15      parameters identified in these studies may serve as useful biomarkers with the availability of
16      suitable technologies and, ultimately, can likely be integrated with epidemiological studies.
17      Interindividual differences in O3 responsiveness have been observed across a spectrum of
18      symptoms and lung function responses but do no yet allow identification of important underlying
19      factors, except a significant role for age.
20
21      8.4.9  Preexisting Disease as a Potential Risk Factor
22           People with preexisting pulmonary disease may be at increased risk from O3 exposure.
23      Altered physiological, morphological and biochemical states typical of respiratory diseases like
24      asthma, COPD and chronic bronchitis may render people sensitive to additional oxidative burden
25      induced by O3 exposure. Based on studies assessed in the 1996 criteria document (U.S.
26      Environmental Protection Agency, 1996), asthmatics appear to be at least as, or more, sensitive
27      to acute effects of O3 as healthy nonasthmatic subjects. The new results reviewed in Chapters 6
28      and 7 from controlled exposure and epidemiological studies also suggest that asthmatics are a
29      potentially sensitive subpopulation for O3 health effects.
30           A number of time-series epidemiological studies have reported increased risk in study
31      subsets of individuals with preexisting lung diseases and tend to implicate asthmatics as

        January 2005                               8-37         DRAFT-DO NOT QUOTE OR CITE

-------
 1      potentially susceptible individuals. The epidemiological studies of acute exposure to O3
 2      discussed in Section 8.2.3 indicate increased risk for exacerbation of disease symptoms during
 3      the warm season.
 4           Newly available human exposure studies by Stenfors and coworkers have shown
 5      differences regarding PMN influx in BALF between asthmatics and healthy human subjects.
 6      In vitro studies (Stenfors et al., 2002) using nasal mucosal biopsies from atopic and nonatopic
 7      subjects exposed to 0.1 ppm O3 found significant difference in the induced release of IL-4, IL-6,
 8      IL-8, and TNF-a.  A subsequent study by the same group (Schierhorn et al., 2002) found a
 9      significant difference in the O3-induced release of the neuropeptides neurokinin A and substance
10      P from allergic patients, compared to nonallergic controls, suggesting increased activation of
11      sensory nerves by O3 in the allergic tissues. Another report from Bayram et al. (2002) using in
12      vitro culture of bronchial epithelial cells recovered from atopic and nonatopic asthmatics
13      indicated the existence of a significant difference in permeability by measuring the paracellular
14      flux of 14C-BSA. Additional controlled O3 exposure studies in human subjects with intermittent
15      asthma (Hiltermann et al., 1999), and asthmatics (Basha et al., 1994; Scannell et al.,  1996)
16      reported increased secretion of IL-8 suggesting increased neutrophilic inflammation in those
17      subjects.
18           The observation of increased pathological  symptoms in long-term animal exposure studies
19      in the absence of observable physiological changes also suggests that chronic exposure may
20      increase susceptibility to adverse health effects,  but this needs to be validated via long-term
21      epidemiological studies.
22
23      8.4.10  Biological Plausibility and Coherence of Evidence for Adverse
24              Respiratory Health Effects
25           The research evidence discussed in the preceding section indicates that injury to lung tissue
26      is the initial step in mediating deleterious health effects  of O3, and, in turn, activates a cascade of
27      events starting with inflammation, altered permeability of the epithelial barrier, impaired
28      clearance mechanisms (including host defense),  and pulmonary structural alterations that can
29      potentially exacerbate a preexisting disease status. Although many or all of the above proposed
30      mechanisms are hypothesized to be implicated in O3 toxicity, scientific evidence is still lacking
31      for clearly establishing a role for one or a group of mechanistic pathways underlying O3 health

        January 2005                               8-38        DRAFT-DO NOT QUOTE OR CITE

-------
 1      effects observed in epidemiologic studies. Most of these mechanisms of action were predicted
 2      based on animal toxicology studies, with some support from human exposure studies.
 3           In this section, the new scientific information reviewed on animal toxicology studies
 4      (Chapter 5) and human exposure studies (Chapter 6) are used to evaluate plausible biological
 5      bases for the health effects observed in epidemiological (Chapter 7) studies.  The interpretations
 6      provided in this section are, in the majority of situations, based on theoretical extrapolations,
 7      while being mindful (based on our understanding in the post-genome era) of the existence of
 8      common biochemical and molecular pathways that operate or function across different species.
 9      In order to help interpret health endpoints (hospital admissions, mortality, and disease
10      exacerbations) purported to be associated epidemiologically with either acute or long-term
11      ambient exposure to O3, this section is organized into two subsections, based on the
12      physiological observations presented in the first section of (a) O3 effects on pulmonary function
13      as supported by cellular and molecular biological observations discussed in the second section,
14      and (b) O3-related lung injury, inflammation and host defense effects.
15           As exposure to O3 progresses, lung injury and inflammation begin to develop and initiate
16      cellular and subcellular changes. Airway hyperreactivity develops more slowly than pulmonary
17      function effects, while inflammation develops even more slowly and reaches its maximum 3 to
18      6 h postexposure. Cellular responses, such as release of inflammatory mediators or cytokines,
19      appear to be active as late as 20-h postexposure (Torres et al., 2000). Although the following
20      discussion is divided into two subsection, there may be cross-references between the sections to
21      better establish meaningful biological plausibility, as physiological and biochemical changes
22      overlap. Each subsection summarizes pertinent key information and then arrives at conclusions
23      as to the plausibility of effects attributable to O3 exposure.
24
25      8.4.10.1 Pulmonary Function
26           Ozone-induced critical respiratory functional deficiencies were monitored by measuring
27      changes in pulmonary function.  Studies detailing  alterations in pulmonary function discussed in
28      Chapters 5, 6, and 7 from animal, human toxicology, and epidemiology studies are summarized
29      in Sections 8.2.2, 8.2.3, and 8.2.7.4. Evaluation of pulmonary function on acute O3 exposure in
30      animals show a positive association with increased breathing frequency, decreased tidal volume
31      (rapid and  shallow breathing), increased resistance, and altered breathing mechanics (compliance

