x-xEPA
             United States
             Environmental Protection
             Agency
              Office of Research and
              Development
              Washington, DC 20460
 EPA/600/AP-93/004d
 February 1994
 External Review Draft
Air Quality
Criteria for
Ozone and
Related
Photochemical
Oxidants
Review
Draft
(Do Not
Cite or
Quote)
             Chapter 1.
             Executive Summary
             and
             Chapter 9.
             Integrative  Summary of
             Ozone Health  Effects
                          Notice
              This document is a preliminary draft. It has not been formally
             released by EPA and should not at this stage be construed to
             represent Agency policy. It is being circulated for comment on its
             technical accuracy and policy implications.

-------
DRAFT-DO NOT QUOTE OR CITE                            EPA/6007AP-93/004c
                                                         February 1994
                                                         External Review Draf
               Air  Quality  Criteria  for Ozone
         and Related  Photochemical Oxidants
                    Chapter 1.  Executive Summary

                                   and

                  Chapter 9.  Integrative Summary of
                         Ozone Health Effects
                                   NOTICE
                  This document is a preliminary draft. It has not been formally
                  released by EPA and should not at this stage be construed to
                  represent Agency policy. It is being circulated for comment on
                  its technical accuracy and policy implications.
                    Environmental Criteria and Assessment Office
                   Office of Health and Environmental Assessment
                        Office of Research and Development
                       U.S. Environmental Protection Agency
                        Research Triangle Park, NC 27711
                                                   ^ZA) Printed on Recycled Paper

                                             U.S. Environmental Protection Agency
                                             Region 5, Library (PL-12J)
                                             77 West Jackson Boulevard, 12th floor
                                             Chicago,  IL 60604-3590

-------
                                  DISCLAIMER

     This document is an external 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.
 February 1994                          IV-ii      DRAFT-DO NOT QUOTE OR CITE

-------
                                      PREFACE

     In 1971, the U.S. Environmental Protection Agency (EPA) promulgated National
Ambient Air Quality Standards (NAAQS) to protect the public health and welfare from
adverse effects of photochemical oxidants.  In 1979, the chemical designation of the
standards was changed from photochemical oxidants to ozone (O3).  This document,
therefore,  focuses primarily on the scientific air quality criteria for O3 and, to a lesser extent,
for other photochemical oxidants like hydrogen peroxide and the peroxyacyl nitrates.
     The EPA promulgates the NAAQS on the basis of scientific information contained in
air quality criteria documents.  The previous O3  criteria document, Air Quality Criteria for
Ozone and Other Photochemical Oxidants, was released in August 1986 and  a supplement,
Summary of Selected New Information on Effects of Ozone on Health and Vegetation, was
released in January  1992.  These documents were used as the basis for a March 1993
decision by EPA not to revise the existing 1-h NAAQS for O3.  That decision, however, did
not take into account some of the newer scientific data that became available after the 1986
criteria document.  This revised air quality criteria document for O3 and related
photochemical oxidants critically evaluates and assesses the latest scientific data associated
with exposure to concentrations of these pollutants found in ambient air.  Emphasis is placed
on presentation of health  and environmental effects data; however, other scientific data are
presented and evaluated in order to provide a better understanding of the nature, sources,
distribution, measurement, and concentrations of O3 and related photochemical oxidants and
their precursors in the environment.  The document is not an exhaustive literature review;
rather it assesses the most pertinent literature available through 1993.
     The chapters in this volume summarize and interpret key information drawn from the
other,  more detailed chapters (Chapters 2 through 8) of the present document, which were
prepared and peer reviewed by experts from various Federal and State governmental offices,
academia, and private industry for use by EPA to support decision making regarding
potential risks to public health and welfare.  The Environmental  Criteria and Assessment
Office of EPA's Office of Health and Environmental Assessment acknowledges with
appreciation the contributions provided by these  authors and reviewers as well as the
diligence of its  staff and contractors in the preparation of this document at the request of the
Office of Air Quality Planning and Standards.

February 1994                           IV-iii     DRAFT-DO NOT QUOTE OR CITE

-------
                           TABLE OF CONTENTS
                                                                  Page

LIST OF TABLES  	     IV-vii
LIST OF FIGURES	     IV-viii
AUTHORS, CONTRIBUTORS, AND REVIEWERS 	      IV-ix
U.S. ENVIRONMENTAL PROTECTION AGENCY PROJECT TEAM
FOR DEVELOPMENT OF AIR QUALITY CRITERIA FOR OZONE
AND RELATED PHOTOCHEMICAL OXIDANTS	      IV-xi

1.   EXECUTIVE SUMMARY 	     1-1
    1.1   INTRODUCTION	     1-1
    1.2   LEGISLATIVE AND REGULATORY BACKGROUND  	      1-1
    1.3   TROPOSPHERIC OZONE AND ITS PRECURSORS  	      1-1
    1.4   ENVIRONMENTAL CONCENTRATIONS, PATTERNS, AND
         EXPOSURE ESTIMATES  	     1-9
    1.5   ENVIRONMENTAL EFFECTS OF OZONE AND RELATED
         PHOTOCHEMICAL OXIDANTS	     1-13
    1.6   TOXICOLOGICAL EFFECTS OF OZONE  	      1-21
    1.7   HUMAN HEALTH EFFECTS OF OZONE AND RELATED
         PHOTOCHEMICAL OXIDANTS	     1-26
    1.8   EXTRAPOLATION OF ANIMAL TOXICOLOGICAL DATA
         TO HUMANS	     1-31
    1.9   INTEGRATIVE SUMMARY OF OZONE HEALTH
         EFFECTS  	     1-32

9.   INTEGRATIVE SUMMARY OF OZONE HEALTH EFFECTS  	      9-1
    9.1   INTRODUCTION  	     9-1
    9.2   HEALTH EFFECTS OF SHORT-TERM EXPOSURES	      9-2
         9.2.1   Exposure-Dose Relationships	     9-2
         9.2.2   Physiological Responses to Ozone Exposure	      9-4
                9.2.2.1 Respiratory Symptom Responses  	      9-5
                9.2.2.2 Changes in Lung Volume	     9-6
                9.2.2.3 Responses to Ambient Ozone Exposures	      9-7
                9.2.2.4 Changes in Airway Resistance	      9-7
                9.2.2.5 Changes in Breathing Pattern   	      9-8
                9.2.2.6 Changes in Airway Responsiveness	      9-8
                9.2.2.7 Small Airways Responses	     9-8
                9.2.2.8 Effects on Exercise Performance  	      9-9
         9.2.3   Exacerbation of Existing Disease	     9-9
                9.2.3.1 Responses of Asthmatics to Controlled
                       Ozone Exposure  	     9-9
                9.2.3.2 Increased Hospital Admissions  and
                       Asthma Attacks	     9-10
         9.2.4   Cellular-Biochemical Responses 	     9-10
                9.2.4.1 Inflammation and Cell Damage  	     9-10
                9.2.4.2 Host Defense  	     9-14

February 1994                        IV-v     DRAFT-DO NOT QUOTE OR CITE

-------
                        TABLE OF CONTENTS (cont'd)
                                                                     Page

          9.2.5   Ozone Exposure-Response Relationships	      9-16
                 9.2.5.1 Prediction and Summary of Mean
                        Responses  	      9-20
                 9.2.5.2 Prediction and Summary of Individual
                        Responses  	      9-25
    9.3    HEALTH EFFECTS OF LONG-TERM EXPOSURES	      9-30
          9.3.1   Repeated Exposures	      9-30
          9.3.2   Prolonged Exposures	      9-31
          9.3.3   Genotoxicity and Carcinogenicity of Ozone  	      9-35
    9.4    COMBINED POLLUTANT EXPOSURES	      9-36
    9.5    CONCLUSIONS	      9-37
    REFERENCES	      9-43
February 1994                        IV-vi      DRAFT-DO NOT QUOTE OR CITE

-------
                               LIST OF TABLES

Number                                                               Page

9-1       Gradation of Physiological Responses to Short-Term Ozone
         Exposure	      9-29
 February 1994                         IV-vii     DRAFT-DO NOT QUOTE OR CITE

-------
                                  LIST OF FIGURES

Number                                                                        Page

9-1       Mean predicted postexposure to preexposure changes in forced
          expiratory volume in one second following two-hour exposures
          to ozone with intermittent exercise	        9-20

9-2       Predicted mean decrements in forced expiratory volume in one
          second for one- and two-hour exposures to ozone with
          intermittent heavy exercise and 6.6-hour exposures with
          moderate prolonged exercise	        9-21

9-3       Derived means of bronchoalveolar lavage protein and the
          exponential model as tune varies from two to eight hours	        9-22

9-4       Predicted mean forced vital capacity for rats exposed to ozone
          while undergoing intermittent carbon dioxide-induced hyperpnea .  .        9-22

9-5       Predicted mean decrements in forced expiratory volume in one
          second following two-hour exposures to ozone while undergoing
          heavy intermittent exercise for three age groups	        9-25

9-6       The distribution of response for 87 subjects exposed to clean
          air and at least one of 0.08, 0.10,  or 0.12 ppm ozone is
          shown here  	       9-26

9-7       Proportion of heavily exercising individuals predicted to
          experience a 10% decrement in forced expiratory volume in one
          second following a one- or two-hour exposure to ozone	        9-27

9-8       Proportion of heavily exercising individuals predicted to
          experience mild cough following a two-hour ozone exposure  ....        9-28

9-9       Proportion of moderately exercising individuals exposed  to ozone
          for 6.6 hours predicted to experience  5, 10,  or 15% decrements
          in forced expiratory volume in one second as a function  of
          concentration times time at 24 years of age	        9-28

9-10      Schematic comparison of the duration-response profiles for
          epithelial hyperplasia, bronchioloalveolar exudation, and
          interstitial fibrosis in the centriacinar region of lung exposed
          to a constant low concentration of ozone  	        9-33
 February 1994                           IV-viii      DRAFT-DO NOT QUOTE OR CITE

-------
                  AUTHORS, CONTRIBUTORS, AND REVIEWERS

                       CHAPTER 1. EXECUTIVE SUMMARY

Principal Authors

Mr. James A. Raub—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Mr. William G. Ewald—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. J.H.B. Garner—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Judith A. Graham—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Ms. Beverly E. Tilton—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711


       CHAPTER 9. INTEGRATIVE SUMMARY OF OZONE HEALTH EFFECTS

Principal Authors

Dr. Daniel L. Costa—Health Effects Research Laboratory (MD-82), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711

Dr. Robert B. Devlin—Health Effects Research Laboratory (MD-58), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711

Dr. Lawrence J. Folinsbee—Health Effects Research Laboratory (MD-58),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Timothy R. Gerrity—Health Effects Research Laboratory (MD-58), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711

Dr. Judith A. Graham—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. William F. McDonnell—Health Effects Research Laboratory (MD-58),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Mr. James A. Raub—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711


February 1994                          IV-ix     DRAFT-DO NOT QUOTE OR CITE

-------
              AUTHORS, CONTRIBUTORS, AND REVIEWERS (cont'd)
Contributors and Reviewers

Dr. Frederick J. Miller—Chemical Industry Institute of Toxicology, P.O. Box 12137,
Research Triangle Park, NC 27709

Dr. Edward S. Schelegle—Department of Human Physiology, School of Medicine,
Building 1-A,  Room 4140, University of California, Davis, CA 95616

Dr. Walter S.  Tyler—Department of Anatomy, School of Veterinary Medicine, University of
California, Davis, CA  95616
February 1994                          IV-x      DRAFT-DO NOT QUOTE OR CITE

-------
                 U.S. ENVIRONMENTAL PROTECTION AGENCY
        PROJECT TEAM FOR DEVELOPMENT OF AIR QUALITY CRITERIA
            FOR OZONE AND RELATED PHOTOCHEMICAL OXIDANTS
Scientific Staff

Mr. James A. Raub—Health Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. A. Paul Altshuller—Physical Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Mr. William G. Ewald—Health Scientist, Environmental Criteria and Assessment  Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. J.H.B. Garner—Ecologist, Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Judith A. Graham—Associate Director, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Ms. Ellie R. Speh—Secretary, Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental  Protection Agency, Research Triangle Park, NC 27711

Ms. Beverly E. Tilton—Physical Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
 Technical Support Staff

 Mr. Douglas B. Fennell—Technical Information Specialist, Environmental Criteria and
 Assessment Office (MD-52), U.S. Environmental Protection Agency, Research Triangle Park,
 NC  27711

 Mr. Allen G. Hoyt—Technical Editor and Graphic Artist, Environmental Criteria and
 Assessment Office (MD-52), U.S. Environmental Protection Agency, Research Triangle Park,
 NC  27711

 Ms. Diane H. Ray—Technical Information Manager (Public Comments), Environmental
 Criteria and Assessment Office (MD-52), U.S.  Environmental Protection Agency, Research
 Triangle Park, NC  27711

 Mr. Richard N. Wilson—Clerk, Environmental Criteria and Assessment Office (MD-52),
 U.S. Environmental Protection Agency, Research Triangle Park, NC  27711
 February 1994                          IV-xi      DRAFT-DO NOT QUOTE OR CITE

-------
                  U.S. ENVIRONMENTAL PROTECTION AGENCY
        PROJECT TEAM FOR DEVELOPMENT OF AIR QUALITY CRITERIA
            FOR OZONE AND RELATED PHOTOCHEMICAL OXIDANTS
                                     (cont'd)
Document Production Staff

Ms. Marianne Barrier—Graphic Artist, ManTech Environmental, P.O. Box 12313,
Research Triangle Park, NC  27709

Mr. John R. Barton—Document Production Coordinator, ManTech Environmental
Technology, Inc., P.O. Box 12313, Research Triangle Park, NC 27709

Ms. Lynette D. Cradle—Lead Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709

Ms. Jorja R. Followill—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709

Ms. Wendy B. Lloyd—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709

Mr. Peter J. Winz—Technical Editor, Mantech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Technical Reference Staff

Mr. John A. Bennett—Bibliographic Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709

Ms. S. Blythe Hatcher—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC  27709

Ms. Susan L. McDonald—Bibliographic Editor, Information Organizers, Inc.,
P.O. Box 14391, Research Triangle Park, NC 27709

Ms. Carol J. Rankin—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC  27709

Ms. Deborah L. Staves—Bibliographic Editor, Information Organizers, Inc.,
P.O. Box 14391, Research Triangle Park, NC 27709

Ms. Patricia R. Tierney—Bibliographic Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
February 1994                         IV-xii      DRAFT-DO NOT QUOTE OR CITE

-------
 i                          1.  EXECUTIVE SUMMARY

 2

 3      1.1   INTRODUCTION

 4           The external review draft document Air Quality Criteria for Ozone and Related
 5      Photochemical Oxidants evaluates the latest scientific information useful in deriving criteria
 6      that form the scientific basis for U.S. Environmental Protection Agency (EPA) decisions
 7      regarding the National Ambient Air Quality Standards (NAAQS) for ozone (O3). This
 8      Executive Summary concisely summarizes key conclusions from the document, which
 9      comprises nine chapters.  Following this Executive Summary is a brief Introduction
10      (Chapter 2) containing information on the legislative and regulatory background for review of
11      the O3 NAAQS. Chapter 3 provides information on the chemistry, sources,  emissions,
12      measurement, and transport of O3 and related photochemical oxidants and their precursors,
13      whereas Chapter 4 covers environmental concentrations, patterns, and exposure estimates of
14      O3 and oxidants.  This is followed by Chapter 5, dealing with the environmental effects;
15      Chapters 6, 7, and 8 discuss, respectively, animal lexicological studies, human health effects,
16      and the extrapolation of animal lexicological data to humans. The last chapter,  Chapter 9,
17      provides an integrative,  interpretative evaluation of heallh effecls associated with exposure to
18      O3.  The following subsections follow the chapter organization of the external review draft.
19
20
21      1.2   LEGISLATIVE AND REGULATORY BACKGROUND
22
23          The EPA is required under Sections  108 and 109 of the Clean Air Act to periodically
24     evaluate Ihe air quality criteria thai reflect the latesl scientific information relevant to review
25     of the O3 NAAQS. These air qualily criteria, contained in Ihe criteria document, are useful
26     for indicating the land and extent of all identifiable effects on public health or welfare lhat
27     may be  expected from Ihe presence of O3  and related photochemical oxidanls in ambienl air.
28     The  lasl criteria documenl was released in 1986, and a supplement was released in  1992.
29     These documents were used as Ihe basis for a March  1993 decision by EPA lo nol  revise Ihe
30     existing 1-h NAAQS for O3 al lhal tune.  That decision,  however, did not take into
31     consideration more recent scientific information that has been published since the last
32     literature review in early 1989. The purpose of this revised criteria documenl,  Iherefore, is
33     lo summarize Ihe pertinenl information conlained hi Ihe previous O3 criteria documenl and to
34     critically evaluate and assess Ihe more recenl scientific dala associaled wilh exposure to the
35     concentrations of O3 and related photochemical  oxidants that are found hi ambient air.
36
37
38     1.3   TROPOSPHERIC OZONE AND ITS PRECURSORS
39
40     Tropospheric Ozone Chemistry
41
42          Ozone is found  in the slratosphere, the "free" troposphere,  and the planelary boundary
43     layer (PEL) of the earth's almosphere.  Ozone occurs in the stratosphere as  Ihe result of
44     chemical reactions initialed by short-wavelength radiation from the sun.  In the "free"

       February 1994                           1-1       DRAFT-DO NOT QUOTE OR CITE

-------
 1     troposphere, O3 occurs as the result of incursions from the stratosphere; upward venting
 2     from the PEL (the layer next to the earth's surface) through certain cloud processes; and
 3     photochemical formation from precursors, notably methane (CF^), carbon monoxide (CO),
 4     and nitrogen oxides (NOX).  In the PEL, O3 occurs as the result of downward mixing from
 5     the stratosphere and free troposphere and as the result of photochemical processes occurring
 6     within the PEL.
 7
 8          The photochemical production of O3 and other oxidants found at the earth's surface
 9     (in the PEL, troposphere, or ambient air, used interchangeably in this summary) is the result
10     of atmospheric physical processes and complex, nonlinear chemical processes involving two
11     classes of precursor pollutants, reactive volatile organic compounds (VOCs) and NOX. The
12     only significant initiator of the photochemical production of O3 in the polluted troposphere is
13     the photolysis of nitrogen dioxide (NO2), yielding nitric oxide (NO) and a ground-state
14     oxygen atom that reacts with molecular oxygen to form O3.  The O3 thus formed reacts with
15     NO, yielding O2 and NO2.  These cyclic reactions attain equilibrium in the absence of
16     VOCs.  In the presence, however, of VOCs, which are abundant in polluted ambient air,  the
17     equilibrium is upset, resulting in a net increase in O3. Methane is the chief VOC found in
18     the free troposphere and in relatively "clean" areas of the PEL.  The VOCs found in polluted
19     ambient air are more complex and more reactive than CH^ but, as with CH4, their
20     atmospheric oxidative degradation is initiated through attack on the VOCs by hydroxyl (OH)
21     radicals.  As in the CH4 oxidation cycle, the conversion of NO to NO2 during the oxidation
22     of VOCs is accompanied by the production of O3 and the efficient regeneration of the OH
23     radical.  The O3 and peroxyacyl nitrates (PANs) formed in polluted atmospheres increase
24     with the  NO2/NO concentration ratio.
25
26          At  night, in the absence of photolysis of reactants, the simultaneous presence of 03  and
27     NO2 results in the formation of the nitrate radical, NO3.  Reactions with NO3 radicals appear
28     to constitute major sinks for alkenes, cresols,  and some other compounds, although alkyl
29     nitrate chemistry is not well characterized.
30
31          Most inorganic gas-phase processes, that is, the nitrogen cycle and its  interrelationships
32     with O3  production, are well understood.  The chemistry of the VOCs in ambient air is not
33     as well understood.  It is well-known, however, that the chemical loss processes of gas-phase
34     VOCs, with concomitant production of O3, include reaction with OH and NO3 radicals and
35     O3,  and  photolysis.  Reaction with the OH radical is the only important atmospheric reaction
36     (loss process) for alkanes, aromatic hydrocarbons, and the higher aldehydes and ketones that
37     lack >C=C< bonds; and the only atmospheric reaction of alcohols and ethers. Photolysis
38     is the major loss process for formaldehyde and  acetone.  Reactions with OH and NO3
39     radicals and with O3 are all important loss processes for alkenes and for carbonyls containing
40      >C=C<  bonds.
41
42          Uncertainties in the atmospheric chemistry of the VOCs can affect quantification of the
43     NO-to-NO2 conversion  and  of O3 yields; and can present difficulties in representation of
44     mechanisms, products, and product yields in O3 air quality models.  Major  uncertainties in
45     current understanding of the atmospheric chemistry of the VOCs include (1) chemistry of
46     alkyl nitrate formation,  (2) products and reaction rates for > C4 alkanes and for branched


       February 1994                             1-2       DRAFT-DO NOT QUOTE OR CITE

-------
 1     alkanes, (3) mechanisms and products of alkene-O3 reactions, and (4) products of aromatic
 2     hydrocarbons in both urban and rural atmospheres.
 3
 4          It should be noted that the atmospheric chemical processes involved in the
 5     photooxidation of VOCs and the formation of O3 and other photochemical oxidants can lead
 6     also to the formation of OH radicals and of particulate-phase organic compounds.  The OH
 7     radicals produced can not only oxidize VOCs but can also react with NO2 and sulfur dioxide
 8     (SO2) to form nitric and sulfuric acids, respectively, which become incorporated into aerosols
 9     as paniculate nitrate and sulfate.
10
11     Meteorological Influences on Ozone Formation and Transport
12
13          The surface energy (radiation) budget of the earth strongly influences the dynamics of
14     the PEL.  The redistribution of energy through the PEL creates thermodynamic conditions
15     that influence vertical mixing. Growing evidence indicates that the strict use of mixing
16     heights in modeling is an oversimplification of the complex processes by which pollutants are
17     redistributed within urban areas; and that it is necessary to treat the turbulent structure of the
18     atmosphere directly and acknowledge the vertical variations in mixing.  Energy balances
19     therefore require study so that more realistic simulations can be made of the structure of the
20     PEL.
21
22          Day-to-day variability in O3 concentrations depends heavily on day-to-day variations in
23     meteorological conditions,  such as the degree of mixing that occurs between release of a
24     pollutant or its precursors and their arrival at a receptor; the occurrence of inversion layers
25     (layers in which temperature increases with height above ground level); and the transport of
26     O3 left overnight in layers aloft and subsequent downward mixing of that O3 to the surface.
27
28          The transport of O3 and its precursors beyond the urban scale (< 50 km) to neighboring
29     rural and urban areas has been well documented and was described in the 1986 EPA criteria
30     document for O3.  Episodes of high O3  concentrations in urban areas are often associated
31     with high concentrations of O3 in the surroundings.  Areas of O3 accumulation are
32     characterized by (1) synoptic-scale subsidence of air in the free troposphere, resulting in
33     development of an elevated inversion layer; (2) relatively low wind speeds associated with
34     the weak horizontal pressure gradient around a surface high pressure system; (3) a lack of
35     cloudiness; and (4) high temperatures.
36
37          Ultraviolet (UV) radiation from the sun plays a key role in initiating the photochemical
38     processes leading to O3 formation and affects individual photolytic reaction steps.  Still, there
39     is little empirical evidence in the literature linking day-to-day variations in observed UV
40     radiation levels with variations in O3 levels.  An association, however,  between tropospheric
41     O3 concentrations and tropospheric temperature has been demonstrated.  Empirical data from
42     four urban areas, for example, show an apparent upper bound on O3 concentrations that
43     increases with temperature. A similar qualitative relationship exists at a number of rural
44     locations.
45
        February 1994                             1-3       DRAFT-DO NOT QUOTE OR CITE