        January 2005                              8-39        DRAFT-DO NOT QUOTE OR CITE

-------
 1      and resistance).  A similar increased breathing frequency observed in human subjects suggests
 2      modulation of ventilatory control on acute O3 exposure. Direct or indirect stimulation of lung
 3      receptors and bronchial C-fibers by O3 and/or its oxidative products have been implicated in
 4      modulating breathing pattern changes. Acute O3-induced biochemical changes suggest potential
 5      interactions of O3 with the extracellular lining fluid and the generation of lipid ozonation
 6      products and reactive oxygen species, ultimately leading to lung injury and/or inflammation.
 7      These reactive species cause inhibition or decrease in the maximal inspiration capacity by
 8      neurogenic mechanisms acting via the C-fiber afferents.  Recent work by Passannante et al.
 9      (1998) implicates stimulation of nociceptive receptors on bronchial C-fibers as the primary
10      mechanism for O3-induced inhibition of inspiration.  Alternately, inhibition in maximal
11      inspiration may also be mediated by mediators such as prostaglandin E2 released due to lung
12      epithelial injury.  This hypothesis gains strength from the observation of the blocking of
13      spirometric response in human subjects who were pretreated with nonsteroidal anti-
14      inflammatory agents such as indomethacin and ibuprofen. While recovery from pulmonary
15      function decline and airway hyperreactivity had been found to be rapid (4 to 6 h) in moderately
16      responsive individuals, persistent small residual lung function effects were found to take more
17      than 48 h to return to baseline values in hyperresponsive individuals (Nightingale et al., 2000).
18      Such an extended recovery from lung function decline, airway hyperresponsiveness, and
19      increased O3-induced pulmonary function decline in  asthmatics (Scannell et al., 1996) compared
20      to normal subjects may be responsible for the increased emergency room visits or hospital
21      admission and the increased use of asthma medication in asthmatics reported in recent time-
22      series epidemiological studies.  The contribution of morphological alterations in the decline of
23      lung function on chronic exposure is not known.
24           Airway hyperresponsiveness (AHR) due to O3  exposure is another important factor
25      involved in the observed decline in pulmonary function.  Intermittent airway obstruction and
26      increased airway responsiveness to physical or chemical  stimuli is also the hallmark of asthma.
27      Asthma-related AHR includes both physiological and morphological components such as
28      inner-wall thickening and mucus secretion.  Airway hyperresponsiveness in response to
29      chemical  challenge is found to be predominant in children compared to adults and older children.
30      Several controlled human O3 exposure studies reported increased airway responsiveness at
31      baseline both in normal and asthmatic subjects. The  mechanisms mediating AHR are not yet

        January 2005                             8-40        DRAFT-DO NOT QUOTE OR CITE

-------
 1      fully understood, but it appears to be mediated by multiple pathways. Involvement of AHR in
 2      pulmonary function decline appears to be mediated by neurogenic mechanisms, as pretreatment
 3      of neonatal rats with capsaicin prevented O3-induced release of neuropetides, suggesting a role
 4      for C-fibers in AHR.  Significant reduction in the immunoreactivity for substance P in the
 5      bronchial biopsies from human subjects 6 h postexposure and its negative correlation with FEVj
 6      decrements suggests involvement of similar neurogenic mechanisms to be persistent in
 7      O3-induced bronchoconstriction (Krishna et al., 1997). Studies carried out in human subjects
 8      using cyclooxygenase inhibitors to block the influx of PMNs and the inhibition of neutrophilic
 9      inflammation by probenecid in dogs (Freed et al., 1999) indicated that O3-induced inflammation
10      and AFIR occur as two independent events. Many of the inflammatory  mediators that exhibit
11      bronchoconstrictive properties may also play a significant role in the persistent spirometry
12      changes observed on exposure to O3 (Blomberg et al., 1999).  Either the inflammation or AFIR
13      may be the underlying biological mechanism responsible for the  observed lung function decline
14      in children and male human subjects reported in epidemiological studies.
15           Repeated-exposure studies in monkeys with a house dust mite antigen-sensitization
16      regimen (Schelegle et al., 2003; Chen et al., 2003) associated lung function changes to the
17      adaptation of respiratory motor responses. Similarly, controlled human exposure studies using
18      asthmatic subjects exposed to house dust mite indicated an immediate increase in airway-antigen
19      reactivity that persisted longer (18 to 20 h) in asthmatics than in normal subjects ( Kehrl et al.,
20      1999).  The enhanced decline in pulmonary function in subjects with allergic rhinitis (Torres
21      et al., 1996) suggests slow recovery from O3-induced pulmonary function declines.  These
22      observations suggest that O3 exposure, therefore, may be a clinically important cofactor in the
23      response to airborne bronchoconstrictor substances in individuals with preexisting allergic
24      asthma.  This phenomenon could plausibly contribute to increased symptom exacerbations and,
25      even increased consequent physician or ER visits and possible hospital  admissions. However,
26      even a small decline in lung function and its persistence in sensitive populations such as
27      asthmatics will have substantial effects and can lead to increased frequencies in emergency room
28      hospital visits or in medication use.
29
        January 2005                              8-41         DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.10.2  Lung Injury, Inflammation, and Host Defense
 2           An extensive biochemical database collected from animal toxicology studies using varied
 3      species indicates that the interaction of O3 with the ELF in the lung leads to the generation of a
 4      series of reactive radical species, including lipid ozonation products and hydroperoxy radicals,
 5      which contribute to the increased injury and inflammatory response. Using in vitro and ex vivo
 6      studies on isolated ELF, it has been increasingly recognized that ozonation of polyunsaturated
 7      fatty acids (PUFA)  and its distribution profiles in the respiratory tract plays a critical role in the
 8      extent of site-specific injury (Postlewait et.al., 1998; Connor et al., 2004).  Ozone-induced lung
 9      injury was also found to cause persistent fibrotic changes with the accumulation of collagen and
10      the thickening of CAR postexposure, suggesting progressive structural changes in the lung
11      tissue.  Bronchial mucosal biopsies after repeated O3 exposure over 5 days in human subjects
12      indicated that inflammation of the bronchial mucosa persisted after repeated O3 exposure,
13      despite attenuation of some inflammatory markers in BALF and attenuation of lung function
14      responses and symptoms.  Along with this, the persistent, although small, decrease in baseline
15      FEVj observed by Torres et al. (2000) suggested a difference in timescales among the functional
16      responses to O3. Elevated protein levels remaining after repeated exposures confirm the findings
17      of others (Christian et al., 1998; Devlin et al.,  1997) and suggest that cellular damage is ongoing
18      irrespective of the attenuation of cellular inflammatory responses.  The results of chronic-
19      exposure studies in animal toxicology evaluating fibrotic changes are inconsistent. Simulated
20      seasonal-exposure studies in infant rhesus monkey (0.5 ppm O3) indicate possible injury-repair
21      processes as observed with chronic exposure studies described in the  1996 O3 AQCD (U.S.
22      Environmental Protection Agency, 1996).
23           Lung inflammation is a host response to  injury, which in turn triggers various biochemical
24      and physiological responses such as epithelial  permeability, PMN influx, and the release of
25      pro- and anti-inflammatory mediators. The inflammatory response is a transient phenomenon
26      and resolves entirely postexposure in all the species studied, including humans. Subchronic
27      exposure to O3 has been found to induce  inflammation after a few days (depending upon species
28      studied and  exposure dose), which resolves to  a normal state, even with continuing exposure;
29      adaptation on repetitive  exposures has also been observed.  Biochemical  and molecular analysis
30      of BALF from various animal species and human subjects exposed to O3 clearly suggest the
31      participation of various inflammatory mediators, which in turn can activate multiple cascades of

        January 2005                               8-42         DRAFT-DO NOT QUOTE OR CITE

-------
 1      physiological events. The adaptive response to repetitive exposure to O3 indicates resolution of
 2      many of these inflammatory markers to different extents, suggesting that persistent mild
 3      inflammation may compromise the abilities of the lung tissue to handle various host defense
 4      functions and allergic responses.
 5           Ozone-induced dysfunction in the barrier function of lung epithelium and subsequent
 6      permeability changes also can impact lung host defense functions.  Altered mucocilliary and
 7      alveolar clearance of inhaled particles or microorganisms observed on acute and subchronic
 8      exposure to O3 in animal toxicology studies indicate compromise in this important host defense
 9      function. Similar compensation observed in the in vitro studies using cells recovered in BALF
10      from normal and asthmatic human  subjects (Newson et al., 2000; Bosson et al., 2003; Bayram
11      et al., 2002) on acute O3 exposure suggest an additional burden on the host defense functions.
12           The new information obtained in the past decade on the morphological, biochemical,
13      cellular, and molecular aspects of O3 toxicology have increased our understanding of the
14      intricate biochemical and molecular mechanisms involved in lung tissue pathology. Combining
15      basic toxicology approaches with the sophisticated molecular technologies and using various O3-
16      sensitive and -resistant animal strains in these investigations have provided additional knowledge
17      in understanding the possible biochemical bases for the  adverse health effects. Newer studies
18      that examined various inflammatory parameters, such as PMN influx (Stenfors et al., 2002) and
19      molecular changes in the nasal or bronchial biopsies from atopic asthmatics and normal human
20      subjects, indicated significant differences.  Nichols et al., (2001) observed a role for oxidative
21      stress in O3-induced inflammation and increased release of TNF-a from nasal epithelial cells.
22      Increased neurogenic involvement in the O3-induced inflammatory response was observed by
23      Schierhorn et al. (1999).  Taken together, the new science gathered from an extensive animal
24      toxicology database and limited human controlled exposure  studies on pulmonary function, lung
25      defense, and biochemistry provide  evidence consistent and coherent with health outcomes
26      endpoints observed in human subjects in clinical or field studies. These biological and
27      toxicological observations will gain additional value and support with future research efforts,
28      particularly those using controlled human exposure studies including subjects with preexisting
29      disease. Such research inquiries will aid in reducing the data gaps and in the development of
30      better extrapolation models that can be used in interpreting the health endpoints monitored in
31      epidemiological studies.