-------
 1          The relationship between wind speed and O3 buildup varies from one part of the
 2     country to another.
 3
 4          Statistical techniques (e.g., regression techniques) can be used to help identify real
 5     trends in O3 concentrations, both intra- and interannual, by normalizing meteorological
 6     variability. In the Southern Oxidant Study, for example, regression techniques were
 7     successfully used to forecast O3 levels to ensure that specialized measurements were made on
 8     appropriate days.
 9
10     Precursors
11
12     Nitrogen Oxides Emissions
13
14          Anthropogenic NOX is associated with combustion processes.  The primary pollutant
15     emitted is NO formed at high combustion temperatures from the nitrogen and oxygen in air
16     and from nitrogen in the combustion fuel.  Emissions of NOX in 1991 in the United States
17     totaled 21.39 Tg. The two largest single NOX emission sources are electric power generation
18     and highway vehicles.  Because a large proportion of anthropogenic NOX emissions come
19     from distinct point sources, published annual estimates are thought to be very reliable.
20
21          Natural NOX sources include stratospheric intrusion, oceans, lightning, soils, and
22     wildfires. Lightning and soil emissions are the only two significant natural sources of NOX
23     in the United States. Combined natural sources contribute about 2.2  Tg of NOX to the
24     troposphere over the continental United States.  Uncertainties in natural NOX inventories  are
25     much larger, however, than for anthropogenic NOX emissions.
26
27     Volatile Organic Compound Emissions
28
29          Hundreds of VOCs, commonly containing from 2 to about 12 carbon atoms, are
30     emitted by evaporative and combustion processes from a large number of source types.
31     Total U.S. VOC emissions in 1991 were estimated at 21.0 Tg. The two largest source
32     categories were industrial processes (10.0 Tg) and transportation (7.9 Tg).  Emissions of
33     VOCs from highway vehicles accounted for almost 75 % of the transportation-related
34     emissions; studies have shown that the majority of these VOC emissions come from about
35     20% of the automobiles in service, many, perhaps most, of which are older cars  that are
36     poorly maintained.  The accuracy of VOC emission estimates is difficult to determine, both
37     for stationary and mobile sources.
38
39          Vegetation emits  significant quantities of VOCs into the atmosphere, chiefly
40     monoterpenes and isoprene, but also oxygenated VOCs, according to recent studies. The
41     most recent biogenic VOC emissions estimate for the United States showed annual emissions
42     of 29.1 Tg/year.
43
44          Uncertainties in both biogenic and anthropogenic VOC emission inventories prevent
45     establishing the relative contributions of these two categories.
46
        February 1994                            1-4       DRAFT-DO NOT QUOTE OR CITE

-------
 1     Concentrations of Volatile Organic Compounds in Ambient Air
 2
 3          The VOCs most frequently analyzed in ambient air are the nonmethane hydrocarbons
 4     (NMHCs).  Morning concentrations (6:00 a.m. to 9:00 a.m.) have been measured most often
 5     because of the use of morning data in the Empirical Kinetics Modeling Approach (EKMA)
 6     and in air quality simulation models. Measurements made in 22 cities in 1984 and 19 cities
 7     in 1985 showed median NMHC concentrations ranging from 0.39 ppm C to 1.27 ppm C for
 8     1984; and from 0.38 ppm C to 1.63 ppm C in 1985.  Overall median values from all urban
 9     sites were about 0.72 ppm C in 1984 and 0.60 ppm C in 1985.
10
11          Concurrent measurements of anthropogenic and biogenic NMHCs have shown that
12     biogenic NMHCs  usually constituted much less than 10% of the total NMHCs.  For
13     example, average  isoprene concentrations ranged  from 0.001  to 0.020 ppm C and terpenes
14     from 0.001 to 0.030 ppm C.
15
16     Concentrations of Nitrogen Oxides in Ambient Air
17
18          Measurements of NOX made in 22 and 19 U.S. cities hi 1984 and 1985, respectively,
19     showed median 6:00-to-9:00 a.m. NOX concentrations ranging from 0.02 to 0.08 ppm hi
20     most of these cities. Nonurban NOX concentrations, reported as average seasonal or annual
21     NOX, range from  <0.005 to 0.015  ppm.
22
23          Ratios of 6:00-to-9:00 a.m. nonmethane organic compounds (NMOC) to NOX are
24     higher in southeastern and southwestern U.S. cities than hi northeastern and midwestern U.S.
25     cities, according to data from EPA's multicity studies conducted hi 1984 and 1985. Median
26     ratios ranged from 9.1 to 37.7 hi 1984; hi 1985,  median ratios ranged from 6.5 to 53.2 hi
27     the cities studied.  Rural NMOC/NOX ratios tend to be higher than urban ratios. Morning
28     (6:00-to-9:00 a.m.)  NMOC/NOX ratios are used in  the EKMA type of trajectory model. The
29     correlation  of NMOC/NOX ratios with maximum 1-h O3 concentrations, however, was weak
30     hi a recent  analysis.
31
32     Source Apportionment and Reconciliation
33
34          Source apportionment (now regarded as synonymous with receptor modeling) refers to
35     determining the quantitative contributions of various sources of VOCs to ambient air
36     pollutant concentrations.  Source reconciliation refers  to the comparison of measured ambient
37     VOC concentrations with emissions inventory estimates of VOC source emission rates for the
38     purpose of validating the inventories.
39
40          Recent findings showed that vehicle exhaust was the dominant contributor to ambient
41     VOCs in seven of eight U.S. cities  studied.  Whole gasoline  contributions have been
42     estimated to be equal to vehicle exhaust in one study and 20 % of vehicle exhaust in a second
43     study.
44
45          Estimates of biogenic VOCs at a  downtown site in Atlanta hi 1990 indicated a lower
46     limit of 2 % (24-h average) for the biogenic percentage of total ambient VOCs at that location
47     (isoprene was used as the biogenic  indicator species). The percentage varies during the 24-h

       February 1994                            1-5       DRAFT-DO NOT QUOTE OR CITE

-------
 1     period because of the diurnal (e.g., temperature, light intensity) dependence of isoprene
 2     concentrations.
 3
 4          Source reconciliation data have shown disparities between emission inventory estimates
 5     and receptor-estimated contributions.  For biogenics,  emission estimates are greater than
 6     receptor-estimated contributions.  The reverse has been true for natural gas contributions
 7     estimated for Los Angeles, Columbus, and Atlanta; and for refinery emissions in Chicago.
 8
 9     Analytical Methods  for Oxidants and Their Precursors
10
11     Oxidants
12
13          Current methods used to measure O3 are chemiluminescence (CL); UV absorption
14     spectrometry; and newly developed spectroscopic and chemical approaches, including
15     chemical approaches applied to passive sampling devices (PSDs) for O3.
16
17          The CL method has been designated as the reference method by EPA.  Detection limits
18     of 0.005 ppm and a response time of < 30 s are typical of currently available commercial
19     instruments.  A positive interference from atmospheric  water vapor was reported in the
20     1970s and has recently been confirmed. Proper calibration can minimize this source of
21     error.
22
23          Commercial UV photometers for measuring O3 have detection limits of about
24     0.005 ppm and a response time of < 1 min.  Because the measurement is absolute, UV
25     photometry is also used  to calibrate O3 methods.  A potential disadvantage of UV photometry
26     is that atmospheric constituents that absorb 254 nm radiation, the wavelength at which 63 is
27     measured, will cause a positive interference in O3 measurements.  Interferences have been
28     reported in two recent studies, but assessment of the potential importance of such
29     interferences (e.g., toluene, styrene, cresols, nitrocresols) is hindered by lack of absorption
30     spectra data in the 250 nm range and by lack of aerometric data for the potentially interfering
31     species.
32
33          Calibration of O3 measurement methods (other than PSDs) is done by UV spectrometry
34     or by gas-phase titration (GPT) of O3 with NO. Ultraviolet photometry is the reference
35     calibration method approved by EPA.  Ozone is unstable and must be generated in situ at
36     time of use to produce calibration mixtures.
37
38          Peroxyacetyl nitrate and the higher PANs are normally measured by gas
39     chromatography using an electron capture detector (GC-ECD). Detection limits have now
40     been extended to 1 to 5  ppt.  The preparation of reliable calibration standards is difficult
41     because PAN is unstable (explosive and subject to surface-related decomposition), but several
42     methods are available.
43
44
        February 1994                            1-6       DRAFT-DO NOT QUOTE OR CITE

-------
 1     Volatile Organic Compounds
 2
 3          Traditionally, NMHCs and other NMOCs have been measured by methods that employ
 4     a flame ionization detector (FID) as the sensing element that measures a change in ion
 5     intensity resulting from the combustion of air containing organic compounds. The method
 6     recommended by EPA for total NMOC measurement involves the cryogenic preconcentration
 7     of NMOCs and the measurement of the revolatilized NMOCs using FID.  The main
 8     technique for speciated NMOC/NMHC measurements is cryogenic preconcentration followed
 9     by GC-FID.  Systems for sampling and analysis of VOCs have now been developed that
10     require no liquid cryogen for operation, yet provide sufficient resolution of species.
11
12          Stainless steel canisters have become the containers of choice for collection of whole-air
13     samples for NMHC/NMOC data.  Calibration procedures for NMOC instrumentation require
14     the generation, by static or dynamic systems, of dilute mixtures at concentrations expected to
15     occur in ambient air.
16
17          Preferred methods for measuring carbonyl species (aldehydes and ketones) in ambient
18     air are spectroscopic  methods; on-line colorimetric methods; and the most common method
19     in current use for measuring gas-phase carbonyl compounds in  ambient air, which is  the
20     high-performance liquid chromatography (HPLC) method employing
21     2,4-dinitrophenylhydrazine (DNPH) derivatization in a silica gel cartridge.  Use of an
22     O3 scrubber has been recommended to prevent interference  in this method by O3  in ambient
23     air.  Several methods are available for preparing stable calibration mixtures.
24
25     Oxides of Nitrogen
26
27          Nitric oxide and NO2 comprise the NOX compounds involved as precursors to O3 and
28     other photochemical oxidants.
29
30           The most common method of NO measurement is the gas-phase CL reaction with O3,
31     which  is essentially specific for NO.  Commercial NO monitors have detection limits of a
32     few parts per billion  by volume (ppbv) in ambient air but may  not have sensitivity sufficient
33     for surface measurements in rural or remote areas, or for airborne measurements.  Direct
34     spectroscopic methods for NO exist that have very high sensitivity and selectivity for NO,
35     but their complexity, size, and cost restrict these methods to research applications. No PSDs
36     presently exist for measurement of NO.
37
38          Chemiluminescence analyzers are the method of choice for NO2 measurement,  even
39     though they do not measure NO2 directly.  Minimum detection levels for NO2 have been
40     reported to be 5 to 13 ppb, but more  recent evaluations have indicated detection limits of
41     0.5 to  1 ppbv.  Reduction of NO2 to NO is required for measurement.  In practice,  selective
42     measurement of NOX by this approach has proved difficult,  and the NO2 value inferred from
43     such measurements may be significantly in error.
44
45          Several spectroscopic approaches to NO2 detection have been developed but share the
46     drawbacks of spectroscopic NO methods.  Passive samplers for NO2 exist, but are still in the
47     developmental stage for ambient air monitoring.

       February 1994                             1-7       DRAFT-DO NOT QUOTE OR CITE

-------
 1          Calibration of methods for NO measurement is done using standard cylinders of NO in
 2     nitrogen.  Calibration of methods for NO2 measurement include use of cylinders of NO2 in
 3     nitrogen or air, use of permeation tubes, and GPT.
 4
 5     Ozone Air Quality  Models
 6
 7     Models and Their Components
 8
 9          Photochemical air quality models are used to predict how O3 concentrations change in
10     response to prescribed changes in source emissions of NOX and VOCs.  They operate on sets
11     of input data that characterize the emissions, topography, and meteorology of a region and
12     produce outputs that  describe air quality in that region.
13
14          Two kinds of photochemical models are recommended in guidelines issued by EPA:
15     (1) the use of EKMA is acceptable  under certain circumstances, and (2) the grid-based Urban
16     Airshed Model (UAM) is recommended for modeling O3 over urban areas.  The  1990 Clean
17     Air Act Amendments mandate the use of three-dimensional (grid-based) air quality models
18     such as UAM in developing State Implementation Plans for areas designated as extreme,
19     severe, serious, or multistate moderate.  General descriptions of EKMA and of grid-based
20     models were given in the 1986 EPA criteria document for O3.
21
22          The EKMA-based method for determining O3 control strategies has some limitations,
23     the most serious of which is that predicted emissions reductions are critically dependent on
24     the initial NMHC/NOX ratio used in the calculations.  This ratio cannot be determined with
25     any certainty and is expected to be quite variable in an urban area.
26
27          Grid-based, photochemical air quality  models have their limitations as well. These are
28     pointed out subsequently.
29
30          Spatial and temporal characteristics of VOC and NOX emissions are major inputs to  a
31     grid-based photochemical air quality model. Greater accuracy in emissions inventories is
32     needed for biogenics and for both mobile and stationary source components. Grid-based air
33     quality models also require as input the three-dimensional wind field for the photochemical
34     episode being simulated.
35
36          A chemical kinetic mechanism (a set of chemical reactions), representing the important
37     reactions that occur  in the atmosphere, is used in an air quality model to estimate the net  rate
38     of formation of each pollutant simulated as  a function of time.
39
40          Dry deposition is an important removal process for O3 on both urban and regional
41      scales and is included in all urban- and regional-scale models. Wet deposition is generally
42      not included in urban-scale photochemical models,  since Oj episodes do not occur during
43      periods of significant clouds or rain.
44
45           Concentration  fields of all  species computed by the model must be specified at the
46      beginning of the simulation ("initial conditions").  These initial conditions are determined


        February 1994                            1-8       DRAFT-DO NOT QUOTE OR CITE

-------
 1     mainly with ambient measurements, either from routinely collected data or from special
 2     studies; but interpolation can be used to distribute the surface ambient measurements.
 3
 4     Use of Ozone Air Quality Models
 5
 6          Photochemical air quality models are used for control strategy evaluation by first
 7     demonstrating that a past episode, or episodes, can be adequately simulated and then
 8     reducing hydrocarbon or NOX emissions or both in the model inputs and assessing the effects
 9     of these reductions on O3 in the region. The adequacy of control strategies based on grid-
10     based models depends in part on the nature of input data for simulations and model
11     validation,  on input emissions inventory data, and on the mismatch between model output
12     and the current form of the NAAQS for O3.   Uncertainties in models can obviously affect
13     their outputs. Uncertainties exist in all components of grid-based O3 air quality models:
14     emissions,  meteorological modules, chemical mechanisms, deposition rates, and
15     determination of initial conditions.
16
17          Grid-based models that have been widely used to evaluate control strategies for O3 or
18     acid deposition, or both, are  (1) the Urban Airshed Model (UAM), (2) the California
19     Institute of Technology/Carnegie Institute of Technology (CIT) model, (3) the Regional
20     Oxidant Model (ROM), (4) the Acid Deposition and Oxidant Model (ADOM), and (5) the
21     Regional Acid Deposition Model  (RADM).
22
23     Conclusions
24
25          Urban air quality models are becoming readily available for application and have been
26     applied in recent years in several urban areas.  Significant progress has also been made in the
27     development of regional models and the integration of state-of-the-art prognostic
28     meteorological models as drivers.
29
30          Although there are still many uncertainties in photochemical air quality modeling, prime
31     among which are emission inventories, models are nevertheless essential for regulatory
32     analysis and constitute one of the major tools for attacking the O3 problem. Grid-based
33     O3 air quality modeling is  superior to the available alternatives for O3 control planning, but
34     the chances of its incorrect use must be minimized.
35
36
37     1.4   ENVIRONMENTAL  CONCENTRATIONS, PATTERNS, AND
38            EXPOSURE ESTIMATES
39
40          Ozone is measured at concentrations above the minimum detectable level at all
41     monitoring locations in the world. In this section of the document, hourly average
42     concentration and exposure information was  summarized for urban, rural forested, and rural
43     agricultural areas in the United States.
44
45          Because O3 produced from urban area  emissions is transported to more rural downwind
46     locations,  elevated O3 concentrations can occur at considerable distances from urban centers.


       February 1994                             1-9       DRAFT-DO NOT QUOTE OR CITE

-------
 1     Urban O3 concentration values are often depressed because of titration by NO.  Because of
 2     the absence of chemical scavenging, O3 tends to persist longer in nonurban than in urban
 3     areas and exposures may be higher than  hi urban locations.
 4
 5     Trends
 6
 7          Ozone hourly average concentrations have been recorded for many years by the state
 8     and local air pollution agencies who report their data to the EPA.  The 10-year (1983 to
 9     1992) composite average trend for the second highest daily maximum hourly average
10     concentration during the O3  season shows that the  1992 composite average  for the trend sites
11     is 21 % lower than the 1983  average.  The 1992 value is the lowest composite average of the
12     past 10 years.  The 1992 composite average is  significantly less than all the previous nine
13     years, 1983 to 1991.  The relatively high O3 concentrations in 1983 and 1988 were likely
14     attributable in part to hot, dry stagnant conditions in some areas of the country that were
15     especially conducive to  O3 formation.
16
17          Between 1991 and 1992, the composite mean of the second highest daily maximum
18     1-h O3 concentrations decreased 7% and the composite average  of the number of estimated
19     exceedances  of the O3 standard decreased by 23%. Nationwide VOC emissions decreased
20     3% between 1991 and 1992. The composite average of the second daily maximum
21     concentrations decreased in 8 of the 10 EPA Regions between 1991 and 1992, and remained
22     unchanged in Region VII. Except for Region Vn, the 1992 regional  composite means are
23     lower than the corresponding 1990 levels.
24
25     Surface Concentrations
26
27          Published data provides evidence showing the occurrence,  at some sites,  of multihour
28     periods  within a day of O3 at levels of potential health effects.  Although most of these
29     analyses were made using monitoring data collected from sites in or near nonattainment
30     areas, in one analysis of five sites (two in New York state, two in rural California, and one
31     in rural Oklahoma), none of which was  in or near a nonattainment area, O3 concentrations
32     showed only moderate peaks but showed multihour levels above 0.10 ppm.
33
34          On the basis of O3 data from isolated monitoring sites, the EPA has indicated that a
35     reasonable estimate of natural O3 background concentration near sea level in the United
36     States today, for an annual average, is from 0.020 to 0.035 ppm.  This estimate included  a
37     0.010- to 0.015-ppm contribution from the stratosphere and a 0.01-ppm contribution from
38     photochemically affected biogenic NMHCs.  An additional 0.010 ppm is possible from the
39     photochemical reaction  of biogenic methane. The EPA  concluded that a reasonable estimate
40     of natural O3 background concentration for a 1-h daily maximum at sea level in the United
41     States during the summer is on the order of 0.03 to 0.05 ppm.  This  estimate may be low,
42     however, because available  data from sites appearing to be isolated from anthropogenic
43     sources in the western United States indicate that maximum hourly concentrations can reach
44     0.06 and 0.075 ppm with infrequent occurrences below  0.02 ppm (i.e., lack of scavenging).
45
46


       February 1994                            1-10       DRAFT-DO  NOT QUOTE OR CITE

-------
 1     Diurnal Variations
 2
 3          Diurnal patterns of O3 may be expected to vary with location, depending on the balance
 4     among the many factors affecting O3 formation, transport, and destruction.  Although they
 5     vary with locality, diurnal patterns for O3 typically show  a rise in concentration from low or
 6     levels near minimum detectable amounts to an early afternoon peak.  The diurnal pattern of
 7     concentrations can be ascribed to three simultaneous processes:  (1) downward transport of
 8     O3 from layers aloft, (2) destruction of O3 through contact with surfaces and through
 9     reaction with NO at ground level, and (3) in situ photochemical production of O3.
10
11     Seasonal Patterns
12
13          Seasonal variations in O3 concentrations in urban areas usually show the pattern of high
14     O3 in late spring or in summer and low levels in the winter; however,  weather conditions in
15     a given year may be more favorable for the formation of  O3 and other oxidants than during
16     the prior or following year.
17
18          Average O3 concentrations tend  to be higher in the second versus  the third quarter of
19     the year for many isolated rural sites.  This observation has been attributed to either
20     stratospheric intrusions or an increasing frequency of slow-moving, high-pressure systems
21     that promote the formation of O3.  However, for several clean rural sites, the highest
22     exposures have occurred in the third quarter rather than in the  second.  For rural O3 sites in
23     the southeastern United  States, the daily maximum 1-h average concentration was found to
24     peak during the summer months.
25
26     Spatial Variations
27
28          Concentrations of O3 vary with altitude and with latitude. There appears  to be no
29     consistent conclusion concerning the relationship between O3 exposure and elevation.
30
31     Indoor Ozone
32
33          Until the early 1970s,  very little was known about the O3 concentrations experienced
34     inside buildings, and to  date, the database on this subject  is not large and a wide range of
35     indoor/outdoor O3 concentration relationships can be found in the literature;  reported
36     indoor/outdoor values for O3 are highly variable.  Indoor/outdoor O3 concentration ratios
37     generally fall in the range from 0.1 to 0.7 and indoor  concentrations of O3 will  almost
38     invariably be less than outdoors.
39
40     Estimating  Exposure
41
42           Both fixed-site monitoring information and human exposure models are used to
43     estimate risks associated with O3 exposure.  Because,  for most cases, it is not possible to
44     estimate population exposure solely from fixed-station data, several human exposure models
45     have been developed. Some of these  models include information on human activity patterns
46     (i.e., the microenvironments people visit and the times they spend there). These models also

       February 1994                            1_H       DRAFT-DO NOT QUOTE OR CITE

-------
 1     contain submodels depicting the sources and concentrations likely to be found in each
 2     microenvinonment, including indoor, outdoor, and in-transit settings.
 3
 4          Few data are available for individuals using personal exposure monitors.  Results from
 5     a pilot study demonstrated that fixed-site ambient measurements may not adequately represent
 6     individual exposures.  Models based on time-weighted indoor and outdoor concentrations
 7     explained only 40% of the variability in personal exposures.
 8
 9          Two distinct types of O3 exposure models exist: those that focus narrowly on predicting
10     indoor O3 levels and those that focus on predicting O3 exposures on a community-wide basis.
11
12     Peroxyacyl  Nitrates
13
14          Peroxyacetyl nitrate and peroxypropionyl nitrate (PPN) are the most abundant of the
15     non-O3 oxidants in ambient air in the United States, other than the inorganic nitrogenous
16     oxidants  such  as NO2, and possibly nitric acid (HNO3).  The concentrations of PAN that are
17     of most concern are those to which vegetation could potentially be exposed, especially during
18     daylight hours in agricultural areas.  Most of the available data on concentrations of PAN
19     and PPN in ambient air are from urban areas.  The levels to be found in nonurban areas will
20     be highly dependent upon the transport of PAN and PPN or their precursors from urban
21     areas because  the concentrations of the NOX precursors to these compounds are considerably
22     lower in nonurban than in urban areas.
23
24     Co-occurrence
25
26          Studies of the joint occurrence of gaseous NO2/O3 and SO2/O3 at rural sites have
27     concluded that the periods of co-occurrence represent a small portion of the potential plant
28     growing  period.  For human ambient exposure considerations, in most cases, the
29     simultaneous co-occurrence of NO2/O3 and SO2/O3 was infrequent.  Some researchers have
30     reported the joint occurrence of O3, nitrogen, and sulfur in  forested areas, combining
31     cumulative exposures of O3 with data on dry deposition of sulfur and nitrogen. One study
32     reported that several forest landscapes with the highest dry deposition loadings of sulfur and
33     nitrogen tended to experience the highest average O3 concentrations and largest cumulative
34     exposure.  Although the authors concluded that the joint concentrations of multiple pollutants
35     in forest landscapes were important, nothing was mentioned about the hourly co-occurrences
36     of O3 and SO2 or O3  and NO2.  Acid sulfates, which are usually composed of sulfuric acid,
37     ammonium bisulfate,  and ammonium sulfate,  have been measured at a number of locations in
38     North  America.  The potential for O3 and acidic sulfate aerosols to co-occur at some
39     locations in some form (i.e., simultaneously,  sequentially, or complex-sequentially) is real
40     and requires further characterization.  For human ambient exposures, the  simultaneous
41     co-occurrence of NO2 and O3 was infrequent.
42
43          In one study, the relationship between O3 and hydrogen ions in precipitation was
44     explored using data from sites that monitored both O3 and wet deposition  simultaneously and
45     within one minute latitude and longitude of each other.  It was reported that individual sites
46     experienced years in which both hydrogen ion deposition and total O3 exposure were  at least


       February 1994                            1_12      DRAFT-DO NOT QUOTE OR CITE

-------
 1     moderately high.  With data compiled from all sites, it was found that relatively acidic
 2     precipitation occurred together with relatively high O3 levels approximately 20%  of the time,
 3     and highly acidic precipitation occurred together with a high O3 level approximately 6% of
 4     the time.  Sites most subject to relatively high levels of both hydrogen ions and O3 were
 5     located in the eastern portion of the United States, often in mountainous areas.
 6
 7          The co-occurrence of O3 and acidic cloudwater in high-elevation forests has been
 8     characterized.  The frequent O3-only and pH-only single-pollutant episodes, as well as the
 9     simultaneous and sequential co-occurrences of O3 and acidic cloudwater, have been reported.
10     Both simultaneous  and sequential co-occurrences were observed a few times each month
11     above  cloud base.
12
13
14     1.5    ENVIRONMENTAL EFFECTS OF  OZONE AND RELATED
15            PHOTOCHEMICAL OXIDANTS
16
17          Ozone is the  gaseous pollutant most injurious to agricultural crops, trees, and native
18     vegetation.  Exposure of vegetation to O3 can inhibit photosynthesis, alter carbon
19     (carbohydrate) allocation, and interfere with mycorrhizal formation in tree roots. Disruption
20     of these important  physiological processes  can suppress the growth of crops, trees, shrubs,
21     and herbaceous vegetation by decreasing their capacity to form the carbon (energy)
22     compounds needed for growth and maintenance and their ability to absorb the water and
23     mineral nutrients they require from the soil. In addition, loss of vigor increases
24     susceptibility of trees and crops to insects and pathogens and impairs their ability to
25     reproduce.  The following section summarizes key environmental effects associated with
26     O3 exposure.
27
28     Effects on Agroecosystems
29
30     Methodologies Used in  Vegetation Research
31
32          Most of the knowledge concerning the effects of O3 on vegetation  comes from the
33     exposure-response  studies of important agricultural  crop plants and some selected forest and
34     urban  tree species, mostly as seedlings.  A variety of methodologies have been used, ranging
35     from field exposures without chambers to open-top chambers and to exposures conducted in
36     chambers under highly controlled climates. In general, the more controlled conditions  are
37     most appropriate for investigating specific responses and for providing the scientific basis for
38     interpreting and extrapolating results.
39
40     Mode  of Action
41
42          Photosynthesis provides plants with the energy and structural building blocks necessary
43     for their existence. The photosynthetic capacity of a  plant (i.e., carbohydrate production) is
44     an important aspect of plant response to stresses in natural environments and  is strongly
45     associated with leaf nitrogen content and with water movement.  Continued acquisition of
46     nitrogen and water uptake are necessary, if photosynthesis is to occur, and involves the


       February 1994                            1-13       DRAFT-DO NOT QUOTE OR CITE

-------
 1     allocation of carbohydrates from the leaves to the roots.  Leaf photosynthetic capacity is age
 2     dependent.  As the plant grows, the canopy structure changes, altering the amount and angle
 3     of light hitting a leaf.  Allocation of carbohydrates and nutrients to new leaves is especially
 4     important in stimulating growth production.
 5
 6           Ozone exposure results in a leaf-mediated plant response.  Leaves are important
 7     regulators of plant stress response and function.  As the major regulators of anatomical and
 8     morphological development of the shoot, they control the allocation of carbohydrates to the
 9     whole plant.  The many regulatory systems contained in leaves change both as a function of
10     leaf development and in response to different environmental stresses.
11
12           Only O3 that enters the plant through stomata (openings in the leaves) can impair plant
13     processes.   In addition, an effect will occur only if sufficient O3 reaches sensitive sites within
14     leaf cells.  Ozone injury will not be detected if (1) the rate of uptake  is small enough for the
15     plant to detoxify or metabolize O3 or its derivatives,  or (2) the plant is able to repair or
16     compensate for its impacts at a rate equal to  or greater than the rate of uptake.  (If the rate
17     of repair is very slow, there may be an effect.)
18
19           The uptake and movement of O3 to sensitive cellular sites are subject to various
20     biochemical and physiological  controls.  The magnitude of O3-induced effects depends upon
21     the physical environment of the plant, including both macro-and microclimatic factors, and
22     the chemical environment of the plant, including other gaseous pollutants and biological
23     factors.  Visible foliar injury is usually the first indication of cellular plant response.  Once
24     sufficient O3 enters the leaf, the plant can  experience reduced photosynthesis, altered
25     carbohydrate production and allocation, reduced plant vigor, reduced growth or yield or
26     both, and sometimes death. The alterations in the biochemical and physiological processes
27     mentioned above may occur with or without  visible injury to the plant.
28
29           Reductions in photosynthesis caused by O3 are likely to  be accompanied by a shift in
30     growth pattern that favors shoots. Plants compensate for differences  in resource imbalances
31     by allocating  energy, water, and minerals to  maintain optimal growth. Allocation of these
32     resources to leaf repair or to new leaf formation and an overall reduction in carbohydrate
33     formation decrease the  availability of carbohydrates for both stem and root growth.
34     Alteration of the normal allocation pattern affects all aspects of plant growth and
35     reproduction.
36
37     Factors That Modify Plant Response
38
39           Plant response to  O3 may be modified by a variety of biological, chemical, and physical
40     factors.  Both the impact of environmental factors on plant response to O3 and the effects  of
41     O3 on the responses of plants to environmental factors have to be considered when
42     determining the impact of oxidants on vegetation in the field.
43
44           Biological factors include those both within and external to the plant.  Within the plant,
45     its genetic composition and stage of development play critical roles in plant response to O3 as
46     well as other stresses.  Different varieties  or cultivars, as well as different individuals of a
        February 1994                             M4      DRAFT-DO NOT QUOTE OR CITE

-------
 1     species, due to differences in their genetic composition, are known to differ greatly in their
 2     responses to a given O3 exposure.
 3
 4          The magnitude of the response of a particular species or variety depends on such
 5     environmental factors as temperature and humidity, soil moisture and nutrition, exposure to
 6     other pollutants or agricultural spray chemicals, and interaction with plant pests and
 7     pathogens. In  other words, the plant's present and past environmental mileu, which includes
 8     the temporal exposure pattern and stage of development, dictates the plant response.  The
 9     corollary  is also true:  exposure to O3 can modify plant response to other environmental
10     variables.  For example, exposure to O3 reduces the ability of trees to withstand winter
11     injury caused by freezing temperatures and influences the success of pest and fungal
12     infections by decreasing  tree vigor.
13
14     Effects Based Air Quality Exposure Indices
15
16           Environmental scientists for many years have attempted to characterize and
17     mathematically represent plant exposures to O3.  A variety of averaging times have been
18     used to characterize exposure-response in plants.   Though most studies have characterized
19     exposure  by using mean concentrations such as seasonal, monthly, weekly, daily, or peak
20     hourly means,  other studies have used cumulative  measures (e.g., the number of hours above
21     selected concentrations).  None of these statistics completely  characterizes the relationships
22     among O3 concentration, exposure duration, interval between exposures, and plant response.
23
24           The use of a mean concentration with long averaging times (1) implies that all
25     concentrations of O3 are equally  effective in causing plant responses, and (2) minimizes the
26     contributions of the peak concentrations to the response.  Present evidence suggests that
27     higher concentrations, episodic peak occurrences,  and duration of exposure  have a relatively
28     greater role in producing plant effects than mean concentrations; therefore, an index that
29     cumulates hourly concentrations during the season and gives  greater weight to higher
30     concentrations appears to be a more appropriate index for relating ambient exposures to
31     growth or yield effects.  From the lexicological perspective,  it is the peak occurrences or
32     concentrations above an unidentified effect level that are most likely to have an impact.
33     Effects on vegetation appear when the amount of pollutant that has entered a plant exceeds its
34     ability to repair or compensate for the impact.
35
36           An  index of ambient exposure that relates well to plant response should directly or
37     indirectly incorporate  both environmental influences, (e.g., temperature, humidity, and
38      soil-moisture status) and exposure dynamics.
39
40           No  experimental studies have been designed specifically to evaluate the adequacy of the
41     various peak-weighted indices that have been proposed.  In retrospective analyses when O3  is
42     the primary source of variation in response, year-to-year variations in plant response are
43      minimized by  peak-weighted, cumulative exposure indices.  However, a number of different
44      forms of peak-weighted, cumulative indices have  been examined for their ability to properly
45      order yield responses  from the large number of studies of the National Crop Loss Assessment
46      Network (NCLAN) program.  These exposure indices (i.e., SUMOO, SUM06, SIGMOID,


        February 1994                            1-15       DRAFT-DO NOT QUOTE OR CITE

-------
 1     and W126) all performed equally well, and it is not possible to distinguish between them on
 2     the basis of statistical fits of the data.
 3
 4     Exposure-Response of Plant Species
 5
 6          The number of crop species/cultivars for which information regarding O3 exposure
 7     responses are available encompasses a mere fraction of the total of cultivated crops and even
 8     a smaller fraction of native plants.  The emphasis of experimental studies has usually been on
 9     the more economically important crop plants and tree species, as seedlings.
10
11          Crop species usually are monocultures that are fertilized and in many cases watered.
12     Therefore, because crop plants are usually grown under optimal conditions, their sensitivity
13     to O3 exposures undoubtedly varies from that of native trees, shrubs, and herbaceous
14     vegetation.
15
16          The concept of limiting values was used in both the 1978 and 1986 criteria documents
17     to summarize visible foliar injury.  Limiting values are defined as concentrations and
18     durations of exposure  below which visible injury does not occur. The limit for visible injury
19     indicating reduced plant  performance was an O3 exposure of 0.05 ppm for several hours per
20     day for greater than 16 days.  When the exposure period was decreased to  10 days, the
21     O3 concentration required to cause injury was increased to 0.1 ppm. A short, 6-day
22     exposure further increased the concentration to 0.30 ppm.  The exposure and concentration
23     periods apply today for those crops where appearance or aesthetic value (e.g, spinach,
24     cabbage, lettuce) is considered important. Limiting values for foliar injury to trees and
25     shrubs range from 0.06 to 0.1 ppm for 4 h.
26
27          Several conclusions were drawn in the 1986 criteria document from the various
28     experimental approaches used to estimate crop  yield loss.
29
30     1.  Ambient O3 concentrations are sufficiently  elevated in several regions of the country to
31         impair growth and yield of plants. This is  clearly indicated by comparison of data
32         obtained from crop yield in  charcoal-filtered and unfiltered (ambient) exposures.  These
33         elevated levels are further supported by  data from  studies using chemical protectants.
34         These response data  make possible extrapolation to plants not studied experimentally.
35         Both approaches mentioned  above indicate that effects occur with only a few exposures
36         above 0.08 ppm.
37
38     2.  Several plant species exhibited growth and  yield effects when the mean O3 concentration
39         exceeded 0.05 ppm for 4 to 6 h/day for at  least 2 weeks.
40
41     3.  Data from regression studies conducted  to develop an exposure-response function for
42         estimating yield loss indicated that at least 50% of the species/cultivars tested could be
43         predicted to exhibit a 10% yield loss at  7-h seasonal  mean O3 concentrations of 0.05 ppm
44         or less.
45
46           Based on research  published since the 1986 criteria document, the following
47     conclusions can be drawn.

       February 1994                            M6       DRAFT-DO NOT QUOTE OR CITE

-------
 1     1. Monitoring data from 80 to 200 nonurban sites for a period of 10 years were analyzed to
 2        determine the ambient 7-h growing-season average O3 concentrations.  For periods of
 3        3 and 5 mo,  concentrations ranged from 0.05 to 0.06 ppm and 0.047 to 0.054 ppm,
 4        respectively.
 5
 6     2. Open-top chamber studies comparing yields at ambient O3 exposures with those in
 7        filtered air, and retrospective analyses of crop data, indicate that current ambient
 8        O3 concentrations at some sites in the United States are sufficient to reduce the yield of
 9        major crops.   These results also indicate that visible injury that reduces the market value
10        of certain crops and ornamentals (spinach, petunia, geranium, for example) occurs at
11        O3 concentrations ranging from 0.04 to 0.10 ppm for 4 h.
12
13     3. A growing season SUMO6 exposure of 26.4 ppm-h, corresponding to a 7-h growing
14        season mean of 0.049 ppm, is estimated to prevent a 10% yield loss from O3 in 50%  of
15        the 54 experimental cases analyzed within this document.  It is estimated that a
16        12-h growing season mean of 0.045 ppm would restrict yield losses to 10% in major crop
17        species.
18
19     Effects on Natural Ecosystems
20
21          Most of the information dealing with the possible responses of ecosystems to O3 stress
22     is based on  the studies in the San Bernardino Forest ecosystem. This mixed conifer forest
23     ecosystem in southern California has been studied more than any other ecosystem in the
24     United States, over 20 years.  Chronic O3 exposures for a period of 50 or more years have
25     been associated with major changes hi the ecosystem.
26
27          Data from  an inventory conducted from 1968 through  1972 indicated that for 5 mo of
28     the year, trees were exposed to O3 concentrations greater than 0.08 ppm for more than
29     1,300 h. Concentrations rarely decreased below 0.05 ppm at night near the crest of the
30     mountain slope,  approximately 5,500 ft.  In addition, during the years 1973 to 1978, average
31     24-h O3 concentrations ranged from a background of 0.03 to 0.04 ppm in the eastern part of
32     the San Bernardino Mountains to a maximum of 0.10 to 0.12 ppm in the western part during
33     May through September.  Although PAN and NO2 were present in these studies,  symptoms
34     of PAN injury on forest herb-layer plant species could not be distinguished from those
35     caused by O3, and NO2 concentrations were too low to cause injury.
36
37          Ecosystem responses to stress are hierachial.  They begin with the response of the
38     most sensitive individuals of a population. Stresses, whose  primary effects occur at the
39     molecular level  (within the leaves), must be propagated progressively up through more
40     integrated levels of organ physiology (e.g., leaf, branch, root) to whole plant physiology, to
41     populations within the stand (community), and then to the landscape level to produce
42     ecosystem effects.
43
44          Therefore, to understand the effects of a stress, one must utilize a framework of
45     hierarchical scales.  This is particularly true of  low-level  stresses because  only a small
46     fraction of stresses at the molecular level become disturbances at the tree, stand, or landscape


       February 1994                            1-17      DRAFT-DO NOT QUOTE OR CITE

-------
 1      level.  Insect defoliation, for example, may severely reduce the growth of one or several
 2      branches, whereas growth of the tree appears not to be affected. The time required for a
 3      stress to be propagated from one level to the next determines how soon the effects of the
 4      stress can be observed or measured.
 5
 6          Foliar injury is usually the first visible indication of O3  exposure in plants.  In the San
 7      Bernardino forest, injury was first observed on ponderosa pine, with Jeffrey pine, white fir,
 8      black oak, incense cedar, and sugar pine following in decreasing order of sensitivity.  Foliar
 9      injury on sensitive ponderosa  and Jeffrey pine was observed when 24-h-average
10      O3 concentrations were 0.05 to 0.06 ppm.
11
12          The biochemical changes within the leaves  were expressed as (1) visible foliar injury;
13      (2) premature needle senescence; (3) a reduction in photosynthesis; (4) a reduction in
14      carbohydrate production; (5) altered carbohydrate allocation, reduced plant vigor; and (6) a
15      reduction in growth and reproduction.
16
17          Tree growth is the culmination of biochemical and physiological processes, during
18      which plants use, accumulate, and store carbon compounds (energy)  to build and maintain
19      their structure.  Within the leaves during the process of photosynthesis, carbon dioxide
20      absorbed from the atmosphere is converted to carbohydrates.  The water and minerals
21      necessary for growth are absorbed from the soil  through the roots.  Growth and seed
22      formation depend not only on photosynthesis and the uptake of water and mineral nutrients,
23      but also on subsequent metabolic processes and the allocation of carbohydrates to the rest of
24      the plant.  Impairment of any of these processes  can decrease plant vigor and affect plant
25      growth and reproduction.
26
27          Trees each year require  energy to grow new leaves and needles, to produce new fine
28      roots,  and to increase radial growth. Factors that inhibit photosynthesis and limit
29      carbohydrate formation shift carbohydrate allocation to new leaves, whereas factors that limit
30      nitrogen or water availability  will shift allocation to the roots.  Increased  carbohydrate
31      allocation to new needles and to repair foliage injured by O3  can place a drain on
32      carbohydrate reserves.  Reduced tree vigor, the result of altered carbohydrate allocation,
33      increases susceptibility of trees  to insect pests and fungal pathogens and reduces the
34      formation of mycorrhizae required for water and nutrient uptake.
35
36          Premature needle senescence, as observed in ponderosa and Jeffrey  pine, not only
37      decreases the amount of foliage available to carry on photosynthesis, but  alters
38      microorganismal succession on  conifer needles, thus altering the detritis-forming process and
39      subsequent nutrient  cycling.
40
41          The primary effect on the more susceptible members of the San Bernardino forest
42      community (e.g., ponderosa and Jeffrey pine) was that they were  no longer able to compete
43      effectively for essential nutrients, water,  light, and space.  As a consequence of altered
44      competitive conditions in the community, there was a decline in the sensitive species,
45      permitting the enhanced growth of more tolerant species.  In addition, changes in the
46      dominant individuals in the ecosystem altered the processes of energy flow and nutrient
47      cycling. These changes that began with the biochemical changes within the leaves of the

        February 1994                            l_lg       DRAFT-DO NOT QUOTE OR CITE

-------
 1      sensitive conifers were ultimately transferred upward to the whole ecosystem, from the
 2      individual, to the population, to the community, and finally, to the ecosystem.
 3
 4           Functional ecosystem changes associated with the deaths in the conifer populations
 5      altered the processes of carbohydrate (energy) flow, mineral nutrient cycling, and water
 6      movement.  The altered processes also changed, either directly or indirectly, the functioning
 7      of other living ecosystem components and ultimately changed the vegetational community
 8      patterns.  The continuum of changes beginning with injury to the individual trees resulted in
 9      ecosystem breakdown and a shift in dominance from ponderosa and Jeffrey pine to
10      O3-tolerant  shrub and oak species.
11
12           The effects that a stress at one level of organization will have on a higher level are
13      determined  by variability and compensation.  Variability in response to stress can mean that
14      not all trees are equally susceptible to O3, as  was observed,  in the San Bernardino Forest, in
15      the Appalachian Mountains, and the Cumberland Plateau of eastern Tennessee.  Individuals
16      of Ponderosa, Jeffrey, and eastern  white pine were designated as being sensitive,
17      intermediate or tolerant, based on their responses  to O3 exposure.  Compensation means that
18      plants tolerant to the stress are able either to detoxify or metabolize O3 entering the leaves or
19      to repair the injury.
20
21           Variability and compensation can also occur at the population level and, all populations
22      do not respond equally. Plant populations can respond in four different ways:  (1) no
23      response, the individuals are tolerant to the stress; (2) mortality of all individuals and local
24      extinction of the extremely sensitive population, the most severe response; (3) physiological
25      accomodation, resulting hi growth and reproductive success of tolerant individuals; and
26      (4) differential response, some individuals of  the population exhibit better growth and
27      reproductive success due to genetically determined traits.
28
29           The properties of variability and compensation at the individual level determine (1) the
30      severity of the response, (2) whether the  effect will be propagated to the next level, and
31      (3) the length of tune required for  the effect to be expressed functionally and structurally. In
32      most instances, the tune required for response at the individual level is a period of at least
33      3 to 5 years.  Variability and compensation at the population level determine whether there
34      will be an ecosystem response.   These properties  help to explain the differences in responses
35      between the San Bernardino Forest and the forest ecosystems in the eastern United States.
36
37           Forest stands differ greatly in age, species composition, stability, and capacity to
38      recover from disturbance.  In addition, the position in the stand or community of the most
39      sensitive species is extremely important.  Ponderosa and Jeffrey pine are controller species.
40      Their removal altered energy flow, nutrient cycling, and water movement in the San
41      Bernardino  forest ecosystem and changed the ecosystem. The removal of the eastern white
42      pine, on the other hand, did not alter the functional ecosystem processes sufficiently to bring
43      about a recognizable change within the ecosystems. For this reason, data dealing with the
44      responses of one forest type (e.g.,  San Bernardino) may not be applicable to other forest
45      types such as those found  in the Appalachian Mountains.
46
        February 1994                             1-19       DRAFT-DO NOT QUOTE OR CITE

-------
 1          Individual eastern white pine trees varied in sensitivity to O3, and, because of
 2     physiological accommodation and compensation, effects at the population level were not
 3     reported in the forest stands in the Appalachian Mountains.  Plantation plantings of eastern
 4     white pine (a monoculture) on the Cumberland Plateau, however, did exhibit overall growth
 5     reductions.
 6
 7          Mycorrhizae (fungus roots) function in the uptake of water and minerals (e.g., nitrogen,
 8     phosphorus) and are required for optimal growth by most plants.  Nitrogen is a critical
 9     element in the process of photosynthesis and is usually the element in shortest supply in both
10     agricultural and natural ecosystems.  Increased carbohydrate allocation to new needles and to
11     foliage repair limits allocation to the roots and reduces formation of the mycorrhizae needed
12     to take up water, nitrogen, and other nutrients.  Mycorrhizae can also protect plants from
13     attack by root pathogens.
14
15     Effects on  Materials
16
17           Over four decades of research show that O3 damages certain materials such as
18     elastomers, textile fibers, and dyes.  The amount of damage to actual in-use materials and
19     the economic consequences of that damage are poorly characterized.
20
21           Natural rubber and synthetic polymers of butadiene, isoprene, and styrene, used in
22     products like automobile tires and protective outdoor electrical coverings, account for most
23     of the elastomer production in the United States. The  action of 03 on these compounds is
24     well known,  and concentration-response relationships have been established and corroborated
25     by several studies.  These relationships, however, must be correlated with adequate exposure
26     information based on product use. For these and other economically important materials,
27     protective measures have been formulated to reduce the rate of oxidative damage. When
28     antioxidants and other protective measures are incorporated in elastomer production, the
29     O3-induced damage is reduced considerably, although the extent of reduction differs widely
30     according to the material and the type and amount of protective measures used.
31
32           Both the type of dye and the material in which it is incorporated  are important factors
33     in the resistance of a fabric to O3.  Some dyed fabrics, such  as royal blue rayon-acetate, red
34     rayon-acetate, and plum cotton, are resistant to O3.  On the other hand, anthraquinone dyes
35     on nylon fibers are sensitive to fading from O3.   Field studies and laboratory work show a
36     positive association between  O3 levels and dye fading of nylon materials.  At present, the
37     available research is insufficient to quantify the amount of damaged materials attributable to
38     O3 alone.
39
40           The degradation of fibers from exposure to O3 is poorly characterized.  In general,
41     most synthetic fibers like modacrylic and polyester are relatively resistant, whereas cotton,
42     nylon, and acrylic fibers have greater but varying sensitivities to  O3. Ozone reduces the
43     breaking strength of these fibers, and the degree of reduction depends  on the amount of
44     moisture present.  The limited research in this area indicates that O3 in ambient air may  have
45     a minimal effect  on textile fibers, but additional research is needed to verify this conclusion.
46
        February 1994                             1_20       DRAFT-DO NOT QUOTE OR CITE