        January 2005                              8-43        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.11  Coherence Between Epidemiological and Experimental Evidence for
 2              Respiratory Health Effects
 3           Recent epidemiological studies (collected from various metropolitan cities in the United
 4      States, Canada, and Europe) reported associations between short-term O3 exposure for various
 5      indices such as respiratory-related mortality, hospital admissions, and emergency department
 6      visits for respiratory diseases. Three U.S. multicity studies of 95 communities and 90 cities
 7      reported positive associations between acute O3 exposure and daily mortality. Data from four
 8      European multicity studies have also reported similar positive associations to daily mortality on
 9      acute O3 exposure.  Several epidemiological individual and multicity studies reported significant
10      O3 effects on hospital admission. Some of the epidemiological studies from Europe reported a
11      strong relationship between unscheduled hospitalizations for COPD and O3 exposure.
12      In addition, new evidence exists for O3 effects on lung function decline/respiratory symptoms
13      from controlled human exposure (exercising) studies and recent exercise panel (field) studies.
14      Recent time-series studies reported excess risk for emergency department visits, particularly in
15      summer seasons with relatively high O3 concentrations.  Epidemiological studies also reported
16      positive associations between the onset of asthma and asthma prevalence due to long-term O3
17      exposure.
18           The respiratory health effects observed in epidemiological studies gain relevance from
19      controlled human exposure studies.  The health responses observed in these studies were
20      indicative of deleterious health effects due to exposure to O3.  The studies indicate an acute
21      O3-induced decline in lung function with inflammatory changes in the lung such as increased
22      levels of infiltrating PMNs, release of inflammatory mediators, lung injury, permeability
23      changes, and altered host defense mechanisms.  These observations gain further support from
24      similar biochemical, morphological, and immunological changes in animal  toxicology studies
25      that suggest perturbations in lung tissue physiology.  This body of evidence provides coherent
26      links between the results of large multicity epidemiological studies reporting increases in
27      hospitalizations and unscheduled emergency department visits with toxicologic evidence of
28      acute O3-induced lung tissue injury and inflammation in humans and animals.
29
        January 2005                              8-44        DRAFT-DO NOT QUOTE OR CITE

-------
 1      8.4.12  Summary and Conclusions for Ozone Health Effects
 2           This section discussed the development of a coherent understanding of O3 health effects
 3      through considering the plausibility and coherence of information derived from epidemiological
 4      evidence and human and animal toxicology studies.  As discussed in Sections 8.4.2 to 8.4.5,
 5      considerations related to the epidemiologic evidence alone appear to support likely causality for
 6      observed associations between O3 and health outcomes. This section further described evidence
 7      from both  epidemiologic and toxicologic studies for health effects that are logically linked
 8      together.
 9           Epidemiological studies have reported positive associations between O3 exposure and
10      health effects across a range of endpoints, from respiratory-related mortality, increased
11      respiratory-related hospital admissions, and emergency department visits, to more subtle effects
12      such as decrements in lung function, lung inflammation, airway responsiveness, and altered
13      mucocilliary clearance.  The new toxicologic and physiological evidence suggest links to
14      potential molecular pathways that may provide reasonable explanations for the observed
15      epidemiological findings, but as described earlier, caveats must be considered in interpreting
16      these studies. The new toxicological information confirms the earlier findings and presents
17      important new evidence supporting the plausibility of associations between O3 and adverse
18      respiratory health effects. While many research questions remain, the convergence of
19      epidemiologic and toxicologic evidence related to respiratory health effects for ambient O3
20      exposure argues for coherence and plausibility for this body of evidence.
21           Controlled human exposure studies have provided new information indicating that age at
22      time of exposure is a major susceptibility factor for observed decrements in lung function.
23      Epidemiological studies and some preliminary supportive data from toxicological  studies
24      suggest that asthmatics are a potential sensitive subpopulation although additional scientific
25      evidence is needed. However, little experimental evidence is available by which to judge the
26      plausibility of any chronic O3 exposure effects observed in epidemiologic studies.  Thus, further
27      study is required on the potential toxicologic or pathologic mechanisms that may underly chronic
28      effects of ambient O3 exposure to relate to observed respiratory health effects in epidemiological
29      studies.
30           Analysis of the body of toxicologic studies suggests plausible  mechanisms for
31      epidemiologic findings. The newly available epidemiological studies on positive associations

        January 2005                              8-45         DRAFT-DO NOT QUOTE OR CITE