-------
 1           A number of artists' pigments and dyes are sensitive to O3 and other oxidants.  Many
 2      organic pigments in particular are subject to fading or other color changes when exposed to
 3      O3. Although most, but not all, modern fine arts paints are more O3 resistant, many older
 4      works of art are at risk of permanent damage due to O3-induced fading.
 5
 6           A great deal of work remains to be done to develop quantitative estimates of the
 7      economic damage to materials from photochemical  oxidants.  Most of the available studies
 8      are now outdated in terms of the O3 concentrations, technologies, and supply-demand
 9      relationships that prevailed when the studies were conducted.  Additionally, little is known
10      about the physical damage functions, so cost estimates have been simplified to the point of
11      not properly recognizing many of the scientific complexities of the impact of O$.
12
13
14      1.6   TOXICOLOGICAL EFFECTS OF OZONE
15
16      Respiratory Tract  Effects of Ozone
17
18      Biochemical Targets of Ozone Interaction
19
20           Knowledge of molecular targets provides a basis for understanding mechanisms of
21      effects and strengthening animal-to-human extrapolations.  Ozone reacts with polyunsaturated
22      fatty acids and sulfhydryl, amino, and some electron-rich compounds.  These elements are
23      shared across  species.  Several types of reactions are  involved, and free radicals may be
24      created.  Based on this knowledge, it has been hypothesized that the  O3 molecule is unlikely
25      to penetrate the liquid linings of the respiratory tract to  reach the tissue, raising the
26      possibility that reaction products exert effects.
27
28      Lung Inflammation and Permeability Changes
29
30           Ozone disrupts the barrier function of the lung,  resulting in (1) the entry of compounds
31      in the airspaces into the blood and (2) the entry of serum components (e.g., protein) and
32      white blood cells (especially polymorphonuclear leukocytes) into the  airspaces and lung
33      tissue.  This latter impact reflects the initial stage of inflammation.  These cells can release
34      biologically active mediators that are capable of a number of actions, including damage to
35      other cells in the  lung.  In lung tissue, this inflammation can also increase the thickness of
36      the air-blood barrier.
37
38           Increases in permeability and inflammation have been observed at levels as low as
39      0.1 ppm (2 h/day, 6 days; rabbits).  After acute exposures, the influence of the time of
40      exposure (from 2 h to several hours) increases as the  concentration of O3 increases.
41      Long-term exposure effects are discussed under lung morphology.
42
43           The impacts of these changes are not fully understood.  At higher O3 concentrations
44      (e.g., 0.7 ppm, 28 days),  the diffusion of oxygen into the blood decreases, possibly because
45      the air-blood barrier is thicker; cellular death may result from the enzymes released by the
46      inflammatory cells; and host defense functions may be altered by mediators.


        February 1994                           1-21       DRAFT-DO NOT QUOTE OR CITE

-------
 1     Effects on Host Defense Mechanisms
 2
 3           Ozone exposure results in alterations of all defense mechanisms of the respiratory tract,
 4     including mucociliary and alveolobronchiolar clearance, functional and biochemical activity
 5     of the alveolar macrophage, and immunologic competence.  These effects can cause
 6     susceptibility to bacterial respiratory infections.
 7
 8           Mucociliary clearance, which removes particles and cellular debris from the conducting
 9     airways, is slowed by acute, but not repeated exposures to O3.  Ciliated epithelial cells that
10     move the mucous blanket are altered or destroyed by acute and chronic exposures.  Neonatal
11     sheep exposed to O3 do  not have normal development of the mucociliary system.   Such
12     effects could prolong the retention of unwanted substances (e.g., inhaled particles) in the
13     lungs, allowing them to  exert their toxic potential for a longer period of time.
14
15           Alveolar clearance mechanisms, which center on  the functioning of alveolar
16     macrophages, are altered by O3.  Short-term exposure  to levels as low as 0.1 ppm (2 h/day,
17     1 to 4 days; rabbits) accelerates clearance, but longer exposures do not.  Even so, after a
18     6-week exposure of rats to an urban pattern of O3,  the retention of asbestos fibers in a region
19     protected by alveolar clearance is prolonged.
20
21           Alveolar macrophages engulf and kill microbes, as well as clear the deeper regions of
22     the lungs of nonviable particles.  They also participate  in immunological responses, but little
23     is known about the effects of O3 on this function. Acute exposures of rabbits to levels as
24     low as 0.1 ppm decrease the ability of alveolar macrophages to ingest particles.  This effect
25     is displayed in decreases in the ability of the lung to kill bacteria after  acute exposure of
26     mice to levels as low as 0.4 ppm.
27
28           Both the pulmonary and systemic immune system are affected by O3,  but in a poorly
29     understood way.  It appears that the part of the immune system dependent on T-cell function
30     is more affected than that part dependent on B-cell  function.
31
32           Dysfunction of host defense systems results in enhanced susceptibility to bacterial lung
33     infections.  For example, exposure as low as 0.08 ppm for 3 h can overcome the ability of
34     mice to resist infection with streptococcal bacteria,  resulting in mortality.  However,
35     prolonged exposures (weeks, months) do not cause  greater effects on infectivity.
36
37           Effects on antiviral defenses are more complex and less well understood.  Only high
38     concentrations (1.0 ppm, 3 h/day, 5 days; mice)  increase viral-induced mortality.
39     Apparently, O3 does not impact antiviral clearance  mechanisms.  Although O3 does not affect
40     acute lung injury from influenza virus infection, it does enhance later phases  of the course of
41     an infection (i.e., postinfluenzal alveolitis).
42
43     Morphological Effects
44
45           Ozone causes similar types of alterations in lung  structure in all laboratory animal
46     species studied, from rats to monkeys.  In the lungs, the most affected cells are the ciliated


       February 1994                            1-22      DRAFT-DO NOT  QUOTE OR CITE

-------
 1      epithelial cells of the airways and Type 1 epithelial cells of the gas exchange region.  In the
 2      nasal cavity, ciliated cells are also affected.
 3
 4           The centriacinar region (the junction of the conducting airways and gas exchange
 5      regions) is the primary target, possibly because this area is predicted to receive the greatest
 6      dose of O3.  The ciliated cells can be killed and are replaced by nonciliated cells (i.e., cells
 7      not capable of clearance functions that also have increased ability to metabolize some foreign
 8      compounds).  Mucous-secreting cells are affected, but to a lesser degree.  Type 1 cells,
 9      across which gas exchange occurs, can be killed.  They are replaced by Type 2 cells, which
10      are thicker and produce more lipids.  An inflammatory response also occurs in the tissue.
11      The tissue is thickened further in later stages when collagen (a structural protein increased in
12      fibrosis) and other elements accumulate.
13
14           The distal airway is remodeled.  More specifically, bronchiolar epithelium replaces the
15      cells present in alveolar ducts. Concurrent inflammation may play a role.  This effect has
16      been observed at 0.25 ppm (8 h/day, 18 mo) in monkeys; at a higher concentration, this
17      remodeling  persists after exposure stops.
18
19           The progression  of effects during and after a chronic exposure is complex.  Over the
20      first few days of exposure, inflammation peaks and then drops considerably, plateauing for
21      the remainder of exposure, after which it largely disappears.  Epithelial hyperplasia increases
22      rapidly over the first few days and rises  slowly or plateaus thereafter; when exposure ends, it
23      begins to return toward normal.  In contrast, fibrotic  changes  in the tissue between the air
24      and blood increase very slowly over months of exposure, and after exposure ceases, the
25      change sometimes persists or increases further.
26
27           The pattern of exposure can make a major difference in  effects.  Monkeys exposed to
28      0.25 ppm O3  (8 h/day) every other month of an 18-mo period had equivalent changes in lung
29      structure, more fibrotic changes, and more of certain types of pulmonary function changes
30      than monkeys exposed every day over the 18 mo.  From this  work and rat studies, it appears
31      that natural seasonal patterns may be of more concern than more continuous exposures.
32      Thus, long-term animal studies with uninterrupted exposures may underestimate some of the
33      effects of O3.
34
35           There is no evidence that O3 causes emphysema.
36
37      Effects on Pulmonary Function
38
39           Pulmonary function changes in animals resemble those observed in humans after acute
40      exposure.
41
42           During acute exposure, the most commonly observed alteration is an increased
43      frequency of breathing and decreased tidal volume (i.e., rapid, shallow breathing). This has
44      been reported at exposures as low as 0.2 ppm for 3 h (rats).   Typically,  higher
45      concentrations (around 1 ppm) are required to affect breathing mechanics (compliance and
46      resistance).  Extended characterizations of pulmonary function show types of changes


        February 1994                            1_23       DRAFT-DO NOT QUOTE OR CITE

-------
 1     generally seen in humans.  For example, there are decreased lung volumes at levels
 2     >0.5 ppm (a few hours; rats).
 3
 4          When rats are exposed to O3 for 2 h/day for 5 days, the pattern of attenuation of
 5     pulmonary function responses is similar to that observed in humans.  Other biochemical
 6     indicators of lung injury did not return to control values by Day 5, and morphological
 7     changes  increased in severity over the period of exposure.  Thus, attenuation did not result in
 8     protection against all the effects of O3.
 9
10          Long-term exposures have provided mixed results on pulmonary function, including no
11     or minimal effects, restrictive effects, or obstructive effects. When changes occurred and
12     postexposure examinations were performed, pulmonary function recovered.
13
14     Biochemical Effects
15
16          In  acute and short-term exposure studies, a variety of lung lipid changes occur,
17     including an increase in arachidonic acid, the further metabolism of which produces a variety
18     of biologically  active mediators that can affect host defenses, lung function,  the immune
19     system,  and other functions.
20
21          The level of lung antioxidant metabolism increases after O3 exposure, probably as a
22     result of the increase in the number of cells (Type 2 cells) rich in antioxidant enzymes.
23
24          Collagen  (the structural protein  involved in fibrosis) increases in O3-exposed lungs in a
25     manner  that has been correlated to structural changes (e.g.,  increased thickness of the tissue
26     between the air and blood after prolonged exposure).   Some, but not all,  studies found that
27     the increased collagen persists after exposure ceases.
28
29          Generally, O3 enhances lung xenobiotic metabolism after  both short- and long-term
30     exposure, possibly as a result of morphological changes (increased numbers of nonciliated
31     bronchiolar epithelial cells).  The  impact of this change is dependent on the xenobiotics
32     involved.  For example, the metabolism of benzo[fl]pyrene to active metabolites was
33     enhanced by O3.
34
35     Genotoxictiy and Carcinogenicity of Ozone
36
37          The chemical reactivities of  O3  give it the potential to be  a genotoxic agent.
38
39          In vitro studies are difficult to interpret because the culture systems used allowed the
40     potential formation of artifacts,  and often high or very high concentrations of O3 were used.
41      Generally, in these studies, O3 causes DNA strand breaks, sometimes is weakly mutagenic,
42      and causes cellular transformation and chromosomal breakage.   The latter finding has been
43      investigated in vivo with mixed results hi animals. A well-designed human clinical study
44      found no such effect.
45
46           The few  earlier long-term carcinogenic studies, with or without coexposure to known
47      carcinogens, are either negative or ambiguous.

        February 1994                            1-24      DRAFT-DO NOT QUOTE  OR CITE

-------
 1           The National Toxicology Program (NTP) completed chronic rat and mouse cancer
 2      bioassays using commonly accepted experimental approaches and designs.  Both males and
 3      females were studied.  Animals were exposed for 2 years (6 h/day, 5 days/week) to 0.12,
 4      0.5, and 1.0 ppm O3 or for a lifetime to the same levels (except 0.12 ppm).  Following their
 5      standard procedures  for determination of weight-of-evidence for carcinogenicity, the NTP
 6      reported "no evidence" in rats,  "equivocal evidence" in male mice,  and "some evidence"  in
 7      female mice.  The increases in adenomas and carcinomas were observed only in the lungs.
 8      There was no concentration response.   One of the reasons for the designation of "some
 9      evidence" in female  mice was that when the 2-year and lifetime exposure studies were
10      combined, there was a statistically significant increase in total tumors at 1.0 ppm.  It is not
11      justified to extrapolate these mouse data to humans at the present time because rats were not
12      affected, only a high level of O3 (1.0 ppm) caused a limited degree of carcinogenic activity
13      in one strain of mice, there was no concentration response,  and there is inadequate
14      information to provide a mechanistic support for the finding in mice.
15
16           In a companion NTP study,  male rats were treated with a tobacco carcinogen and
17      exposed for 2 years to 0.5 ppm O3. Ozone did not affect the response and therefore had no
18      tumor promoting activity.
19
20      Systemic Effects  of Ozone
21
22           Ozone causes a variety of effects on tissues/organs distant from the lung.  Because
23      O3 itself is not thought to penetrate the lung, these systemic effects  are either secondary to
24      lung alterations or result from reaction products of O3. Effects have been observed on
25      clinical chemistry, white blood cells, red blood cells, the circulatory system, the liver,
26      endocrine organ(s),  and the central nervous system.  Most of these effects cannot be
27      adequately interpreted at this time and have not been investigated in humans, but it is  of
28      interest to note that O3 exposures  causing effects on the respiratory  tract of animals cause a
29      wide array of effects on other organs also.
30
31           Several behavior changes occur in response to O3.  For example, 0.12 ppm (6 h, rats)
32      decreases wheel-running activity, and 0.5 ppm (1 min) causes mice to avoid exposure.
33      These effects are not fully understood,  but they may be related to lung irritation or decreased
34      ability to exercise.
35
36           Although cardiovascular effects,  such as slowed heart rate and decreased blood
37      pressure, occur in O3-exposed rats, some observed interactions with thermoregulation  prevent
38      qualitative extrapolation of these effects to humans at this time.
39
40           Developmental toxicity studies in pregnant rats summarized in the 1986 O3 criteria
41      document showed that levels up to about 2.0 ppm did not cause birth defects.  Rat pups from
42      females exposed to 1.0 ppm O3 during certain periods of gestation weighed less or had
43      delays in development of behaviors (e.g., righting, eye opening). No "classical"
44      reproductive assays with O3 were found.
45
       February  1994                            1_25       DRAFT-DO NOT QUOTE OR CITE

-------
 1          Other studies have indicated that O3 can affect some endocrine organs (i.e., pituitary-
 2     thyroid-adrenal axis  and parathyroid gland).  It appears that the liver has less ability to
 3     detoxify drugs after  O3 exposure, but assays of liver enzymes involved in xenobiotic
 4     metabolism are not consistent with each other.
 5
 6     Interactions of Ozone with Other  Co-occurring Pollutants
 7
 8          Animals studies of the effects of O3 in combination with other air pollutants show that
 9     antagonism, additivity, and synergism can result, depending upon the animal species,
10     exposure regimen, and health endpoint.  Thus, they clearly demonstrate the major
11     complexities and potential importance of interactions, but do not provide a scientific basis for
12     predicting the results of interactions under untested ambient exposure scenarios.
13
14
15     1.7  HUMAN HEALTH EFFECTS OF OZONE AND RELATED
16           PHOTOCHEMICAL OXIDANTS
17
18          This section summarizes key health effects associated with exposure to O3, the major
19     component of photochemical oxidant air pollution that is clearly of most health concern to the
20     human population.  Another, often co-occurring photochemical  oxidant component of "smog"
21     is PAN, but this compound has been demonstrated to be primarily responsible for induction
22     of smog-related eye  irritation (stinging of eyes).  Limited pulmonary function studies have
23     shown no effects of PAN at concentrations below 0.13 to 0.30 ppm, which are much higher
24     than the generally encountered ambient air levels in most cities.
25
26     Controlled Human Studies of Acute Ozone Effects
27
28     Effects on Lung Function
29
30          Controlled studies in healthy adult subjects have demonstrated O3-induced decrements
31     in pulmonary function, characterized by alterations in lung volumes and  flow, airway
32     resistance, and airway responsiveness.  Respiratory symptoms,  such as cough and pain on
33     deep inspiration, are associated with these changes  in lung function.
34
35          Ozone-induced decreases in lung volume, specifically forced vital capacity (FVC) and
36     forced expiratory volume in 1 s (FEVj),  can be largely attributed to decreases in inspiratory
37     capacity, or the ability to take a deep breath; although at higher exposure concentrations,
38     there is clearly an additional component that is not volume dependent. They recover to a
39     large extent within 2 to 6 h; normal baseline function is typically reestablished within 24 h,
40     but not fully with more severe exposures.
41
42          Ozone causes increased airway resistance and may cause reductions in expiratory flow
43     and the FEV^FVC  ratio.
44
        February 1994                           1-26      DRAFT-DO NOT QUOTE OR CITE

-------
 1          Ozone causes an increase in airway responsiveness to nonallergenic stimuli (e.g.,
 2     histamine or methacholine) in healthy and asthmatic subjects. There is no clear evidence of a
 3     relationship between O3-induced lung volume changes and changes in airway responsiveness.
 4
 5     Inflammation and Host Defense Effects
 6
 1          Controlled studies in healthy  adult subjects also indicate that O3 causes an inflammatory
 8     response in the lungs characterized by elevated levels of neutrophils (a type of white blood
 9     cell), increased epithelial permeability, and elevated levels of biologically active substances
10     (e.g., prostaglandins, proinflammatory mediators, and cytokines).
11
12          Inflammatory responses to O3 can be detected within 1 h after a single 1-h exposure
13     with exercise to concentrations >0.30 ppm; the increased levels of some inflammatory cells
14     and mediators persist for at least 18 h.  The temporal response profile is not adequately
15     defined, although it is clear that the tune course of response varies for different mediators
16     and cells.
17
18          Lung function and respiratory symptom responses to O3 do not seem to be correlated
19     with airway inflammation.
20
21          Ozone also causes inflammatory responses in the nose, marked by increased numbers of
22     neutrophils (neutrophilia) and protein levels  suggestive of increased permeability.
23
24          Alveolar macrophages removed from the lungs of human subjects after 6.6 h exposure
25     to 0.08 and 0.10 ppm O3 have a decreased ability to ingest microorganisms, indicating some
26     impairment of host defense capability.
27
28     Ozone Exposure-Response Relationships
29
30          Functional, symptomatic, and inflammatory responses to O3 increase with increasing
31     exposure dose of O3. The major determinants of the exposure dose are O3 concentration
32     (C), exposure duration  (T), and the amount  of ventilation (VE).  Of these factors, C has
33     more influence on the response than T or VE.
34
35          Exercise increases response to O3 by increasing VE (greater mass delivered), tidal
36     volume, inspiratory flow (greater percentage delivery), and the intrapulmonary
37     O3 concentration.
38
39          Repeated daily exposures to  relatively  high levels of O$ doses (C x T x VE)  causing
40     substantial reductions in FEVj (>20%  decrement) typically cause exacerbation of the lung
41     function and respiratory symptom  responses on the second exposure day.  However,
42     attenuation of these responses occurs with continued exposures for a few days. Most
43     inflammatory responses also  attenuate; for example, neutrophilia is absent after five
44     consecutive exposures.
45
46           Multihour exposures  (e.g., for up to 7 h) to O3 concentrations as low as  0.08 ppm
47     cause small but significant (1) decrements in lung function, (2) increases in respiratory

       February 1994                             1-27       DRAFT-DO NOT QUOTE OR CITE

-------
 1     symptoms, and (3) increases in neutrophils and protein levels.  Ozone C is a more important
 2     factor than exercise VE or T in predicting responses to multihour low-level O3 exposure.
 3     There is now clear evidence of a response plateau in terms of lung volume response to
 4     prolonged O3 exposure.  This evidence suggests that for a given combination of exercise and
 5     O3 concentration (i.e., dose rate), there is a response plateau; continued exposure (i.e,
 6     increased T) at that dose rate will not increase response.  Therefore, quantitative
 7     extrapolation of responses to longer exposure durations is not valid.
 8
 9     Mechanisms of Acute Pulmonary Responses
10
11           The mechanisms leading to the observed pulmonary responses induced by O3 are
12     beginning to be better understood. The available descriptive data suggest a number of
13     mechanisms leading to the alterations in lung function and respiratory symptoms, including
14     (1) O3 delivery to the tissue (i.e., the inhaled concentration, breathing pattern, and airway
15     geometry; (2) O3 reactions with the airway lining fluid and/or epithelial cell membranes;
16     (3) local tissue  responses, including injury and inflammation; and (4) stimulation of neural
17     afferents (bronchial C fibers) and the resulting reflex responses and symptoms.  The
18     cyclooxygenase inhibitors block production of prostaglandin £2 and interleukin-6 as well as
19     reduce lung volume responses;  however,  these drugs do not reduce neutrophilic inflammation
20     and levels of cell damage markers such as lactate dehydrogenase.
21
22     Effects on Exercise Performance
23
24           Maximal oxygen uptake, a measure of peak exercise performance capacity, is reduced
25     hi healthy young adults if preceded by O3 exposures sufficient to cause marked changes in
26     lung function (i.e., decreases of at least 20%) and increased subjective symptoms of
27     respiratory discomfort.  Limitations in exercise performance may be related to increased
28     symptoms, especially those related to breathing discomfort.
29
30     Factors Modifying Responsiveness to Ozone
31
32           Many variables have the potential for influencing responsiveness to O3; most, however,
33     are inadequately addressed in the available clinical data to make definitive conclusions.
34
35           Active smokers  are less responsive to O3 exposure, which may reverse following
36     smoking cessation, but these results should be interpreted with caution.
37
38           The possibility of age-related differences in response to O3 has been explored, although
39     young adults historically have provided the subject population for controlled human studies.
40     Children and adolescents have similar lung volume responses  to O3 as young adults, but lack
41     respiratory symptoms at levels to which they have been  exposed.  Pulmonary function
42     responsiveness  in adults appears to decrease with age, whereas symptom rates remain similar
43     to young adults. Group mean lung function responses of adults over 50 years of age are less
44     than those  of children, adolescents,  and young adults.
45
46           The available data have not conclusively demonstrated that men and women respond
47     differently to O3.  Likewise, pulmonary function responses of women have been compared

       February 1994                            1-28        DRAFT-DO NOT QUOTE OR CITE

-------
 1     during different phases of the menstrual cycle, but the results are conflicting.  If gender
 2     differences exist for lung function responsiveness to O3, they are not based on hormonal
 3     changes, differences in lung volume, or the ratio of FVC to VE.
 4
 5          There is no compelling evidence, to date, suggesting that any ethnic or racial groups
 6     have a different distribution of responsiveness to O3.
 7
 8          Seasonal and ambient factors may vary  responsiveness to O3, but further research is
 9     needed to determine how they affect individual subjects.  Individual sensitivity to O3 may
10     vary throughout the year, related to seasonal  variations in ambient O3 concentrations.
11
12          The specific inhalation route appears to be of minor importance in exercising adults.
13     Exposure to O3 by  oral breathing (i.e., mouthpiece) yields similar results as oronasal
14     breathing (i.e., chamber exposures).
15
16     Population Groups at Risk from Ozone Exposure
17
18          Population groups that have demonstrated responsiveness to ambient concentrations  of
19     O3 consist of exercising healthy and asthmatic individuals,  including children, adolescents,
20     and adults.
21
22          Available evidence from controlled human studies on subjects with preexisting disease
23     suggests that (1) mild asthmatics have  similar lung volume responses, but greater airway
24     resistance responses to O3 than nonasthmatics; and (2) moderate asthmatics may have, in
25     addition, greater lung volume responses than nonasthmatics.
26
27          Of all the other population groups studied, those with preexisting limitations in
28     pulmonary function and exercise capacity (e.g., chronic obstructive pulmonary disease,
29     chronic bronchitis,  ischemic  heart disease) would be of primary concern in evaluating the
30     health effects of O3. Unfortunately, limitations of (1) subject selection, (2)  standardized
31     methods of subject characterization,  and (3) range of exposure hamper the ability to  make
32     definitive conclusions regarding the relative responsiveness of most chronic  disease subjects.
33
34     Effects of Ozone Mixed with Other Pollutants
35
36          No significant enhancement of respiratory  effects has been consistently demonstrated
37     for simultaneous exposures of O3 mixed with SO2, NO2, sulfuric or nitric acid, paniculate
38     aerosols,  or with multiple combinations of these pollutants. It is fairly well established that
39     simultaneous exposure of healthy adults and  asthmatics to mixtures of O3 and other pollutants
40     for short periods of time (< 2 h) induces pulmonary function responses not  significantly
41     different from those following O3 alone when studies are conducted at the same
42     O3 concentration.  Exposure to PAN has been reported to  induce greater pulmonary function
43     responses than exposure to O3 alone, but at PAN concentrations (>0.27 ppm) much higher
44     than ambient levels.  Unfortunately, only a limited number of pollutant combinations and
45     exposure protocols have been investigated, and  subject groups are small and are
46     representative of only small portions of the general population.  Thus, much is unknown
47     about  the relationships between  O3  and the complex mix of pollutants found in the ambient air.