-------
 1      between acute exposure to O3 and a range of health outcomes support the general conclusion that
 2      O3 is causally related to respiratory-related mortality and morbidity. A very limited database
 3      (epidemiologic and toxicologic) is available on the long-term effects of O3 on respiratory-related
 4      mortality and morbidity, but given our understanding of the plausible biological mechanisms
 5      implicated in acute response and differential recovery, O3 may also be causally related to long-
 6      term respiratory-related health risks. Additional scientific information to support the predicted
 7      sensitivity of asthmatics reported in epidemiological studies is still needed.  Substantial scientific
 8      evidence gathered in the past decade provides additional support for the conclusions stated in the
 9      1996 O3 AQCD with regard to health effects shown to be associated with ambient O3 exposure.
10      The recent epidemiological studies provide further strong evidence supporting potential
11      morbidity health risks associated with exposure to ambient O3. Furthermore, newly available
12      epidemiological data linking acute O3 exposure to respiratory-related mortality and to the
13      exacerbation of respiratory-related disease symptoms suggest even larger health impacts and
14      costs to society than previously demonstrated, warranting additional research.
15
        January 2005                              8-46        DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Alexis, N.; Urch, B.; Tarlo, S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000) Cyclooxygenase
  3             metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
  4             normals. Inhalation Toxicol. 12: 1205-1224.
  5       Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, I, Jr. (1994) Bronchoalveolar lavage
  6             neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
  7             Chest 106: 1757-1765.
  8       Bayram, H.; Rusznak, C.; Khair, O. A.; Sapsford, R. J.; Abdelaziz, M. M. (2002) Effect of ozone and nitrogen
  9             dioxide on the permeability of bronchial epithelial cell cultures of non-asthmatic and asthmatic subjects.
10             Clin. Exp. Allergy 32: 1285-1292.
11       Bell, M. L.; McDermott, A.; Zeger, S. L.; Samet, J. M.; Dominici, F. (2004) Ozone and short-term mortality in
12             95 US urban communities, 1987-2000. JAMA J. Am. Med. Assoc. 292: 2372-2378.
13       Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S.  T.; Sandstrom,
14             T. (1999) Ozone-induced lung function decrements do not correlate with early airway inflammatory or
15             antioxidant responses. Eur. Respir. J. 13: 1418-1428.
16       Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate,  S. T.; Kelly, F. J.; Sandstrom, T.; Wilson, S.;
17             Frew, A. J.; Blomberg, A. (2003) Ozone-induced bronchial epithelial cytokine expression differs between
18             healthy and asthmatic subjects.  Clin. Exp. Allergy 33: 777-782.
19       Burnett, R. T.; Brook,  J. R.; Yung, W. T.; Dales, R. E.; Krewski, D. (1997) Association between ozone and
20             hospitalization for respiratory diseases in 16 Canadian cities. Environ. Res. 72: 24-31.
21       Burnett, R. T.; Cakmak, S.; Brook, J. R.; Krewski, D. (1997) The role of paniculate size and chemistry in the
22             association between summertime ambient air pollution and hospitalization for cardiorespiratory diseases.
23             Environ. Health Perspect. 105: 614-620.
24       Burnett, R. T.; Smith-Doiron, M.; Stieb, D.; Raizenne, M. E.; Brook, J. R.; Dales, R. E.; Leech, J. A.; Cakmak, S.;
25             Krewski, D. (2001) Association between ozone and hospitalization for acute respiratory diseases in children
26             less than 2 years of age. Am. J.  Epidemiol. 153: 444-452.
27       Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
28             contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone  exposure in infant
29             primates. J. Appl. Physiol.  94: 819-827.
30       Chen, L.; Jennison, B. L.; Yang, W.; Omaye, S. T. (2000) Elementary school absenteeism and air pollution.
31             Inhalation Toxicol. 12: 997-1016.
32       Chock, D. P.; Winkler, S. L.; Chen, C. (2000) A study of the association between daily mortality and ambient air
33             pollutant concentrations in Pittsburgh, Pennsylvania. J. Air Waste Manage. Assoc. 50:  1481-1500.
34       Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998)  Ozone-induced
35             inflammation is attenuated with multiday exposure. Am. J. Respir. Crit. Care Med. 158: 532-537.
36       Coleridge, J. C. G.; Coleridge, H. M.;  Schelegle, E. S.; Green, J. F. (1993) Acute inhalation of ozone stimulates
37             bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
38       Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
39             reactive absorption-mediated cytotoxicity in vitro. Am. J. Physiol. 286: L1169-L1178.
40       Devlin, R. B.; Folinsbee, L. J.; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997)
41             Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhalation Toxicol.
42             9:211-235.
43       Dockery, D. W.; Schwartz, J.; Spengler, J. D. (1992) Air pollution and daily  mortality: associations with particulates
44             and acid aerosols. Environ. Res. 59: 362-373.
45       Dominici, F.; McDermott, A.; Daniels, M.; Zeger,  S. L.; Samet, J. M. (2003) Mortality among residents of 90 cities.
46             In:  Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA: Health
47             Effects Institute; pp. 9-24. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [12 May, 2004].
48       Fairley, D. (2003) Mortality  and air pollution for Santa Clara County, California, 1989-1996.  In: Revised analyses of
49             time-series studies of air pollution and health. Special report. Boston, MA: Health Effects Institute; pp.
50             97-106. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf [18 October, 2004].