       February 1994                             1-29      DRAFT-DO NOT QUOTE  OR CITE

-------
 1          Prior exposure to O3 in asthmatics may cause an increase in response to other pollutant
 2     gases,  especially SO2.  Likewise, prior exposure to other pollutants can enhance responses to
 3     O3 exposure.
 4
 5     Controlled Human Studies of Ambient Air Exposures
 6
 7          Mobile laboratory studies of lung function and respiratory symptoms in a local subject
 8     population exposed to ambient photochemical oxidant pollution provide quantitative
 9     information on exposure-response relationships for O3.  A series of these studies from
10     Los Angeles, CA, have demonstrated pulmonary function decrements at O3 concentrations of
11     0.14 ppm in exercising healthy adolescents, and increased respiratory symptoms and
12     pulmonary function decrements at 0.15 ppm in heavily exercising athletes and at 0.17 ppm in
13     lightly exercising healthy and asthmatic subjects.  Comparison of the observed effects in
14     exercising athletes with controlled chamber studies at comparable O3 concentrations showed
15     no significant differences in lung function and symptoms, suggesting that coexisting ambient
16     pollutants have a minimal contribution to the measured responses under typical summer
17     ambient conditions in Southern California.
18
19     Field and Epidemiology Studies of Ambient Air  Exposures
20
21          Individual-level field studies and aggregate level time-series studies have addressed the
22     acute effects of O3 on lung function decrements and increased morbidity and mortality in
23     human populations exposed to real-world conditions of O3 exposure.
24
25          Camp and exercise studies of lung function provide quantitative information on
26     exposure-response relationships linking lung function declines with O3 exposure occurring in
27     ambient air.  Combined statistical analysis of six recent camp studies in children yields an
28     average relationship between decrements in FEVl and previous-hour O3 concentration of
29      —0.64 mL/ppb. Two key studies of lung function measurements before and after well-
30     defined outdoor exercise events in adults have yielded exposure-response slopes of
31      -0.40 and  -1.35 mL/ppb.  The magnitude of pulmonary function declines with
32     O3 exposure is consistent with the results of controlled human studies.
33
34          Daily life studies support a consistent relationship between O3 exposure and acute
35     respiratory morbidity in the population.  Respiratory symptoms (or exacerbation of asthma)
36     and decrements in peak expiratory flow rate are associated with increasing ambient O3,
37     particularly in asthmatic children; however, concurrent temperature, particles, acidity
38     (hydrogen ions), aeroallergens, and asthma severity or medication  status may also contribute
39     as independent or  modifying factors.  Aggregate results show greater responses in asthmatic
40     individuals than in nonasthmatics, indicating that asthmatics constitute a sensitive group in
41     epidemiologic studies of oxidant air pollution.
42
43          Summertime daily hospital  admissions for respiratory causes in various locations of
44     eastern North America have consistently shown a relationship with ambient levels of O3,
45     accounting for approximately one to three  excess respiratory hospital admissions per hundred
46     ppb O3 per million persons.  This association has been shown to remain even after


       February  1994                            1-30      DRAFT-DO NOT QUOTE OR CITE

-------
 1     statistically controlling for the possible confounding effects of temperature and copollutants
 2     (e.g., hydrogen ions, sulfate, and particles less than 10 jwm), as well as when considering
 3     only concentrations below 0.12 ppm O3.
 4
 5          Two recent time-series epidemiologic studies indicate a small but statistically significant
 6     association between O3 and total daily mortality in Los Angeles, CA, and New York City,
 7     NY, where peak 1-h maximum O3 reached concentrations greater than 0.2 ppm during the
 8     study period. The results were significant even after controlling for the potentially
 9     confounding effects of temperature and particles.  A third study in regions with lower
10     (<0.15 ppm) maximum 1-h O3 concentrations (St. Louis, MO, and Kingston-Harriman, TN)
11     did not detect a significant O3 association with mortality.
12
13          Only suggestive epidemiologic evidence exists for health effects of chronic ambient
14     O3 exposure in the population. All of the available studies of chronic respiratory system
15     effects in exposed children and adults are limited by a simplistic assignment of exposure or
16     by their inability to isolate potential effects related to O3 from those of other pollutants,
17     especially particles.
18
19
20     1.8   EXTRAPOLATION OF ANIMAL TOXICOLOGICAL DATA TO
21            HUMANS
22
23          There have been significant advances in O3 dosimetry since 1986 that better enable
24     quantitative extrapolation with marked reductions in uncertainty.  Experiments and models
25     describing the uptake efficiency and delivered dose of O3 in the respiratory tract (RT) of
26     animals and humans are beginning to present a clearer picture than has previously existed.
27
28          The total RT uptake efficiency of rats at rest is approximately 50%. Within the RT of
29     the rat, 50% of the O3 taken up by the RT is removed in the head, 7% in the larynx/trachea,
30     and 43% in the lungs.
31
32          In humans at rest, the total RT uptake  efficiency is between 80 and 95%. Total RT
33     uptake efficiency falls as flow increases. As tidal volume increases, uptake efficiency
34     increases and flow dependence lessens.  Pulmonary function response data and O3 uptake
35     efficiency data in humans generally indicate that the mode of breathing (oral versus nasal
36     versus oronasal) has little effect on upper RT or on total RT uptake efficiency, though one
37     study suggests that the nose has a higher uptake efficiency than the mouth.
38
39          When all of the animal and human in vivo O3 uptake efficiency data are compared,
40     there is a good degree  of consistency across data sets.  This agreement raises the level of
41     confidence with which these data sets can be used to support dosimetric model formulations.
42
43           Several mathematical  dosimetry models have been developed since  1986.  Generally,
44     the models predict that net  O3 dose to lung lining fluid plus tissue gradually decreases
45     distally from the trachea toward the end of the tracheobronchial region, and then rapidly
46     decreases in the pulmonary region.


       February  1994                           1-31       DRAFT-DO NOT QUOTE OR CITE

-------
 1          When the dose of O3 to lung tissue is computed theoretically, it is found to be very low
 2     in the trachea, to increase to a maximum in the terminal bronchioles of the first generation
 3     pulmonary region,  and then to decrease rapidly, moving further into the pulmonary region.
 4     The increased tidal volume and flow, associated with exercise in humans,  shifts O3 dose
 5     further into the periphery of the lung and causes a disproportionate increase in distal lung
 6     dose.
 7
 8          Preditions of delivered dose have been used to investigate O3 responses in the context
 9     of intra- and interspecies comparisons.  In the case of intraspecies comparisons, for example,
10     the distribution of predicted O3 tissue dose to a ventilatory unit in a rat as a function of
11     distance from the bronchoalveolar duct junction is very consistent  with the distribution of
12     alveolar wall thickening.  In the case of interspecies comparisons (using the delivered
13     O3 dose to the proximal alveolar regions), although the tachypneic responses (i.e., rapid,
14     shallow breathing)  differ markedly between rats and humans, there is similarity of dose-
15     response patterns in inflammation among species, with humans and guinea pigs more
16     responsive than rats and rabbits.  In other words, the quantitative relationship between animal
17     and human responses is dependent on the animal species and the endpoint.
18
19          In summary,  there is an emerging consistency among a variety of O3 dosimetry data
20     sets and between the experimental data and theoretical predictions of O3 dose. The
21     convergence of experimental data with theoretical predictions lends a degree of confidence to
22     the use  of theoretical models to predict total and regional O3 dose. The use of O3 dosimery
23     data and models is beginning to bear fruit in attempts to extrapolate effects between animals
24     and humans.  The data and models have thus far helped demonstrate that humans may be
25     more  responsive to O3 than rats with respect to inflammatory responses.   Thus, chronic
26     effects data in rats may not accurately reflect the degree to which comparably exposed
27     humans would respond.
28
29
30     1.9  INTEGRATIVE SUMMARY OF OZONE HEALTH EFFECTS
31
32          This section summarizes the primary conclusions derived from an integration of the
33     known health effects of O3 provided by animal lexicological, human clinical, and
34     epidemiological studies.
35
36     1.  What  are the health effects of short-term  (<8 h) exposures to ozone?
37
38          Acute O3 exposure of laboratory animals and humans causes changes hi pulmonary
39     function,  including tachypnea (rapid, shallow breathing), decreased lung volumes and flows,
40     and increased airway responsiveness to  nonspecific stimuli. Increased airway resistance
41     occurs in  both humans and laboratory animals, but typically at higher exposure levels than
42     other functional endpoints.  In addition, adult  human subjects experience O3-induced
43     symptoms of airway irritation such as cough or pain  on deep inspiration.  The changes in
44     pulmonary function and respiratory symptoms occur as a function of exposure concentration,
45     duration,  and level of exercise.  Recovery of pulmonary function  and the  absence of
        February  1994                           1-32       DRAFT-DO NOT QUOTE OR CITE

-------
 1     O3-induced symptoms is usually complete within 24 h of the end of exposure, although other
 2     responses may persist somewhat longer.
 3
 4            •  Pulmonary function decrements are generally observed in healthy subjects
 5               (8 to 45 years of age) after 1 to 3 h of exposure as a function of the level of
 6               exercise performed and the O3 concentration inhaled during the exposure.
 7               Group mean data from numerous controlled human exposure and field
 8               studies indicate that, in general, statistically significant pulmonary function
 9               decrements beyond the range of normal measurement variability (e.g., 3 to
10               5% for FEVi) occur
11
12               (1) at >0.5 ppm when at rest,
13               (2) at >0.37 ppm  with light exercise (slow walking),
14               (3) at >0.30 ppm  with moderate exercise (brisk walking),
15               (4) at >0.18 ppm  with heavy exercise (easy jogging), and
16               (5) at >0.16 ppm  with very heavy exercise (running).
17
18               For a number of studies, small group mean changes (e.g., <5%)  in FEV1;
19               the medical significance of which is a matter of controversy, have been
20               observed at lower O3 concentrations than those listed above. For example,
21               data from one specific study indicate that FEVj decrements occur with very
22               heavy exercise in healthy adults at 0.15 to  0.16 ppm O3, and data from two
23               studies indicate that such effects may occur in healthy adults at levels as
24               low as 0.12 ppm.  Also, pulmonary function decrements have been
25               observed in children and adolescents at concentrations of 0.12  and
26               0.14 ppm O3 with  heavy exercise. Pulmonary function  decrements were
27               observed at 0.12 ppm O3 in healthy young adults undergoing heavy exercise
28               in a recent study.  Some individuals within a study may experience FEVj
29               decrements in excess of 15% under these exposure conditions,  even when
30               the group mean  decrement is less than 5 %.
31
32            •  For exposures of healthy subjects performing moderate exercise during
33               longer duration exposures  (6 to 8 h),  5% group mean decrements  in FEV^
34               were observed at
35
36               (1)  0.08 ppm O3 after 5.6 h,
37               (2)  0.10 ppm O3 after 4.6 h, and
38               (3)  0.12 ppm O3 after 3 h.
39
40               For these same subjects, 10% group mean FEVj  decrements were observed
41               at 0.12 ppm O3  after 5.6 and 6.6 h.  As in the shorter duration studies,
42               some individuals experience changes larger than those represented by the
43               group mean changes.
44
45            •  An increase in the incidence of cough has been reported at
46               O3 concentrations as low as 0.12 ppm in healthy  adults  during 1 to 3 h of
47               exposure with very heavy  exercise.  Other respiratory symptoms,  such as

       February 1994                           1-33      DRAFT-DO NOT QUOTE OR CITE

-------
 1              pain on deep inspiration, shortness of breath, and lower respiratory scores
 2              (a combination of several symptoms), have been observed at 0.16 to
 3              0.18 ppm O3 with heavy and very heavy exercise.  Respiratory symptoms
 4              have also been observed following exposure to 0.08, 0.10, and 0.12 ppm
 5              O3 for 6.6 h with moderate levels of exercise.
 6
 7            • Increases in nonspecific airway responsiveness have been observed after
 8              1 to 3 h of exposure to 0.40 ppm, but not 0.20 ppm, O3 at rest, and have
 9              been observed at concentrations as low as  0.18 ppm, but not to 0.12 ppm,
10              O3 during exposure with very heavy exercise.  Increases in nonspecific
11              airway responsiveness  during 6.6-h exposures with moderate levels of
12              exercise have been observed at 0.08, 0.10, and 0.12 ppm O3.
13
14          Acute O3 exposure of laboratory animals and humans disrupts the barrier function of
15     the lung epithelium, permitting materials in the airspaces to enter lung tissue, allowing cells
16     and serum proteins to enter the airspaces (inflammation), and setting off a cascade of
17     responses.
18
19            • Increased levels of neutrophils and protein in lung lavage fluid have been
20              observed following exposure of humans to 0.20,  0.30, and 0.40 ppm with
21              very heavy exercise and have not been studied at lower concentrations for
22              1- to 3-h exposures.  Increases in protein and/or  neutrophils have also been
23              observed at 0.08 and 0.10 ppm O3 during 6.6-h exposures with moderate
24              exercise; lower concentrations have not been tested.
25
26          Acute O3 exposure of laboratory animals and humans impairs alveolar macrophage
27     clearance of viable and nonviable particles from the  lungs and decreases the effectiveness of
28     host defenses against bacterial lung infections in animals and perhaps humans.  The ability of
29     alveolar macrophages to engulf microorganisms is decreased in humans exposed to 0.08 and
30     0.10 ppm for 6.6 h with moderate exercise.
31
32     2.  What are the health effects of repeated, short-term  exposures to ozone?
33
34          During repeated acute exposures, some of the O3-induced responses  are partially or
35     completely attenuated. Over a 5-day exposure, pulmonary function changes are typically
36     greatest on the second day, but return to control levels by the fifth day of exposure.  Most of
37     the inflammatory markers (e.g.,  neutrophil influx) also attenuate by the fifth day of exposure,
38     but markers of cell damage (e.g., lactate dehydrogenase  enzyme activity) do not attenuate
39     and continue to increase.  Attenuation of lung function decrements is reversed following 7 to
40     10 days without O3.  Some inflammatory markers are also reversed during this time period,
41     but others still show attenuation  even after 20 days without O3.  The mechanisms and
42     impacts involved in attenuation are not known, although the underlying cell damage continues
43     throughout the attenuation process. In addition, attenuation may alter the normal distribution
44     of O3 within the lung, allowing more O3 to reach sensitive regions, possibly affecting normal
45     lung defenses (e.g., neutrophil influx in response  to inhaled microorganisms).
46
47

       February 1994                            1-34      DRAFT-DO NOT QUOTE OR CITE

-------
 1     3.    What are the health effects of long-term exposures to ozone
 2
 3           Exposure to O3 for months and years causes structural changes in several regions of the
 4     respiratory tract, but effects may be of the greatest importance in the centriacinar regions
 5     where the alveoli and conducting airways meet because this region is typically affected in
 6     most chronic diseases of the human lung. This information on O3 effects in the distal lung is
 7     derived from animal lexicological studies because, to date, such data are not available in
 8     humans.  Epidemiological studies attempting to associate chronic health effects in humans
 9     with long-term O3 exposure have yet to provide unequivocal evidence that such a linkage
10     exists.
11
12           Chronic exposure of one strain of female mice to high 63 levels (1 ppm) caused a
13     small, but statistically significant increase hi lung tumors.  There was no concentration-
14     response  relationship and rats were not affected.  Genotoxicity data are either negative or
15     weak.  Given the nature of the database, the effects in one strain of mice cannot yet be
16     qualitatively extrapolated to humans. Ozone did not show tumor-promoting activity in a
17     chronic rat study (at 0.5 ppm O3).
18
19     4.    What are the health effects of binary pollutant mixtures containing ozone?
20
21           Combined data from laboratory animal and controlled human exposure studies on
22     O3 support the hypothesis that coexposure to pollutants, each at low-effect levels, may result
23     in effects of significance. The data from human studies of O3  in combination with NO2,
24     SO2, sulfuric  acid, nitric acid, or CO show no more than an additive response on lung
25     spirometry or respiratory symptoms. The larger number of laboratory animal studies with
26     O3 in mixture with NO2 and sulfuric acid show that effects can be additive, synergistic, or
27     even antagonistic, depending upon the exposure regimen and the endpoint studied.  This issue
28     of exposure to copollutants remains poorly understood, especially with regard to chronic
29     effects.
30
31     5.    What population groups are  at-risk as a result of exposure to ozone?
32
33           Identification of population groups that may  show increased susceptibility to O3 are
34     based on their (1) biological responses to O3; (2) physiological status (e.g., preexisting lung
35     disease);  (3) activity patterns; (4) personal exposure history; and (5) personal  factors (e.g.,
36     age, nutritional status).
37
38           The predominant information on the health effects of O3  noted above  comes from
39     studies on healthy, nonsmoking, exercising subjects, 8 to 45 years of age.  These studies
40     demonstrate that among this group, there is a large variation in sensitivity and responsiveness
41     to O3, with at least a 10-fold difference  between the most and  least responsive individuals.
42     Individual sensitivity to O3  may also vary throughout the year, related to seasonal variations
43     in ambient O3 exposure.  The specific factors that contribute to this large intersubject
44     variability, however, remain undefined.   Although differences  may be due to the dosimetry
45     of O3 in  the respiratory tract, available data show little effect on O3 deposition after
46     inhalation through the nose or mouth.