51       Freed, A. N.; Cueto, R.; Pryor, W. A. (1999) Antioxidant transport modulates peripheral airway reactivity and
52             inflammation during ozone exposure. J. Appl. Physiol. 87: 1595-1603.
53       Friedman, M. S.; Powell, K.  E.; Hutwagner, L.; Graham, L. M.; Teague, W. G. (2001) Impact of changes in
54             transportation and commuting behaviors during the 1996 summer Olympic games in Atlanta on air quality and
55             childhood asthma. JAMA J. Am. Med. Assoc. 285:  897-905.
         January 2005                                    8-47         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Gamble, J. L. (1998) Effects of ambient air pollution on daily mortality: a time series analysis of Dallas, Texas,
  2             1990-1994. Presented at: 91st annual meeting and exhibition of the Air & Waste Management Association;
  3             June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; paper no. 98-MP26.03.
  4       Gent, J. F.;  Triche, E. W.; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Leaderer, B. P. (2003)
  5             Association of low-level ozone and fine particles with respiratory symptoms in children with asthma. JAMA
  6             J. Am. Med. Assoc. 290: 1859-1867.
  7       Gilliland, F. D.; Berhane, K.; Rappaport, E. B.; Thomas, D. C.; Avol, E.; Gauderman, W. J.; London, S. J.;
  8             Margolis, H. G.; McConnell, R.; Islam, K. T.; Peters, J. M. (2001) The effects of ambient air pollution on
  9             school absenteeism due to respiratory illnesses. Epidemiology 12: 43-54.
10       Gong, H., Jr.; McManus, M. S.; Linn, W. S. (1997) Attenuated response to repeated daily ozone exposures in
11             asthmatic subjects. Arch. Environ. Health 52: 34-41.
12       Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997) Responses of older men with and without
13             chronic obstructive pulmonary disease to prolonged ozone exposure. Arch. Environ. Health 52: 18-25.
14       Gryparis, A.; Forsberg, B.; Katsouyanni, K.; Analitis, A.; Touloumi, G.; Schwartz, J.; Samoli, E.; Medina, S.;
15             Anderson, H. R.; Niciu, E. M.; Wichmann, H.-E.; Kriz, B.; Kosnik, M.; Skorkovsky, J.; Vonk, J. M.;
16             Dortbudak, Z. (2004) Acute effects of ozone on mortality from the "air pollution and health: a European
17             approach" project. Am.  J. Respir.  Crit. Care Med. 170:  1080-1087.
18       Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003) Distribution and reproducibility of spirometric response to
19             ozone by gender and age. J.  Appl. Physiol. 95: 1917-1925.
20       Hazucha, M. J.; Sant'Ambrogio, G. (1993) Effects of ozone on the activity of slowly (SAR) and rapidly adapting
21             (RAR) receptors in cats. FASEB J. 7: 407A.
22       Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.; Sterk, P. J.;
23             Hiemstra, P. S.; Stolk, J. (1999) Ozone-induced inflammation assessed in sputum and bronchial lavage fluid
24             from asthmatics: a new  noninvasive  tool in epidemiologic studies on air pollution and asthma. Free Radical
25             Biol. Med. 27: 1448-1454.
26       Ito, K. (2003) Associations of paniculate matter components with daily mortality and morbidity in Detroit,
27             Michigan. In: Revised analyses of time-series studies of air pollution and health. Special report. Boston, MA:
28             Health Effects Institute; pp.  143-156. Available: http://www.healtheffects.org/Pubs/TimeSeries.pdf
29             [12 May, 2004].
30       Ito, K.; Thurston, G. D. (1996) Daily PM10/mortality associations: an investigation of at-risk subpopulations.
31             J. Exposure Anal.  Environ. Epidemiol. 6: 79-95.
32       Jaffe, D. H.; Singer, M. E.; Rimm, A. A.  (2003) Air pollution and emergency department visits for asthma among
33             Ohio Medicaid recipients, 1991-1996. Environ. Res. 91: 21-28.
34       Jorres, R. A.; Holz, O.; Zachgo, W.; Timm, P.; Koschyk, S.; Mtiller, B.; Grimminger, F.; Seeger, W.; Kelly, F. J.;
35             Dunster,  C.; Frischer, T.; Lubec, G.; Waschewski, M.; Niendorf, A.; Magnussen, H. (2000) The effect of
3 6             repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
37             Am.  J. Respir. Crit. Care Med. 161:  1855-1861.
38       Jorres, R.; Nowak,D.; Magnussen, H.; Speckin, P.; Koschyk,  S. (1996) The effect of ozone exposure on allergen
39             responsiveness in  subjects with asthma or rhinitis. Am.  J. Respir. Crit. Care Med. 153: 56-64.
40       Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased specific airway reactivity
41             of persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J. Allergy. Clin.
42             Immunol. 104: 1198-1204.
43       Kinney, P. L.; Ito, K.; Thurston, G. D. (1995) A sensitivity analysis of mortality/PM10 associations in Los Angeles.
44             In: Phalen, R. F.; Bates, D. V., eds. Proceedings of the colloquium on  paniculate air pollution and human
45             mortality and morbidity; January 1994; Irvine, CA. Inhalation Toxicol. 7: 59-69.
46       Kinney, P. L.; Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County.
47             Environ.  Res. 54: 99-120.
48       Klemm, R.  J.; Lipfert, F. W.; Wyzga, R.  E.; Gust, C. (2004) Daily mortality and air pollution in Atlanta: two years
49             of data from ARIES. Inhalation Toxicol. 16(suppl. 1): 131-141.
50       Klemm, R.  J.; Mason, R. M., Jr. (2000) Aerosol Research and Inhalation Epidemiological Study (ARIES): air
51             quality and daily mortality statistical modeling-interim results. J. Air.  Waste Manage. Assoc. 50: 1433-1439.
52       Kopp, M. V.; Ulmer, C.;  Ihorst, G.; Seydewitz,  H. H.;  Frischer, T.; Forster, J.; Kuehr, J. (1999) Upper airway
53             inflammation in children exposed  to ambient ozone and potential signs of adaptation. Eur. Respir. J.
54             14:854-861.
55       Korrick, S.  A.; Neas, L. M.; Dockery, D. W.; Gold, D. R.; Allen, G. A.; Hill, L. B.; Kimball, K. D.; Rosner, B. A.;
56             Speizer, F. E. (1998) Effects of ozone and other pollutants on the pulmonary function of adult hikers.
57             Environ.  Health Perspect.  106: 93-99.