       February 1994                             1-35       DRAFT-DO NOT  QUOTE OR CITE

-------
 1          Controlled studies on mild asthmatics suggest that they have similar lung volume
 2     responses but greater airway resistance changes to O3 than nonasthmatics.  Furthermore,
 3     limited data from studies of moderate asthmatics suggest that they may have greater lung
 4     volume responses than nonasthmatics.  Daily life studies reporting an exacerbation of asthma
 5     and decrease in peak expiratory flow rates, particularly in asthmatic children, appear to
 6     support the controlled studies; however, those studies are confounded by concurrent
 7     temperature, particle or aeroallergen exposure, and asthma severity of the subjects or their
 8     medication use. In addition, field studies of summertime daily hospital admissions  for
 9     respiratory causes show a consistent relationship between asthma and ambient levels of O3 in
10     various locations in the northeastern United States, even after controlling for independent
11     contributing factors.
12
13          Other population groups with preexisting limitations in pulmonary function and exercise
14     capacity (e.g., chronic obstructive pulmonary  disease,  chronic bronchitis, ischemic  heart
15     disease)  would be of primary concern  in evaluating the health effects of O^.  Unfortunately,
16     not enough is known about the responses  of these individuals to make definitive conclusions
17     regarding their relative responsiveness to  O3.  Indeed, functional effects in these individuals
18     with reduced lung function may have greater clinical significance than comparable changes in
19     healthy individuals.
20
21          Currently available data on personal factors or personal exposure history known or
22     suspected of influencing responses  to O3 are the following.
23
24            •  Human studies have identified a decrease in pulmonary function
25               responsiveness to O3 with increasing age, although symptom rates remain
26                similar. Toxicological studies are not easily interpreted but suggest that
27               young animals are not more responsive than adults.
28
29            •  Available lexicological and human data have not conclusively  demonstrated
30               that males and females respond differently to O3. If gender differences
31               exist for lung function responsiveness to O3, they are not based on
32                differences in baseline pulmonary function.
33
34            •   There is no compelling evidence to date to suggest that any ethnic or racial
35                group has a different distribution of responsiveness  to O3.  However,  data
36                are not adequate to rule out the possibility of such differences.
37
38            •  Information derived from O3 exposure of smokers is limited.  The general
39                trend is that smokers are less responsive than nonsmokers. This reduced
40                responsiveness may wane after smoking cessation.
41
42            •  Although nutritional status (e.g., vitamin E  deficiency) makes laboratory
43                rats more susceptible to O3-induced effects, it is not clear if vitamin E
44                supplementation has an effect in human populations.  Such supplementation
45                has no or minimal effects in animals.  The role of such antioxidant vitamins
46                in O3 responsiveness, especially their deficiency, has not been well studied.
47

        February  1994                             1-36       DRAFT-DO NOT QUOTE OR CITE

-------
1          Based on information presented in this document, the population groups that have
2     demonstrated responsiveness to ambient concentrations of O3 consist of exercising healthy
3     and asthmatic individuals, including children, adolescents, and adults.
       February 1994                            1-37       DRAFT-DO NOT QUOTE OR CITE

-------
 i              9.  INTEGRATTVE SUMMARY OF OZONE
 2                                HEALTH EFFECTS
 3
 4
 5     9.1   INTRODUCTION
 6          This chapter integrates the knowledge gained in animal toxicological, human clinical,
 7     and epidemiological studies of ozone (O3) that were discussed in Chapters 6, 7, and 8
 8     respectively, of the criteria document. Because each of these approaches has different
 9     strengths and weaknesses, a combined evaluation can better describe the full array of health
10     effects that are known to occur with exposure to 03.
11          The chapter is organized according to the health effects of short- and long-term
12     exposures of O3 alone and exposures to binary mixtures with O3.  The section on short-term
13     exposures (i.e., less than 8 h) begins with a discussion  of the relationship between exposure
14     and dose, as this lays a  foundation for inter- and intraspecies extrapolation.  Effects on lung
15     function, exacerbation of existing disease, and cellular-biochemical responses are then
16     presented descriptively.  Finally, quantitative exposure-response relationships for the effects
17     of O3 on pulmonary function (e.g., changes in lung volume) are summarized separately
18     because the large number of studies allows more complex evaluations and modeling. For the
19     other classes of effects,  the exposure-response information is integrated with the description
20     of the effects.  The section on long-term exposures encompasses repeated exposures (i.e.,
21     1 to 5 days), prolonged exposures (i.e., months), and genotoxicity and carcinogenicity.
22     Because the  data base on binary exposure studies has little predictive value, the emphasis is
23     placed on the principles of interaction.  The conclusions section is organized according to
24     key questions.
25          Because this chapter integrates the results of a large number of studies from the current
26     and earlier O3  criteria documents, it is not practical to  provide experimental details  or cite
27     specific references. Rather, emphasis is given to main findings which are supported by
28     theory or other confirmatory studies, unless noted otherwise.  Comprehensive  details and
29     references are  provided in the previous chapters.
30

       February 1994                           9-1       DRAFT-DO NOT QUOTE OR  CITE

-------
 1     9.2   HEALTH EFFECTS OF SHORT-TERM EXPOSURES
 2     9.2.1    Exposure-Dose Relationships
 3          Qualitative and quantitative health assessment requires, among other things, the ability
 4     to relate exposure to dose and dose to effect. In the case of O^ health assessment, this
 5     ability is necessary for two major reasons:  (1)  to develop unified predictive models of
 6     human population responses based upon exposure; and (2) to enable extrapolation from
 7     animals to humans for chronic effects.  Physically and biologically  based models of dose
 8     simplify the methods of predicting population responses and in turn significantly reduce the
 9     uncertainty of these predictions.  For animal-to-human extrapolations,  splitting the problem
10     of exposure and response into an exposure-dose problem and a dose-response problem
11     separates the issue of interspecies sensitivity from purely dosimetric considerations.
12     Responses in animals may be homologous with humans but follow  different dose-response
13     curves.   By measuring or computing delivered O3 dose to relevant  tissues in animals and
14     humans, transfer functions can,  in principle, be developed relating  dose-response curves
15     among different species.  This section discusses the understanding of exposure-dose
16     relationships and how they  improve the ability to interpret and predict O3 responses.
17          Historically, the first step beyond  describing responses solely in terms of exposure
18     concentration was the use of the product of concentration x time x minute ventilation
19     (C x T x  VE), yielding what has been often referred to as an "effective dose".  Response
20     modeling has examined the interaction of individual pairs of variables.  However, no single
21     model has been able to simply unify any response in terms  of the product C x T  x VE.
22     This is  due to the fact that C x T x  VE is a metric of exposure dose and not delivered dose
23     and, furthermore, does not account for  the mediation of responses  in localized regions of the
24     lung that would be responding to local O3 doses. Advances in O3  dosimetry modeling and
25     experimental determinations of regional O3 dose in animals and humans have enabled
26     extensions beyond simple C x T X VE modeling to interpret responses.
27          Ozone dosimetry models provide predictions on the dose distribution of O3 in the
28     respiratory tract from the trachea to the alveolar spaces of the lung. These models utilize the
29     best available anatomical, physiological, and biochemical data available for animals and
30     humans. These data are incorporated into mathematical formulations of convection,
31     diffusion, and chemical reaction processes  in the lung. The models predict that under resting

       February 1994                             9-2      DRAFT-DO NOT QUOTE OR CITE

-------
 1     ventilatory conditions, the O3 dose per airway generation to all respiratory tract constituents
 2     (tissue plus fluid) slowly decreases from the trachea to the terminal and respiratory
 3     bronchioles and then declines in the alveolated generations.  When dose of O3 to tissue alone
 4     is considered (taking account of reaction and diffusion kinetics in the liquid lining layer),
 5     there is a three order of magnitude increase in tissue dose from the trachea to the proximal
 6     alveolar regions, after which the tissue and total dose are virtually equal and fall rapidly in
 7     the alveolated generations.   Currently, relationships between delivered regional dose and
 8     response are derived assuming that O3 is the active agent directly responsible for effects.
 9     There is uncertainty, however, whether this assumption is correct.  Reactive intermediates,
10     such as peroxides and aldehydes formed when O3 interacts with constituents of lung lining
11     fluid, may be the agents mediating responses. Thus, the dose of the reactive intermediates
12     may be more relevant than the dose of O3.  Despite this suggestion, the histopathological
13     findings from chronic  O3 exposures in animals match the predicted distribution of O3 dose
14     (i.e., the sites of the highest predicted O3 doses  correspond with those regions of the lung
15     with the greatest tissue alterations).
16           Experimental studies in humans have revealed some important features needed for
17     health assessment.  Among these is the  observation that the dose of O3 delivered to the lower
18     respiratory tract is independent of  the mode of breathing (i.e., oral versus nasal versus
19     oronasal).  This observation simplifies health assessment by eliminating  the need for precise
20     information on modes of breathing when considering population responses. Experimental
21     studies in humans have also shown that increasing VE with exercise (increasing both
22     breathing frequency and tidal volume) only causes a small decrease in O3 uptake efficiency
23     by the total respiratory tract.  Based on models of O3 dose, it appears that the increased
24     VE in exercise,  though having little effect on uptake efficiency by the total respiratory tract,
25     causes the distribution of delivered O3 dose to shift deeper into the respiratory tract. The
26     shift in O3 dose as a function of VE could help explain the complex relationships seen
27     between response and  C, T, and VE. An important observation from the human
28     experimental dosimetry studies is the general agreement between O3 dosimetry models and
29     the measured data.
30           Experiments in laboratory  animals (particularly rats) have been valuable in providing,
31     in conjunction with human experimental data and mathematical  dosimetry models, the basis

       February 1994                             9.3       DRAFT-DO NOT QUOTE OR CITE

-------
 1      for dosimetric extrapolation.  Whereas the human total respiratory tract has an O3 uptake
 2      efficiency between 70 and 100%, the respiratory tract of the rat only takes up about 50% of
 3      the inhaled O3.  Unlike the case with humans, the dosimetry models overestimate the uptake
 4      efficiency of the rat respiratory tract by approximately 25 to 50% (i.e., the predicted uptake
 5      efficiency is between 65 and 75%), but the models are still highly valuable for extrapolation
 6      purposes. An important finding has been that the models correctly relate the regional dose of
 7      O3 to the increase in alveolar wall thickness,  both of which decline with distance from the
 8      junction of the conducting airways and the alveolar regions of the lung.
 9           Experimental O3 dosimetry and predictive O3 dosunetry models are informative about
10      the feasibility of extrapolating animal responses to humans. Some responses to 03 can be
11      compared across species on a strict dose-response basis.  For example, both animals and
12      humans respond to O3 in a dose-dependent manner by increasing breathing frequency and
13      decreasing tidal volume (tachypnea).  A qualitative comparison between rat and human
14      tachypneic responses at a variety of O3 concentrations and exercise levels  indicate that when
15      exercising, rats and humans have a similar response, but at rest rats are somewhat more
16     responsive.  However,  when dose to the proximal alveolar region of the lung (normalized to
17     body weight) is considered as the dose metric for tachypneic responses, rats appear to be
18      much more responsive than humans.  Another example is influx of protein into the alveolar
19      spaces following O3 exposure as measured in bronchoalveolar lavage (BAL) fluid. When
20     BAL protein is plotted as a function of pulmonary tissue dose,  the rat, guinea pig, rabbit,
21      and human all respond  with a similar dose-response pattern, reflecting a common mechanism
22     of response.  However, each curve is offset from the other, reflecting overall sensitivity
23     differences among the species, with the human and guinea pig being more responsive than
24     the rat and rabbit.
25
26     9.2.2   Physiological Responses to Ozone Exposure
27          Typical acute physiological responses to 03 exposure in humans include a reduction in
28     forced vital capacity (FVC), decreased expiratory flow rates, and increased respiratory
29     symptoms. The most common symptoms include cowjj|flairway  irritation, and chest
         I'V' X. I WiiaJ                     : irtJH |W]i) i        i,li  i.n V4HWI I I
30     discomfort associated with a deep inhalation.  These responses are often accompanied by
31     increased airway resistance and tachypnea.   The voluntary spirometry and symptom

       February 1994                            9.4      DRAFT-DO NOT QUOTE OR CITE

-------
  1      responses cannot be elicited from animals, but their tachypneic response is well documented.
  2      Ozone exposure also increases airway responsiveness to nonallergenic airway stimuli (e.g.,
  3      histamine) in humans and animals.  There is a large range of responses among humans with
  4      at least a  10-fold difference between the most and least responsive individuals.
  5
  6      9.2.2.1   Respiratory Symptom Responses
  7       \   The principal symptoms associated with  O3  exposure in humans include cough,
         -.-$;      ,   '                               .          .'-„..,.--;
  8   :   irritation of the airways  (described as a scratchy throat or discomfort under the sternum), and
                 when Sati|ffa deep breath (described as chest ti^^inMrpaii% fie chest).
                ^-sometimesReported"as a symptom in field or epidemiological studies witti
11   )  to oxidant mixtures, is not associated with exposure to O^ alone. The receptors responsible
12      for cough may be unmyelinated C-fibers or rapidly adapting receptors located in the larynx
13      and the largest conducting airways.  Thus, there appears to be a potential mechanistic linkage
14      to changes in spirometry.  Field and epidemiological studies also indicate an  association
15 fc*tjetween hourly or daily ambient 03 levels and the presence of symptoms, particularly
16   I  Such associations may-lie most evident in  aslhmatic children? ^Although symptoms"cahtiotTxT
17      elicited from animals, indirect measures of symptom responses in animals include behavioral
18      responses indicative of aversion to O3 exposure.
19          Symptom responses to O3  exposure follow a monotonic exposure-response relationship
20      that has a similar form to that for spirometry responses.  Increasing exposure levels elicit
21      increasingly more severe symptoms that persist for longer periods.  Symptom and spirometry
22      responses follow a similar time course during  an acute exposure and the subsequent recovery
23      as well as over the course of several days in a repeated exposure study.  Furthermore,
24      medication interventions that block or reduce spirometry responses  have a similar effect on
25      symptom responses. Levels at which symptoms occur under various exposure conditions are
26      discussed in Section 9.2.5.1.  As with spirometry responses, symptom responses vary
27      considerably among subjects,  although the individual correlations between spirometry and
28      symptom responses are relatively low. In several heavy or severe exercise studies of athletes
29      exposed to O3, the discomfort associated with the respiratory symptoms caused by
        O3  concentrations in excess of 0.18 ppm was of sufficient severity that the athletes reported^*
        that they would have been unable to perform maximally, if the conditions of the exposure >

        February  1994                            9.5       DRAFT-DO NOT QUOTE OR CITE

-------
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
19
20
21
22
23
24
25
26
27
28
29
30
31
            present during athletic competition.  In workers or active people exposed to O3,
       respiratory symptoms may cause reduced productivity or curb the desire to pursue leisure
       activity.
            Symptom responses have also been reported in asthmatics exposed-to-Qg-  In contrast to
       nonasthmatics, wheezing is a prevalent symptom in addition to cough, chest tightness, aa4'f:v;:
                                               ,-.-,-                        '            •'Hr-'i'iS
                                              -' '-' ' 	 "                       ,             :., " =3
       shortness of breath that are reported by subjects without asthma.

       9.2.2.2   Changes in Lung Volume
            In humans, 0j exposure reduces FVC primarily by decreasing iaspiratory
       This is believed to be the result of neurogenic inhibition of maximal inspiration, possibly
       caused by stimulation off C-fibesr afferents. C-fibers are also thought to be the receptors
       responsible for the cough reflex in humans.  After exposure to O3, coughing is frequently
       elicited during the deep  inspiration prior to the forced expiratory maneuver used in dynamic
       spirometry tests such as  FVC, forced expiratory volume in 1 s (FEVj), and forced expiratory
       flow at 25 to 75% of FVC (FEF25_75%).  The observation that nonsteroidal antiinflammatory
       drugs (e.g.,  indomethacin, ibuprofen) reduce or block spirometric responses to O3 exposure
17     and reduce levels of prostaglandin £2 (PGE2) within the lung suggest that mediators released
18    , by damaged epithelial cells .and/or alveolar macrophages may play
       m^dmal inspiration.  Although it seems clear that the reduction in total lung capacity (TLC)
       is not attributable to reduced lung compliance (i.e., a stiffer lung ) or inspiratory  muscle
       weakness, the understanding of the mechanism(s) which cause this response remains
       incomplete.  The  O3-induced tachypneic response, seen in many animal species and in
       exercising humans, may be related to the decrease in vital capacity.  The tachypneic response
       hi humans may not be entirely involuntary because it has been reported that O3-exposed
       subjects may consciously modify their breathing pattern to relieve discomfort.
            The time course of the spirometry responses to O3 exposure depends on  the exposure
       conditions.
           |8ppm), responses  are induced slowly
                of response depending upon the duration of the exposure.  At higher levels of
                (e.g., very heavy exercise andO? concentration >Q.25 ppmj, responses occur Hf
      StHwwf%••/ r  -  .        -,-  . '•;         -    -"*-..    -  •-   - --  :   •"*  •, .  .-.-,. f<--j/i:-: ^?i-*»*?**.-21s;
     / rapidly (within 15 min) and the largest portion of the response tends to occur early in
       February 1994
                                                 9-6
DRAFT-DO NOT QUOTE OR CITE

-------
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
                               mwtthta 1 i to 2 h-  The exposure-response relationships are
discussed more extensively in Section 9.2.5.
 9.2.2.3   Responses to Ambient Ozone Exposures
                                           "is-av •*-»*#>
                          and adolescents exposed to
                                 changes mspnometry similar to those
                                 «rimeotal <)aaditiofes, il^re is a substantial range of
                                                * f    „„
"respcmselinTong individuals in camp studies arid "between various locations.  However, the
 average FEV^ was lower when ambient O3 was higher.  Although direct comparisons cannot
 be made because of incompatible differences in experimental design and analytical approach,
 this range of response is comparable to the range of responses seen in chamber studies at low
 O3 concentrations.  The similarity  of response between field and chamber studies leaves little
 doubt as to the relevance of controlled human exposure studies.

 9.2.2.4  Changes in Airway Resistance
                         airway resistance following O3 exposure are small relative to those
                  with an inhalation exposure to a bronchoconstricting drug (methacholine),
 a specific antigen, or to sulfur dioxide (SO2).  However, because the airways of healthy
 nonallergic people are relatively nonresponsive to a number of stimuli, including allergens,
 methacholine, and histaminej
                            .whose airways are much more sensitive,  u3 aeenrrorcause
                             although the changes are clearly greater than in healthy young
 subjects. In rats exposed to O3, changes in resistance also tend to be small. The observation
 that changes in airway resistance are modest clearly indicates that reductions in maximum
 expiratory flow are not caused primarily by narrowing of large airways.  The increase in
 airway resistance appears to be vagally mediated because it is sensitive to inhibition by
 atropine.
        February 1994
                                           9-7
DRAFT-DO NOT QUOTE OR CITE

-------
 1      9.2.2.5   Changes in Breathing Pattern
 2          k          "  '	"
       M; 5 :•--•>•.
 3      breathing
     i»"*-'» .   -, ""   "^
 4                         ^fc^^^pjMal breaths.OT particulif IntereitToFBo^aring
 5     interspecies responses is that the responses of rats and guinea pigs fall within the same range
 6     as seen for humans from rest to heavy exercise.
 7
 8     9.2.2.6   Changes in Airway Responsiveness
 9          Ozone exposure causes increased responsiveness of the pulmonary airways to
10     subsequent challenge with bronchoconstrictor drugs such as histamine or methacholine.
11     Although this phenomenon is seen even after regression of spirometric changes,  it is typically
12     no longer present after 24 h.  One animal study has demonstrated decreased antigen-induced
13     bronchoconstriction after O3 exposure, and one human study is suggestive of an  increase in
14     such a response.  An increased response to a specific antigen to which a human  is sensitized
15     is a plausible outcome of O3 exposure but needs to be further investigated. Although
16     changes in airway responsiveness tend to resolve somewhat more slowly and appear to be
17     less likely to be attenuated with repeated exposure, the evidence for a persistent  increase in
18     responsiveness from animal studies is inconsistent.  Changes in airway responsiveness in rats
19     and guinea pigs tend to occur at higher O3 concentrations and, as in humans, tend to be most
20     pronounced shortly after the exposure and less so 24 h postexposure.  Changes in airway
21     responsiveness appear to  occur independently of changes in pulmonary function.   This
22     response does not appear to be due to inflammation (at least PMNs) or to the release of
23     arachidonic acid metabolites, but may be due to epithelial damage and the consequent
24     increased access of these chemicals to  smooth muscle in the airways or to the receptors in the
25     airways responsible for reflex bronchoconstriction.  The clinical relevance of this observation
26     is that after O3  exposure, human airways may be more susceptible to a variety of stimuli,
27     including antigens, chemicals,  and particles.
28
29     9.2.2.7   Small Airways Responses
30          There are several tests purported to be indicators of "small airways" function, although
31     the morphological correlate of these functional tests is not clearly established.

       February  1994                            9-8      DRAFT-DO NOT QUOTE OR CITE

-------
 1     Morphologically, the small airways of the centriacinar region of the lung, that segment
 2     between the last conducting airway and the gas exchange region, is highly susceptible to
 3     damage by O3 and is the site of epithelial cell necrosis and remodelling of respiratory
 4     bronchioles.  Numerous pulmonary function tests reputed to measure responses in small
 5     airways (e.g., closing volume, aerosol bolus) have been used in O3 studies.  Responses have
 6     been demonstrated, but it is not clear that these tests correlate with the morphological lesion
 7     observed in animals, which is presumed to occur but has not been demonstrated in humans.
 8     In dogs whose peripheral airways are directly exposed to O3 through a wedged
 9     bronchoscope, the collateral resistance to airflow through nonairway channels is increased
10     almost immediately.
11
12     9.2.2.8   Effects on Exercise Performance
13          A large number of studies show that O3 exposure can interfere with exercise
14     performance, either by reducing  maximal sustainable levels of activity or reducing  the
15     duration of activity which can be tolerated at a particular work level.  Such effects can be
16     seen with very heavy exercise at O3 concentrations of 0.18 ppm and above.  Although there
17     are a large number of factors which can alter maximal or submaximal exercise performance,
18     many investigators have implicated the respiratory symptoms caused by O3 exposure.  Small
19     changes in airway resistance have no effect on maximal or submaximal exercise
20     performance, and even modest mechanical restriction of total lung capacity will not induce a
21     respiratory limitation to maximal or submaximal exercise.  Animals studies indicate
22     decreased wheel running activity and decreased activity associated with obtaining food.
23     Decreased worker productivity has also been associated with elevated O3 levels.
24
25     9.2.3   Exacerbation of Existing Disease
26     9.2.3.1   Responses of Asthmatics to Controlled Ozone Exposure
27          Asthmatics have qualitatively similar responses to O3 exposure as nonasthmatics.
28     Although symptom and volume related responses (i.e., decreased FVC) tend to be  similar,
29     airway  resistance increases relatively more, from an already higher baseline, hi asthmatics
30     exposed to O3.  Altered responsiveness to bronchoconstrictor drugs shows similar changes in
31     asthmatics and nonasthmatics.  There is no evidence at this time that O3 induces a  persistent

       February  1994                            9.9       DRAFT-DO NOT QUOTE OR CITE

-------
 1     increase in airway responsiveness or that O3-exposed asthmatics are more likely to have a
 2     late phase response to specific antigen challenge.
 3
 4     9.2.3.2  Increased Hospital Admissions and Asthma Attacks
 5          A number of epidemiological studies have shown a consistent relationship between
 6     ambient oxidant exposure and acute respiratory morbidity in the population.  Decreased lung
 7     function and increased respiratory symptoms, including exacerbation of asthma, occur with
 8     increasing ambient O3, especially in children.  Modifying factors, such as ambient
 9     temperature, aeroallergens, and other copollutants (e.g., particles) can also contribute to this
10     relationship. Ozone air pollution may at least account for a portion of summertime hospital
11     admissions for respiratory causes, because studies conducted in various locations in the
12     eastern United States have consistently shown a relationship with increased incidence of
13     admissions, even after controlling for modifying factors, as well as when considering only
14     concentrations below 0.12 ppm O3. It has been  estimated from these studies that O3 may
15     account for roughly one to three excess respiratory hospital admissions per hundred parts per
16     billion O3, per million persons.
17          The association between elevated ambient O3 concentrations during the summer months
18     and increased hospital admissions has a plausible biologic basis in the physiologic,
19     symptomatic, and field study evidence  discussed earlier.  Specifically, increased airway
20     responsiveness, airway inflammation, increased airway permeability, and increased incidence
21     of asthma attacks suggest that ambient O3 exposure could be a cause of the increased hospital
22     admissions, particularly for asthmatics.
23
24     9.2.4   Cellular-Biochemical  Responses
25     9.2.4.1  Inflammation and Cell Damage
26          Ozone-induced cell injury may lead to effects including inflammation, altered
27     permeability of the epithelial barrier, impaired host defense and particle clearance,
28     irreversible structural alterations in the lung, exacerbation of preexisting disease (e.g.,
29     asthma), and increased sensitivity to biocontaminants (e.g., allergens).  Of these, O3-induced
30     inflammation of the respiratory tract has been best documented and occurs in all species that
31     have been studied.  The mechanisms leading to the observed inflammatory responses induced

       February 1994                            9-10       DRAFT-DO NOT QUOTE OR CITE

-------
 1     by O3 are just beginning to be studied.  Both animal morphological studies and in vitro
 2     studies indicate that airway ciliated epithelial cells and Type 1 cells are the most O3-sensitive
 3     cells, and are initial targets of O3. These cells are damaged by O3 and produce a number of
 4     proinflammatory mediators (e.g., IL-6, JCL-8, PGI^) capable of initiating a cascade of events
 5     leading to neutrophil influx into the lung, activation of alveolar macrophages, inflammation,
 6     and increased permeability across the epithelial barrier.
 7
 8     Ozone-Induced Inflammation in the Lower Respiratory  Tract
 9           In general, inflammation can be considered  as the host response to injury  and the
10     induction of inflammation as evidence that injury has occurred.  Inflammation induced by
11     exposure of humans to O3 can have  several outcomes:  (1) inflammation induced by a single
12     exposure (or several exposures over  the course of a summer) can resolve entirely;
13     (2) continued acute inflammation can evolve into  a chronic inflammatory state;  (3) continued
14     inflammation can alter the structure and function  of other pulmonary tissue, leading to
15     diseases such as fibrosis; (4) inflammation can alter the body's host defense response to
16     inhaled microorganisms, particularly in potentially vulnerable populations such  as the very
17     young and old; and (5) inflammation can alter the lung's response to other agents such as
18     allergens or toxins. It is also possible that the profile of response can be altered in persons
19     with preexisting pulmonary  disease (e.g., asthma or COPD) or smokers.
20           The recent use of BAL as a research tool in humans has afforded the opportunity to
21     sample the lung and lower airways of humans exposed to O3 and to ascertain the extent and
22     course of inflammation and  its constitutive elements.  Several studies have shown that
23     humans exposed acutely (1 to 3 h) to 0.2 to 0.6 ppm O3  had O3-induced inflammation, cell
24     damage, and altered permeability of epithelial cells lining the respiratory tract (allowing
25     components from plasma to enter the lung).  The lowest  concentration of O3 tested,
26     0.08  ppm for 6.6 h, also induced increases in these endpoints.  Short-term (< 8 h) exposure
27     of animals to O3 also results in cell damage, inflammation, and altered permeability,
28     although, in general, higher O3 concentrations are required to elicit a response  equivalent to
29     that of humans. Because humans were exposed to O3 while exercising and most animal
30     studies were done at rest, differences in  ventilation likely play a significant role in the
31     different response of humans and rodents to the same O3 concentration.  Studies in which