         January 2005                                    8-48         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G.; Frew, A.; Polak, J.; Holgate, S.
  2             (1997) Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am. J.
  3             Respir. Crit. Care Med. 156: 943-950.
  4       Liao, D.; Duan, Y.; Whitsel, E. A.; Zheng, Z.-J.; Heiss, G.; Chinchilli, V. M.; Lin, H.-M. (2004) Association of
  5             higher levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based
  6             study. Am. J. Epidemiol. 159: 768-777.
  7       Linn, W. S.; Szlachcic, Y.; Gong, H., Jr.; Kinney, P. L.; Berhane, K. T. (2000) Air pollution and daily hospital
  8             admissions in metropolitan Los Angeles. Environ. Health Perspect. 108: 427-434.
  9       Lipfert, F.  W.;  Morris, S. C.; Wyzga, R. E. (2000) Daily mortality in the Philadelphia metropolitan area and
10             size-classified paniculate matter. J. Air Waste Manage. Assoc. 50: 1501-1513.
11       Lippmann, M.; Ito, K.; Nadas, A.; Burnett, R. T. (2000) Association of paniculate matter components with daily
12             mortality and morbidity in urban populations. Cambridge, MA: Health Effects Institute; research report no.
13             95.
14       Luginaah,  I. N.; Fung, K. Y.; Gorey, K. M.; Webster, G.; Wills, C. (2004) Association of Ambient Air Pollution
15             with Respiratory Hospitalization in a Government Designated sArea of Concern^: The Case of Windsor,
16             Ontario. Environ. Health Perspect. 10.1289/ehp.7300. Available: http://dx.doi.org/ (14 December 2004).
17       Moolgavkar, S. H.; Luebeck, E. G.; Hall, T. A.; Anderson, E. L.  (1995) Air pollution and daily mortality in
18             Philadelphia. Epidemiology 6: 476-484.
19       Mortimer,  K. M.; Neas, L. M.; Dockery, D. W.; Redline, S.; Tager, I. B. (2002) The effect of air pollution on
20             inner-city children with asthma. Eur. Respir. J. 19: 699-705.
21       Newson, E. J.;  Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000) Effects of
22             short-term exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum, exhaled nitric oxide, and
23             lung function in subjects with mild atopic asthma. J. Occup. Environ. Med. 42: 270-277.
24       Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
25             factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
26             60:356-362.
27       Nightingale, J.  A.; Rogers, D. F.; Chung, K.  F.; Barnes, P. J. (2000) No effect of inhaled budesonide on the response
28             to inhaled ozone in normal subjects. Am. J. Respir. Crit. Care Med. 161: 479-486.
29       Ostro, B. (1995) Fine paniculate air pollution and mortality in two Southern California counties. Environ. Res.
30             70:98-104.
31       Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998) Nociceptive
32             mechanisms modulate ozone-induced human lung function decrements. J. Appl. Physiol. 85:  1863-1870.
33       Peel, J. L.; Tolbert, P. E.; Klein, M.; Metzger, K. B.; Planers, W. D.; Knox, T.; Mulholland, J. A.; Ryan, P. B.;
34             Frumkin, H. (2004) Ambient air pollution and respiratory  emergency department visits. Epidemiology: in
35             press.
36       Pope, C. A., Ill; Burnett, R. T.; Thun, M. J.;  Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. (2002) Lung cancer,
37             cardiopulmonary mortality, and long-term exposure to fine paniculate air  pollution.  JAMA J. Am. Med.
38             Assoc. 287: 1132-1141.
39       Postlethwait, E. M.; Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
40             estimated surface concentrations and lining layer effects. Am. J. Physiol. 274: L1006-L1016.
41       Romieu, I.; Meneses, F.; Ramirez, M.; Ruiz, S.; Padilla, R. P.; Sienra, J. J.; Gerber, M.; Grievink, L.; Dekker, R.;
42             Walda, I.; Brunekreef, B. (1998) Antioxidant supplementation and respiratory functions among workers
43             exposed to high levels of ozone. Am.  J. Respir.  Crit. Care Med. 158: 226-232.
44       Romieu, I.; Sienra-Monge, J.  J.; Ramirez-Aguilar, M.; Tellez-Rojo, M. M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.;
45             Del Rio-Navarro, B. E.; Ruiz-Navarro, M. X.; Hatch, G.; Slade, R.; Hernandez-Avila, M. (2002) Antioxidant
46             supplementation and lung functions among children with asthma exposed to high  levels of air pollutants.
47             Am. J. Respir. Crit. Care Med. 166: 703-709.
48       Samet, J. M.; Zeger, S. L.; Dominici, F.; Curriero, F.; Coursac, L; Dockery, D. W.; Schwartz, J.; Zanobetti, A.
49             (2000) The national morbidity, mortality, and air pollution study. Part II: morbidity, mortality, and air
50             pollution in the United States. Cambridge, MA: Health Effects Institute; research report no. 94, part II.
51       Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab,  L.; Bromberg, P. A.; Levine, M.; McDonnell,
52             W. F.; Devlin, R. B. (2001) Effect of antioxidant supplementation on ozone-induced lung injury in human
53             subjects. Am. J. Respir. Crit. Care Med. 164: 819-825.
54       Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996)
55             Greater ozone-induced inflammatory responses in subjects with asthma. Am. J. Respir. Crit. Care Med.
56             154:24-29.
         January 2005                                   8-49         DRAFT-DO NOT QUOTE OR CITE

-------
  1       Schelegle, E. S.; Miller, L. A.; Gershwin, L. I; Fanucchi, M. V.; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.;
  2             Mitchell, V.; Tarkington, B. K.; Wong, V. J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.;
  3             Wu, R.; Evans, M.  J.; Hyde, D. M.; Plopper, C. G. (2003) Repeated episodes of ozone inhalation amplifies
  4             the effects of allergen sensitization and inhalation on airway immune and structural development in Rhesus
  5             monkeys. Toxicol.  Appl. Pharmacol.  191: 74-85.
  6       Schierhorn, K.; Hanf, G.; Fischer, A.; Umland, B.; Olze, H.; Kunkel, G. (2002) Ozone-induced release of
  7             neuropeptides from human nasal mucosa cells. Int. Arch. Allergy Immunol. 129: 145-151.
  8       Schierhorn, K.; Zhang, M.; Matthias, C.; Kunkel, G. (1999) Influence of ozone and nitrogen dioxide on histamine
  9             and interleukin formation in a human nasal mucosa culture system. Am. J. Respir. Cell Mol. Biol.
10             20: 1013-1019.
11       Schwartz, J. (2004) How sensitive is the association between ozone and daily deaths to control for temperature?
12             Am. J. Respir. Crit. Care Med.: in press, 10.1164/rccm.200407-933OC.
13       Schwartz, J.; Spix, C.; Touloumi,  G.; Bacharova, L.; Barumamdzadeh, T.; le Tertre, A.; Piekarksi, T.; Ponce de
14             Leon, A.; Ponka, A.; Rossi, G.; Saez, M.; Schouten, J. P. (1996) Methodological issues in studies of air
15             pollution and daily counts of deaths or hospital admissions. In: St Leger, S., ed. The APHEA project.
16             Short term effects of air pollution on health: a European approach using epidemiological time series data.
17             J. Epidemiol. Commun. Health 50(suppl. 1): S3-SI 1.
18       Sherwin, R. P.; Richters, V.; Kraft, P.; Richters, A. (2000) Centriacinar region inflammatory disease in young
19             individuals: a comparative  study  of Miami and Los Angeles residents. Virchows Arch. 437: 422-428.
20       Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, I.; Helleday, R.; Kelly, F. J.; Frew, A. J.;
21             Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects.
22             Respir. Med. 96: 352-358.
23       Stieb, D. M.; Burnett, R. T.; Beveridge, R. C.; Brook, J. R. (1996) Association between ozone and asthma
24             emergency department visits in Saint John, New Brunswick, Canada. Environ. Health Perspect.
25             104: 1354-1360.
26       Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M. G. (1990) Measurements of cardiopulmonary response in
27             awake rats during acute exposure to near-ambient concentrations of ozone. J. Appl. Toxicol. 10:7-15.
28       Tolbert, P. E.; Mulholland, J. A.; Macintosh, D. L.; Xu, F.; Daniels, D.; Devine, O. J.; Carlin, B. P.; Klein, M.;
29             Dorley,  J.; Butler, A. J.; Nordenberg, D. F.;  Frumkin, H.; Ryan, P. B.; White, M. C. (2000) Air quality and
30             pediatric emergency room visits for asthma in Atlanta, Georgia. Am. J. Epidemiol. 151: 798-810.
31       Trenga, C. A.;  Koenig, J. Q.; Williams,  P. V. (2001) Dietary antioxidants and ozone-induced bronchial
32             hyperresponsiveness in adults with asthma. Arch. Environ. Health 56: 242-249.
33       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
34             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
35             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
36             http://cfpub2.epa.gov/ncea/.
37       U.S. Environmental Protection Agency. (2004) Air quality criteria for paniculate matter. Research Triangle Park,
3 8             NC: National Center for Environmental Assessment; report no. EPA/600/P-99/002aF-bF. 2v. Available:
39             http://cfpub.epa.gov/ncea/ [9 November, 2004].
40       Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004)  Uptake distribution of ozone in human lungs: intersubject
41             variability in physiologic response. Boston, MA: Health Effects Institute.
42       Vedal, S.; Brauer, M.; White, R.; Petkau, J. (2003) Air pollution and  daily mortality in a city with low levels of
43             pollution. Environ. Health Perspect.  Ill: 45-51.
44
         January 2005                                    8-50          DRAFT-DO NOT QUOTE OR CITE

-------