       February 1994                            9-11      DRAFT-DO NOT QUOTE OR CITE

-------
 1      animals were exposed at night (during their active period), or in which ventilation was
 2      increased with CO2, tend to support this idea.
 3           Studies utilizing BAL techniques sample only free or loosely adherent cells in the lung;
 4      thus, it is possible that cellular changes  have occurred in the interstitium that are not
 5      reflected in BAL studies.  However, morphometric analyses of inflammatory cells present in
 6      lung and airway tissue sections of animals exposed to O3 are in general agreement with BAL
 7      studies. Short-term O3 exposure (< 8 h) causes similar types of alterations in lung
 8      morphology in all laboratory animal species studied.  The most affected cells are the ciliated
 9      epithelial cells of the airways and Type 1 cells in the alveolar region.  The centriacinar
10      region (the junction of the conducting airways and gas exchange region) is a primary target,
11      possibly because it receives the greatest dose of O3 delivered to the lower respiratory tract.
12      Sloughing of ciliated epithelial and Type 1 cells occurs within  2 to 4 h of exposure of rats to
13      0.5 ppm O3.
14
15      Ozone-Induced Inflammation in the Upper Respiratory  Tract
16          Ozone causes inflammatory changes throughout the respiratory tract, including the
17     nose.  Humans and laboratory animals exposed to O3 develop  inflammation and increased
18     permeability in the nasal passages. Thus O3-induced changes in the nasal passages may
19     reflect similar changes occurring in the lung.  A recent study reported a positive correlation
20     between nasal inflammation in children and measured ambient O3 concentrations.  Studies
21     with rats suggest a potential competing  mechanism between the nose and lung, with
22     inflammation occurring preferentially in the nose at low O3 concentrations and shifting to the
23     lung at higher concentrations.  It is unclear if this represents a specialization restricted to rats
24     or is a more general phenomenon.
25
26     Time Course of Ozone-Induced Inflammatory Response
27          Findings from human and animal  studies agree that the O3-induced  inflammatory
28     response occurs rapidly and persists for at least 24 h.  Increased levels of neutrophils and
29     protein are observed in the BAL fluid within 1 h following a 2-h exposure of humans to
30     O3 and continue for at least 20 h.  The kinetics of response during this time have not been
31     well studied in humans, although a single study shows that neutrophil levels are higher at

       February 1994                            9-12      DRAFT-DO NOT QUOTE OR CITE

-------
 1     6 h postexposure than at 1 or 20 h.  Several animal studies suggest that neutrophil and BAL
 2     protein levels peak 12 to 16 h after an acute O3 exposure, and begin to decline by 24 h,
 3     although some studies report detectable BAL neutrophils even 36 h after exposure. It is also
 4     clear that hi humans the pattern of response differs for different inflammatory mediators.
 5     Mediators of acute inflammation, such as IL-6 and PGE^, are more elevated immediately
 6     after exposure; whereas mediators that could potentially play a role in resolving
 7     inflammation, such as fibronectin and plasminogen activator, are preferentially elevated
 8     18 h after exposure.  The rapidity with which cellular and biochemical mediators  are induced
 9     by O3  makes it conceivable that  some of them may play a role in O3-induced changes in lung
10     function—indeed there is some evidence that BAL PGE2 levels  are correlated with
11     decrements in FEV^, and anti-inflammatory medications that block PGE2 production also
12     reduce or block the spirometric responses to O3. Although earlier studies suggested that
13     O3-induced PMN influx might contribute to the observed increase in airway hyperreactivity,
14     animal studies show that when neutrophils are prevented from entering the lung, O3-induced
15     hyperreactivity or increases in many inflammatory mediators still occur.  In addition, studies
16     hi which anti-inflammatory drags are used to  block O3-induced lung function decrements,
17     still show increases in neutrophils and most other inflammatory mediators (although PGEj is
18     not increased).
19
20     Individuals and Populations  Susceptible to Ozone
21          To date, there have been no studies that have examined the cellular/biochemical
22     response of potentially susceptible subpopulations, such as asthmatics, to O3; nor  are there
23     any data in humans addressing whether age, gender, or racial differences can modify the
24     inflammatory response to O3.  However, inflammation is not induced to the same extent in
25     all individuals.  In moderately exercising humans exposed to 0.08 ppm O3 for 6.6 h, the
26     mean changes in inflammatory indices were low, but some individuals had increases
27     comparable to those reported hi  heavily exercising subjects exposed to 0.4 ppm O3 for 2 h,
28     suggesting that some segments of the population may be more responsive to low levels of  O3.
29     It has  not yet been studied whether intersubject differences hi inflammatory response  to
30     O3 are reproducible over time for the same subject, as has been shown for intersubject
31     differences in lung function.  There seems to be no strong correlation between the various

       February 1994                           9-13      DRAFT-DO NOT QUOTE OR CITE

-------
 1     mediators of inflammation, cell damage, and permeability (i.e., those individuals with the
 2     greatest neutrophil response are not necessarily those with the greatest BAL protein, PGI^,
 3     or IL-6 response).  Furthermore, the magnitude of lung function decrements and respiratory
 4     symptoms has not yet been shown  to be correlated with mediators of inflammation, with the
 5     possible exception of PGE^.
 6          Animal studies also show large interspecies and interstrain differences hi response to
 7     O3 and suggest that genetic factors may play a role in susceptibility to O3.  Different rat
 8     strains respond to O3 differently; for example, Wistar rats have the greatest neutrophil
 9     influx, whereas Fisher rats demonstrate the most epithelial cell damage. In addition, limited
10     data suggest that  dietary antioxidant levels may affect the response of rodents to O3 and that
11     very young rats produce more PGF^ in response to O3 than do older rats.  Taken as a whole,
12     both the human and animal studies suggest that the inflammatory response to O3 is complex
13     and that determinants of susceptibility may occur at several different genetic loci.
14
15     9.2.4.2   Host Defense
16          The mammalian respiratory tract has a number of closely integrated defense
17     mechanisms that, when functioning normally, provide protection from the  adverse effects of
18     a wide variety of inhaled particles and microbes.  Impaired mucociliary clearance can result
19     in unwanted accumulation of cellular secretions and increased numbers of particles and
20     microorganisms in  the lung, leading to increased  infections and bronchitis.
21
22     Mucociliary Clearance of Inhaled Particles
23          Animal studies consistently show that clearance of inhaled  insoluble particles is slowed
24     after acute exposure to O3.  Ozone-induced damage to cilia and  increased  mucus secretion
25     likely  contribute  to a slowing of mucociliary transport  rates.  Interestingly, retarded
26     mucociliary clearance  is not observed in animals  exposed repeatedly to O3.  The effects of
27     O3 on mucociliary  clearance in humans has not been well studied, and the results are
28     somewhat conflicting;  one study reports an O3-induced increase  in particle clearance in
29     subjects exposed to 0.4 ppm O3 for 2  h, and another study reports no O3-induced change  in
30     particle clearance with a similar exposure regimen.
31

       February 1994                            9-14      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Alveolar Macrophage Function
 2           Macrophages represent the first line of defense against inhaled microorganisms and
 3      particles that reach the lower airways and alveoli.  Studies in both humans and animals have
 4      shown that there is an immediate decrease in the number of BAL macrophages following
 5      O3 exposure.  Alveolar macrophages also have been shown to be crucial to the clearance of
 6      certain gram-positive bacteria from the lung.  Several studies in both humans and laboratory
 7      animals have also shown that O3 impairs the phagocytic capacity of alveolar macrophages,
 8      and some studies suggest that mice may be more impaired than rats. The production of
 9      superoxide anion (an oxygen radical used in bacterial killing) by alveolar macrophages also is
10      depressed in both humans and animals exposed to O3, and the ability of alveolar
11      macrophages to directly kill bacteria is impaired.  Decrements in alveolar macrophage
12      function have been observed  in moderately exercising humans exposed to the lowest
13      concentration tested, 0.08 ppm O3 for 6.6 h.
14
15      Interaction with Infectious Agents
16           Concern about the effect of O3  on susceptibility to respiratory infection derives
17     primarily from animal studies in which O3-exposed mice die following a subsequent
18     challenge with aerosolized bacteria.  Mortality has been shown to be concentration-
19     dependent, and exposure to as little as 0.08 ppm O3 for 3  h can increase mortality of mice to
20     a subsequent challenge with Streptococcus bacteria. In addition, younger mice are more
21     susceptible to infection than older mice; this has been related to increased PGF^ production
22     in these animals, which likely decreases alveolar macrophage activity.
23           It has been suggested that impaired alveolar macrophage function is the mechanism
24     likely responsible for enhanced susceptibility to bacteria.  However, mortality is not observed
25     with  other rodent species, raising the question of whether  this phenomenon is restricted to
26     mice. Although both mice and rats show impaired macrophage killing of inhaled bacteria
27     following O3 exposure; rats mount a faster neutrophil response to O3 to compensate for the
28     deficit in alveolar macrophage function.  The resulting slower clearance time in mice allows
29     the Streptococcus strain to persist in lung tissue and subsequently elaborate a number of
30     virulence factors that evade secondary host defense and lead to bacterial multiplication and
31     death of the host.  Mortality as an endpoint is not directly relevant to humans; however,

       February 1994                           9-15       DRAFT-DO NOT QUOTE OR CITE

-------
 1     humans and laboratory animals share many host defense mechanisms being measured by
 2     mortality in the mouse model.
 3          There is no compelling evidence from lexicological, human clinical,  or epidemiological
 4     studies that O3 increases the incidence of respiratory viral infection in humans.  A study of
 5     experimental rhinovirus infection in susceptible volunteers failed to show any effect of
 6     5 consecutive days of O3 exposure (0.3 ppm, 8 h/day) on the clinical picture or on host
 7     response.   Studies in which O3-exposed mice were challenged with influenza virus report
 8     conflicting results,  with some studies showing increased mortality, some showing  decreased
 9     mortality, and still others showing no change at all.  However, even when increased
10     mortality was demonstrated, there was no difference in viral tilers in the lung, suggesting
11     virus-specific immune functions were not altered.
12          Taken as a whole, Ihe data clearly  indicate lhat an acute O3 exposure impairs Ihe  host
13     defense capability of both humans and animals, primarily by depressing alveolar macrophage
14     function and perhaps also by decreasing  mucociliary clearance of inhaled particles and
15     microorganisms. This suggesls lhal humans exposed to O3 are likely to be predisposed to
16     bacterial infections in Ihe lower respiratory iracl.  The seriousness of such infections may
17     depend on how quickly bacteria develop virulence factors and how rapidly neutrophils are
18     mobilized to compensate for Ihe deficil in alveolar macrophage function.
19          Ozone also has been reported to suppress natural killer cell activity in the lung, to
20     suppress proliferative responses to bacterial antigen (Lisleria) in bolh spleen and bronchial
21     lymph nodes, and to induce delayed hypersensilivily responses lo Lisleria  antigen. However,
22     these effects occur al higher exposure levels (0.75 lo  1.0 ppm O3) lhan those thai affecl
23     macrophage function.
24
25     9.2.5   Ozone Exposure-Response Relationships
26          A quantitative understanding of Ihe relationship between O3 exposure and subsequenl
27     response is useful bolh for a better understanding  of Ihe processes underlying oulcomes of
28     interest and for purposes of prediction.  Examples of Ihe ulilily of Ihe latter include
29     identification of exposures unlikely lo produce effecls, risk and benefils assessmenl, and
30     prediction of responses based on exposures for which empirical dala do nol  exisl.  Exposure-
31     response relationships exisl bolh for Ihe population as a whole and for individuals wilhin Ihe

       February 1994                           9-16       DRAFT-DO NOT QUOTE OR CITE

-------
 1     population.  The form of the exposure-response relationship can be different for populations
 2     and individuals, and predicted magnitudes of effect for the population generally do not reflect
 3     the experience of all individuals.
 4          Exposure-response relationships can be described in terms of concentration of
 5     O3, dose rate of exposure, total inhaled dose, or dose at the active site. Most of the
 6     completed work to date has focused on describing the relationships between C, VE, and
 7     T (as introduced in Section 9.2.1) for either pulmonary function or BAL outcomes. Parallel
 8     work has been performed in both humans and in laboratory animals.  While exposure-
 9     response relationships for individuals have been examined to a limited degree and are  likely
10     to be generally similar in form to those for populations,  little work has focussed upon
11     understanding or quantifying individual exposure-response relationships.
12          No single exposure-response model form has been  adequately tested  and identified as
13     providing an accurate,  precise description of the relationship between exposure and response
14     in both humans and laboratory animals for lung function and BAL endpoints.  Rather, for a
15     given study, a particular model may have been selected a priori to describe the exposure-
16     response data, or may  have been identified as providing the best fit among several competing
17     models. In many cases,  models have been found  to be deficient, but rarely has the
18     performance of a number of possible models been systematically compared.  The limiting
19     factor in most of this work has  been that no single study to date has included a wide enough
20     range of the three exposure variables of interest, (i.e., C, VE, and T) to choose between
21     models or to identify the appropriate method of describing exposure. From the individual
22     studies, however, have come a  number of observations which are qualitatively true for
23     describing BAL and pulmonary function responses in both humans and laboratory animals
24     and which should be considered in the selection of a model to describe population response
25     as a function of exposure.
26
27             1. Response monotonically increases with increases in C,  VE, and T.
28
29             2. C has generally been found to be a stronger predictor of response than
30                VE or T.
31
32             3. The relationship between response and T depends upon the level of C, and
33               by logical extension upon VE.  For example, the  effect of duration of

       February 1994                            9-17       DRAFT-DO NOT QUOTE OR CITE

-------
 1               exposure upon response will be different for high- and low-concentration
 2               exposures.
 3
 4            4. The relationship between response and each of the exposure variables is
 5               curvilinear over a wide range of exposure conditions, although it may
 6               appear linear over certain narrow ranges of exposure.
 7
 8            5. With increasing duration of exposure (and possibly with increasing
 9               concentration), the FEVj response may approach a plateau in humans.
10               Some evidence exists suggesting that the level of the time-dependent
11               response plateau is a function of C.  Respiratory symptom responses may
12               behave similarly, but have not been adequately studied, and this plateau has
13               not been observed in animal studies or for BAL endpoints.
14

15          Models that meet these criteria to a greater or lesser degree and that have been utilized

16     for purposes of prediction include the following.
17
18            1. Polynomial models including linear or second order terms of C, T, or VE,
19               but not including crossproduct terms.  These  models describe linear or
20               curvilinear functions, with each exposure variable having a different
21               weight.  They do not allow the level of C, however, to affect the
22               relationship between response and T, nor do they allow a plateau in
23               response.  There is no simple biological interpretation of individual terms
24               in such a model.
25
26            2. Polynomial models utilizing linear or higher order terms of the variable
27               C x T x  VE.  The product C X T x  VE is conceptually pleasing in that
28               it represents the total inhaled dose of 0$ for a given exposure and in the
29               past has been  referred to as "effective dose". These models  describe a
30               linear or curvilinear function in which the level of C affects  the
31               relationship between response and T, but do not allow the individual
32               variables to be weighted differently. These models do not allow a plateau
33               in response.
34
35            3. Exponential model utilizing C XT ajs the exposure variable (model
36               previously tested only at constant Vg).  This model describes a nonlinear
37               exposure-response relationship, with C and T having equal weight,
38               allowing the level of C to affect the relationship between response  and T.
39               This model does not allow a plateau in response.
40
41            4. Cumulative normal probability model or logistic model utilizing Cy XT as
42               the exposure variable (models previously tested only at constant VE).
43               These models describe sigmoid-shaped exposure-response relationships with
44               C having a different weight than T  (identified by the exponent y) and allow
45               the level of C to affect the relationship between response and T. These

        February 1994                            9-18       DRAFT-DO NOT QUOTE OR CITE

-------
 1               models do not allow the level of the plateau in response to be a function of
 2               concentration.
 3
 4            5.  Logistic model of CxT as the exposure variable, with C having a further
 5               effect upon the level of the time-dependent plateau (model previously tested
 6               only at constant VE).  This model describes a sigmoid-shaped exposure-
 7               response relationship which allows C to have a different weight than T and
 8               allows the level of C to affect the relationship between response and T.
 9               This model describes a concentration-dependent plateau in response.
10               A model of this form has recently been found to describe the relationship
11               between change in FEVj and C and T for heavily exercising humans over
12               a range of exposure conditions.
13
14          Each of the above models has been found under some circumstance to describe the
15     relationship between exposure and response for a particular data set.  Most single data sets,
16     however,  do  not include a wide enough range of data to adequately test the performance of a
17     particular model across  a wide range of exposure conditions or to identify an appropriate
18     exposure metric.  In particular, recent efforts have focussed upon the relationship between
19     response and C and T at constant VE. No definitive work has addressed the  modeling of
20     response and VE  for a given endpoint or for consideration of VE changing as a function of
21     T.  Because animal and human studies are often conducted at different relative levels of
22      VE and because techniques to mathematically adjust for these differences are  only now being
23     developed, efforts to compare responses across species or to develop extrapolation models
24     have been hampered.
25          Evidence indicates that, for humans and animals, the exposure-response relationship of
26     BAL and pulmonary function outcomes may be modified by previous recent exposure to O3,
27     and  the relationship for FEVj changes in humans may be modified by age. Previous
28     exposure to O3 has  not  been included in any exposure-response models.  For young adults,
29     the modification of the exposure-response relationship by age has been modeled.
30          In summary, no single universal model form has been identified which accurately and
31     precisely describes the relationship between population exposure and response under all
32     circumstances. In general, the ability of a predictive model based on one study to predict
33     responses from an independent study has not been studied adequately.  For purposes of
34     prediction or risk estimation, the adequacy of fit of a given model in a given data set and the
        February 1994                            949      DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
size and representativeness of the sample should be assessed.  Extrapolation beyond the range
of observed data introduces additional uncertainty into predictions or risk estimates.

9.2.5.1  Prediction and Summary of Mean Responses
     A selection of reports in which models of mean FEV^ or BAL response have been
developed is listed below, along with figures  summarizing examples of predicted exposure-
response relationships.  Following this section on prediction of mean responses will be a
further section which describes the individual responses within the population.
       1.  Hazucha (1987) predict mean FEVj decrements in humans as a function of
          C (0.0 to 0.75 ppm) for four levels of VE for 2-h exposures (Figure 9-1).
                         110
                                        0.2          0.4         0.6
                                           Ozone Concentration (ppm)
                                                                    08
        Figure 9-1. Mean predicted postexposure to preexposure changes in forced expiratory
                   volume in 1 s (xlOO%) following 2-h exposures to ozone with intermittent
                   exercise.
        Source: Hazucha (1987).
  1
  2
  3
        2. McDonnell and Smith (1994) predict mean ¥EVl decrements in heavily
           exercising humans as a function of C (0.0 to 0.4 ppm) and T (1.0 to
           6.6 h) (Figure 9-2).
        February 1994
                                         9-20
DRAFT-DO NOT QUOTE OR CITE

-------
              -5
               0.00    0.10   0.20    0.30   0.40    0.50
                         Ozone (ppm)
                                                           Time (hours)
       Figure 9-2.  Predicted mean decrements in forced expiratory volume in 1 s for 1- and
                   2-h exposures to ozone with intermittent heavy exercise (A) and 6.6-h
                   exposures with moderate prolonged exercise (B).

       Source:  McDonnell and Smith (1994).
 1
 2
 3
 4
 5
 6
 7
 8
 9

10
11
12
13
14
15
16
17
       3.  Highfill et al. (1992) predict the BAL responses of resting rats and guinea
           pigs as a function of C (0.0 to 0.8 ppm) and T (2 to 8 h) (Figure 9-3).

       4.  Tepper et al. (1994) predict the FVC changes  as a function of C (0.0 to
           0.8 ppm) and T (2 to 7 h) for exposures conducted with rats breathing at
           three times resting VE (Figure 9-4).

     Other reports in which models are developed or that contain data potentially useful for
further development or testing of models  are listed below.
        5.  Seal et al. (1993) present data which would allow modeling of
           FEV^ decrements in humans as a function of C (0.0 to 0.4 ppm) for 2-h
           exposures with moderate exercise.

       6.   Folinsbee et al.  (1978) predict lung function changes in humans as a
           function of C (0.0 to 0.50 ppm) and VE (10 to 65 L/min) for 2-h
           exposures.
       February 1994
                                        9-21
DRAFT-DO NOT QUOTE OR CITE

-------
                             Rats
                                                          Guinea Pigs
                 600-
                 500
                 400
                 300-
                 200-
                 100-
                      — o2h
                      — a4h
                   -  —*8h
0    02    0.4    0.6
        Ozone (ppm)
                                                       02   0.4    0.6
                                                          Ozone (ppm)
Figure 9-3.  Derived means of BAL protein (BALP) denoted by symbols and the
            exponential model shown by lines as time of exposure varies from 2 to 8 h.

Source:  Highfill et al. (1992).
                       i Observed
                       i Predicted
                                  Ozone (ppm)
Figure 9-4. Predicted mean forced vital capacity for rats exposed to ozone while
            undergoing intermittent carbon dioxide-induced hyperpnea.

Source: Tepper et al. (1994).
February 1994
                     9-22
DRAFT-DO NOT QUOTE OR CITE

-------
 1           7.    Adams et al. (1981) predict lung function changes in humans as a
 2                 function of the product of C x T X VE for C  = 0 to 0.4 ppm, T = 30
 3                 to 80 min, and VE = 33 or 66 L/min.
 4
 5           8.    Rombout et al. (1989) predict the concentration of protein in BAL fluid
 6                 of rats as a function of C (0.25 to 4.0 mg/m3) and T (0 to 12 h) for
 7                 daytime and nighttime exposures.
 8
 9           9.    Highfill and Costa (1994) compare the fits of quadratic, exponential, and
10                 sigmoid-shaped models to published human lung function data and
11                 O3 laboratory animal BAL data.
12
13          The concentrations and durations of exposure which  result in mean lung function and
14     airway reactivity changes and mean changes in BAL endpoints in humans can be summarized
15     as follows. The estimates are based directly upon studies  that were performed at exposure
16     conditions of interest rather than upon mathematical models.
17          Pulmonary function decrements are generally observed in healthy subjects (8 to
18     45 years of age) after 1 to 3 h of exposure as a function of the level  of exercise performed
19     and the O3 concentration inhaled during the exposure. Group mean data from numerous
20     controlled human exposure and field studies indicate that,  in general, statistically significant
21     pulmonary function decrements beyond the range of normal measurement variability (e.g.,
22      > 3 to 5 % for FEV^ occur:
23          (1) at >0.5 ppm when at rest,
24          (2) at >0.37 ppm with light exercise (slow walking),
25          (3) at >0.30 ppm with moderate exercise (brisk walking),
26          (4) at >0.18 ppm with heavy exercise (jogging), and
27          (5) at >0.16 with very heavy exercise (running).
28
29          For  a number of studies,  small group mean changes (e.g., <5%) in FEVl5 the medical
30      significance of which is a matter of controversy, have been  observed at lower
31      O3 concentrations than those listed above. For example,  data from one specific study
32      indicate that FEV^ decrements  occur with very heavy exercise in healthy adults at 0.15  to
33      0.16 ppm O3, and data from two studies indicate that such effects may occur in healthy
34      adults at levels as low as 0.12 ppm.   Also, pulmonary function decrements have been
35      observed in children and adolescents at concentrations of  0.12 and 0.14 ppm  O3 with heavy
36      exercise.  Pulmonary function decrements were observed  at 0.12 ppm O3 in healthy young
37      adults undergoing heavy exercise in a recent study.  Some individuals may experience FEV1

        February  1994                           9-23      DRAFT-DO NOT QUOTE OR CITE

-------
 1     decrements in excess of 15% under these exposure conditions, even when the mean
 2     decrement is less than 5 %.
 3          An increase in the incidence of cough has been reported at levels as low as 0.12 ppm
 4     O3 hi healthy adults during exposure with very heavy exercise and at levels as low as
 5     0.18 ppm with heavy exercise.  Other respiratory  symptoms, such as pain on deep
 6     inspiration, shortness of breath, and lower respiratory scores (a combination of several
 7     symptoms), have been observed at 0.16 to 0.18 ppm O3 with heavy and very heavy exercise.
 8          Increases in nonspecific airway responsiveness have been observed following exposure
 9     to 0.40, but not 0.20 ppm O3 at rest, and have been observed at concentrations as low as
10     0.18 ppm but not to 0.12 ppm O3 during exposure with very heavy exercise.
11          Increases in BAL protein concentration and/or PMNs have been observed following
12     exposure to 0.20, 0.30, and 0.40 ppm with very heavy exercise and have not been studied at
13     lower concentrations for 1 to 3 h exposures.
14          For exposures of healthy subjects performing moderate exercise during longer duration
15     exposures (6 to 8 h), 5%  group mean decrements  hi FEVj were observed at:
16          (1) 0.08 ppm O3 after 5.6 h,
17          (2) 0.10 ppm O3 after 4.6 h, and
18          (3) 0.12 ppm O3 after 3 h.
19
20     For these same subjects, 10% mean FEVj decrements were observed at 0.12 ppm O3 after
21     5.6 and 6.6 h.  As in the shorter duration studies  some individuals experience changes larger
22     than those represented by the group mean changes.
23          Increases in the respiratory symptoms (i.e.,  cough, shortness of breath, and pain on
24     deep inspiration) and hi nonspecific airway responsiveness have been observed following
25     exposure to 0.08, 0.10, and 0.12 ppm O3 for 6.6 h with moderate levels of exercise.
26          Increases in nonspecific airway responsiveness have been observed at 0.08, 0.10, and
27     0.12 ppm O3 during 6.6 h exposures with moderate levels of exercise.
28          Increases in BAL protein and/  or PMN have also been observed following 6.6 h
29     exposure to 0.08 and 0.10 ppm O3.  No 6- to 8-h exposure studies have been conducted
30     using O3 concentrations less than 0.08 ppm.
31
32

       February 1994                           9-24      DRAFT-DO NOT QUOTE OR CITE

-------
1     9.2.5.2  Prediction and Summary of Individual Responses
2          It is well known that considerable inter-individual differences in the magnitude of
3     response to O3 exposure exist.  The individual lung function and, to a lesser extent,
4     respiratory symptom responses to O3 have been demonstrated to be reproducible over a
5     period of time, indicating that some individuals are consistently more responsive to O3 than
6     are others. The basis for these differences are not known, with the exception that young
7     adults have been observed to be more responsive than older adults (see Figure 9-5).
8
                         1.0
                     
-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
     Calculation of group mean responses for a population that includes both more and less

responsive individuals is useful for making inferences regarding the probability that a

population effect is present or absent for a given exposure.  Since the frequency distribution

of individual responses to O3 changes with changing exposure conditions, however,

knowledge of the mean and variance of population responses does not provide reliable

information on the distribution of individual responses for a given exposure, and hence, is

not particularly useful for estimating risks to members of the population. One method  of

presenting individual data is illustrated in Figure 9-6 in which histograms are presented for

individual responses of subjects participating in four 6.6-h studies of low-levels  03 exposure.
                     Distribution  of Percent Change in FEV,
•
^^™

N-87
AIR
:—
•

N-60
0.08
                                                           N-32
                                                            0.10
                                                                       N = 49
                                                                         0.12
                                       Percent of Subjects
           > 10% Drop

           •HFEV,
                                26%
  31%
46%
       Figure 9-6.  The distribution of response for 87 subjects exposed to clean air and at least
                   one of 0.08, 0.10, or 0.12 ppm O3 is shown here.  The O3 exposures lasted
                   6.6 h, during which time the subjects exercised for 50 min of each hour
                   with a 35-min rest period at the end of the third hour.  The abscissa
                   indicates the number of subjects.  The bar labels on the ordinate indicate
                   decreases in FEVj, expressed as percent change from baseline. For
                   example, the bar labeled 10 indicates the number of subjects with a
                   decrease in FEV,  of >5% but <10%; the bar labeled -5 indicates
                   improvement in FEVj of >0% but <5%.  The rectangle across the bottom
                   of the graph indicates the percentage of subjects at each O3 concentration
                   with a decrease of FEVX in excess of 10%.

       Source:  Folinsbee et al. (1991).
       February 1994
                                        9-26
DRAFT-DO NOT QUOTE OR CITE

-------
1          Another method that allows interpolation between observed data points involves
2     definition of the effect of interest (e.g., a 10% decrement in FEVj) and modeling of the
3     proportion of individuals who experience such an effect as a function of exposure conditions.
4     For example, Figure 9-7 shows the prediction of the proportion of individuals (humans)
5     experiencing 10% FEVj decrements and respiratory symptom responses (Figure 9-8) as a
6     function of C (0.0 to 0.4 ppm O3) for three studies conducted at either 1 or 2 h of exposure
7     with heavy exercise.
8
                                                                    • Avoletal. (1984)
                                                                    D Kulleetal. (1985)
                                                                    A McDonnell etal. (1983)
                0.0
                   o.o
0.1         0.2         0:3
Ozone Concentration (ppm)
       04
      Figure 9-7.  Proportion of heavily exercising individuals predicted to experience a 10%
                   decrement in forced expiratory volume in 1 s following a 1- or 2-h exposure
                   to ozone.
      Source: U.S. Environmental Protection Agency (1989).
1           Predictions of the proportion of individuals experiencing a 5, 10, or 15% FEVl
2      decrement as functions of C (0.0 to 0.12 ppm), T (1 to 6.6 h), and age (18 to 34 years) for
3      exposures with moderate exercise are shown in Figure 9-9.
4           As an example of differences between the mean and individual responses, it was stated
5      earlier that exposure for 5.6 h to 0.08 ppm was the shortest duration for  which a 5 % mean
       February 1994
                 9-27
DRAFT-DO NOT QUOTE OR CITE

-------
                                             a  Kulleetal. (1985)

                                             A  McDonnell et al. (1983)
                               0.1          0.2         0.3
                               Ozone Concentration (ppm)
Figure 9-8. Proportion of heavily exercising individuals predicted to experience mild
            cough following a 2-h ozone exposure.

Source: U.S. Environmental Protection Agency (1989).
                •5
                g
                          0.1
0.2  0.3   0.4  0.5   0.6  0.7
       Dose (ppm x hours)
                                                         0.8   0.9  1.0
Figure 9-9.  Proportion of moderately exercising individuals exposed to ozone for
             6.6 h predicted to experience 5, 10, or 15% decrements in forced expiratory
             volume in 1 s as a function of CxT for age = 24 years.

Source: McDonnell et al. (1994).
February 1994
           9-28
DRAFT-DO NOT QUOTE OR CITE

-------
 1      decrement in FEVj was observed.  For those same exposure conditions, 41, 17, and 10% of
 2      the subjects studied experienced FEVl  decrements larger than 5, 10, and 15%, respectively.
 3           The clinical significance of individual responses to O3 exposure depends on the
 4      magnitude of the changes in spirometry (e.g., FEVj, FVC) and airway responsiveness to
 5      nonspecific stimuli, the severity of respiratory symptoms (e.g., cough and pain on deep
 6      breath), and the duration of the response.  In nonasthmatic individuals, O3-induced changes
 7      in specific airway resistance (SR^), a measure of large airway narrowing, are small and of
 8      minimal significance. Asthmatics, however, often have baseline airway narrowing and
 9      experience larger changes in SR^ upon exposure to O3 than do nonasthmatics.  Because of
10      these baseline differences, the clinical significance of increases in SR^ depends both upon
11      percent changes from baseline and absolute increases in SR^. Table 9-1 categorizes these
12      physiological responses to O3 exposure as normal (or none), mild, moderate, or severe.
13      Qualitative relationships among the categories provide an indication of the severity of the
14      response.
15
        TABLE 9-1. GRADATION OF PHYSIOLOGICAL RESPONSES TO SHORT-TERM
                                      OZONE EXPOSURE3
Response
Cough symptom
Pain on deep
breath symptom

FEV! (also FVC)
SRaw (asthmatics)
Airway
responsiveness
Duration of response
None/Normal
Infrequent cough
None

Within normal
range (±3%)
Within normal
range (±20%)
Within person's
normal range
None
Mild
Cough with deep
breath or FVC test
Discomfort just
noticeable on FVC
test

FEVi decreased
less than 10%
Increase less than
100%
Increase less than
100%
Less than 4 h
Moderate
Frequent
spontaneous cough
Marked discomfort
on exercise or
FVC test

FEVj decreased
more than 10%
but less than 20%
Increase up to
200% or up to
15 cm H2O/s
Increase up to
300%
Less than 24 h
Severe
Persistent
uncontrollable cough
Exercise or
breathing tests cause
chest pain; FVC
tests cannot be
performed properly
FEV1 decreased
more than 20%
Increase greater than
200% or more than
15 cm H2O/s
Increase greater than
300%
Longer than 24 h
        See text for discussion.
       February 1994
9-29
DRAFT-DO NOT QUOTE OR CITE

-------
 1     9.3   HEALTH EFFECTS OF LONG-TERM EXPOSURES
 2          In both humans and test animals, the response to a single O3 exposure can nominally be
 3     characterized by lung dysfunction, lung cell injury and inflammation, and leakage of plasma
 4     proteins into the airspace lumen.  However, when such an exposure is repeated for several
 5     consecutive days, these effects appear to wane, suggesting adaptation or the development of
 6     tolerance to the continued intermittent challenge.  In spite of this apparent state of tolerance,
 7     long-term O3 exposures have been linked to subtle pulmonary effects, some of which have
 8     irreversible components, thereby enhancing concern about chronic effects. The following
 9     section will provide an overview attempting to synthesize the current understanding of the
10     phenomenon of adaptation during brief repeated exposures and the evidence for potential
11     health impairments resulting from protracted exposures to this oxidant.
12
13     9.3.1    Repeated Exposures
14          It is well established that a brief exposure of laboratory rodents to a minimally  toxic
15     concentration of O3 will protect the animals from a subsequent lethal challenge of O3 a week
16     later. This phenomenon, called tolerance, bears a similarity to the pattern of attenuated
17     nonlethal effects (sometimes referred to as "adaptation") observed in both human volunteers
18     and animals when exposed to episodic levels of O3 (<0.5 ppm) for 1 to 7 h/day over a
19     succession of 5 or more days.  Generally, over a 5-day exposure period,  the effects of Day 1
20     are accentuated on Day 2 and diminish thereafter. Attenuation of the functional effects
21     include spirometric deficits and associated symptoms  as well as irritative  alterations of
22     breathing; nonspecific airway responsiveness, however, does not revert to normal  levels.
23     Measures of tissue effects which attenuate include inflammation and impaired phagocytic
24     capabilities of  alveolar macrophages. However,  some evidence from animal studies  suggests
25     that tissue alterations persist, although the observed changes may be part  of a transition  to a
26     chronically affected state of the lung.  Thus, in general, cell associated indicators of injury or
27     damage within the lung appear to diminish in spite of the continued O3 exposure.
28          A number of mechanisms have been shown to be involved in the evolution of this
29     "adapted" state. These mechanisms range from the replacement of sensitive cells  in the
30     alveolar lining (epithelium) by more resistant cells (with or without a thickened fluid barrier
31     on the lumenal surface) to the enhancement of antioxidant metabolism providing cell

       February 1994                           9-30      DRAFT-DO NOT QUOTE  OR CITE

-------
 1     resistance and more biochemical defenses at the lung surface. However, controlled human
 2     studies show that after a one-week period without O3 exposure,  subjects regain their
 3     spirometric responsiveness to O3 challenge, although this abrupt transition between
 4     unresponsiveness and responsiveness appears less distinct in field-related studies.  For
 5     example, studies of Southern Californians suggest that they are  significantly less responsive
 6     to the spirometric effects of an acute episodic-like controlled challenge with O3 when studied
 7     for a period after the "high" O3 season than after the relatively  "low" O3 season.  Likewise,
 8     there is some evidence that O3-exposed urban populations are also somewhat more resistant
 9     to the oxidant than populations that receive minimal exposure.  This would appear to be in
10     conflict with hospital admissions data suggesting the aggravation of respiratory diseases, like
11     asthma, within such populations.  It remains to be shown whether these latter data reflect the
12     responsiveness of a sensitive subpopulation,  perhaps less  adapted or having less reserve
13     function.
14
15     9.3.2   Prolonged Exposures
16          Most long-term exposure studies in animals have evaluated structural and functional
17     changes.  In the few investigations of the immune system or antibacterial host defenses,
18     prolonged exposures of animals either caused no effects or did not increase the magnitude of
19     effects observed after acute exposures. Thus,  the following discussion centers on the larger
20     body of knowledge on other endpoints.
21          Epidemiologic studies attempting to associate chronic lung effects in humans with long-
22     term O3 exposure have yet to provide unequivocal evidence that such a linkage exists.  Most
23     studies have been cross-sectional in design and have been compromised by incomplete
24     control of confounding variables and inadequate  exposure information.  Other studies have
25     attempted to follow variably exposed groups prospectively. Studies of such design have been
26     conducted in communities of the Southwest Air Basin as  well as in Canada where
27     comparisons could be drawn between lung function changes over several years in populations
28     from high or low oxidant pollution.  The findings suggest small, but consistent decrements in
29     lung function among inhabitants of the more highly polluted communities.  However,
30     associations between O3 and other copollutants and, in  some cases, problems  with study
31     population loss undermine the confidence in the  study conclusions.  Likewise, recent

       February  1994                            9.3!      DRAFT-DO NOT QUOTE  OR CITE

-------
 1      associations found between O3 and the incidence/severity of asthma over a decade of study,
 2      though derived from well-designed studies, also tend to be weakened by colinearity of O3
 3      with other air pollutants.  Nevertheless, in all of the studies assessing lung function, the
 4      pattern of dysfunction associated with the long-term exposure has been consistent with the
 5      small airway lesions seen in animal studies.
 6           The advantage of animal studies is the ability to examine closely the distribution and
 7      intensity of the O3-induced lesions throughout the  respiratory tract.  Indeed, cells of the nose
 8      like the distal lung are clearly affected by 03.  Perhaps of greater health concern would be
 9      the lesions that occur in the centriacinar regions of the lung where the alveoli meet the end-
10      airways.  Altered function of the distal airways, the proximal conduits of air to the gas-
11      exchange regions, can result in reduced communication of fresh air with the alveoli, air-
12      trapping, and reduced oxygenation of the blood.  In fact,  chronic O3 lesions as found in
13      animal studies are reminiscent of the earliest lesions found in autopsied cigarette  smokers,
14     many of whom would have theoretically progressed to chronic obstructive  lung diseases.
15           As shown in Figure 9-10,  the temporal pattern of effects during and after a chronic
16     exposure is complex. During the early days of exposure, the end-airway lumenal and
17     interstitial inflammation peaks, and thereafter appears to subside at a lower plateau of activity
18     sometimes referred to as a "smoldering" lesion. Several cytokines remain elevated beyond
19     the apparent adaptation phase of the response and may be conceptually linked to  the
20     development of chronic lesions in the distal lung.   To date, however, a clear association of
21     these BAL-derived mediators and cells with long-term toxicity has not been demonstrated.
22     Some evidence of molecular changes within the matrix of the lung may also link to the
23     chronic  effects, but these too remain poorly defined.  When exposures to O3 continue for
24     weeks or months, the diminished O3-induced exudative response in the distal bronchoalveolar
25     areas is  supplanted by hyperplastic epithelia hi the alveoli and end-airways.  Damaged cells
26     in centriacinar alveoli are replaced by metabolically active progenitor cells that are more
27     resistant to oxidant challenge.  Junctional areas between conducting and gas exchange
28      regions, where the O3 changes are typically most intense, also undergo epithelial hyperplasia,
29     giving the appearance that airway cells are extending into the mouth of the alveolus, hence
30     the term "bronchiolization". The functional result of this concentration-dependent process is
31      the effective elongation of distal bronchioles, which functionally may alter air distribution

        February 1994                             9-32      DRAFT-DO NOT QUOTE OR CITE

-------
                     CD
                     CO
                     O
                     a.
                     
-------
 1     that similar end-airway lesions result from ambient O3 exposure in areas with high oxidant
 2     exposures.
 3           Studies of prolonged exposures in monkeys and rats reveal generally similar
 4     morphologic responses, although it appears that the monkey exhibits somewhat more tissue
 5     injury than does the rat under roughly similar exposure conditions.  Although adequate
 6     dosimetric data are not yet available for a direct comparison of interspecies sensitivity, the
 7     monkey, with its similarity in distal airway structure relative to the human, provides data that
 8     may best reflect the potential effects in humans.  As such, monkeys exposed to O3 at
 9     0.15 ppm for 8 h each day for 6 to 90 days exhibit significant distal airway remodeling. Rats
10     show similar but more modest changes at 0.25 ppm after exposures of longer duration, up to
11     18 mo and beyond in both species (near-lifetime in the case of the rat).  The chronic distal
12     lung and airway alterations appear consistent with incipient peribronchiolar fibrogenesis
13     within the interstitium. Attempts to correlate functional deficits have been variable, perhaps
14     due in part to the degree and distribution of the lesions and the general insensitivity of most
15     measures of the distal lung function.  The  interstitial changes may progress, however.
16     Moreover, one recent primate study revealed evidence that intermittent challenge with a
17     pattern of  O3 exposure more reflective of seasonal episodes, with extended periods of clean
18     air in between extended periods of O3, actually leads to greater injury.  The reasons for this
19     are unclear, but may relate to the known loss of tolerance which occurs in both humans and
20     animal test species with removal of the oxidant burden.
21           In conclusion, the collective toxicologic data on chronic exposure to O3 garnered in
22     animal exposure  and human population studies have some ambiguities.  What is clear is that
23     the distribution of the O3 lesions is roughly similar across species, is, in part,  concentration
24     dependent (and perhaps time or exposure pattern dependent), and, under certain conditions,
25     has irreversible structural attributes.  What is unclear is whether ambient exposure scenarios
26     encountered by humans result in similar lesions and whether there are  resultant functional or
27     unpaired health outcomes, particularly since the human exposure scenario may involve much
28     longer exposures than can be  studied in the laboratory. The epidemiologic lung function data
29     generally parallel those of the animal studies, but they lack the confidence of O3 exposure
30     history and are frequently confounded by personal or copollutant variables.
31

       February 1994                           9.34      DRAFT-DO NOT QUOTE OR CITE

-------
 1     9.3.3   Genotoxicity and Carcinogenicity of Ozone
 2          Numerous in vitro exposure studies suggest that O3 either has a weak or no potential to
 3     cause mutagenic, cytogenetic, or cellular transformation effects.  Most of these experiments
 4     utilized high concentrations of O3 (>5.0 ppm). Because of the exposure systems used, there
 5     are unknowns about the formation of artifacts and the dose of O3.  Therefore, these studies
 6     are not very useful in health assessment.  Cytogenetic effects have been observed in some,
 7     but not all, laboratory animal and human studies of short-term O3 exposure.  However, well-
 8     designed human clinical cytogenetic studies were negative.
 9          Until recently, in vivo exposure studies of carcinogenicity,  with and without
10     co-exposure to known carcinogens, were either negative or ambiguous. A well-designed
11     cancer bioassay study has recently been completed by the National Toxicology Program
12     (NTP) using male and female F344/N rats and B6C3FJ mice.  Animals were exposed for
13     2 years to 0.12, 0.5, and 1.0 ppm O3  (6 h/day, 5 days/week). A similar lifetime exposure
14     was conducted, but 0.12 ppm was not used.  The NTP evaluated the weight-of-evidence for
15     this study.  They  found "no evidence" of carcinogenicity in rats.  They reported  "equivocal
16     evidence" of carcinogenicity in O3-exposed male mice and "some evidence" of carcinogenic
17     activity in O3-exposed female mice. The increases in adenomas and carcinomas were only
18     observed in the lungs.  There was no concentration response.  In the male mice, the
19     incidence of neoplasms in the 2-year study was not significantly  elevated by O3 and was
20     within the range of historical controls.  Also, the lifetime exposure did not significantly
21     increase the incidence of neoplasms, even though the incidence of carcinomas was increased.
22     In the female mice, a 2-year (but not lifetime) exposure to 1.0 ppm only increased the
23     incidence of animals with neoplasms.  When the female mouse data from the two exposure
24     regimens (at 1.0 ppm) were combined, there was a statistically significant  increase (about a
25     doubling) in neoplasms.  In a companion study, male rats were treated with a tobacco
26     carcinogen and exposed for 2 years to 0.5 ppm O3.  Ozone did not affect the response and
27     therefore had no tumor promoting activity.
28          In summary, only chronic exposure to a high concentration of O3 (1.0 ppm) has been
29     shown to evoke a limited degree of carcinogenic activity in one  strain of mice. Rats were
30     not affected. Furthermore, there was no concentration response, and there is inadequate
        February 1994                            9-35      DRAFT-DO NOT QUOTE OR CITE

-------
 1     information from other research to provide a mechanistic support for the finding in mice.
 2     Thus, it is not justified to extrapolate these mouse data to humans at the present time.
 3
 4
 5     9.4   COMBINED POLLUTANT EXPOSURES
 6          In the ambient air, people are exposed to mixtures of pollutants, making it important to
 7     understand interactions.  Epidemiological studies, which inherently evaluate O3 as part of
 8     complex mixtures, are discussed in other subsections dealing with classes of effects. In the
 9     laboratory it becomes possible to sort out the role of O3 in simple mixtures.  Complex
10     mixtures are typically not investigated in the laboratory because even if only six pollutants
11     were involved, the experimental design required to unequivocally sort out which pollutant or
12     pollutant interactions were responsible for the responses or portions of the responses could
13     require as many as 719 additional separate experiments, and this would be true only if the
14     concentrations of the six  pollutants remained the same.
15          The summary will focus only on binary mixtures since these are by far the predominant
16     type of experiments.   Responses to a binary pollutant mixture may represent the sum of the
17     independent responses to the two chemicals (i.e., an additive response).  If there is some
18     interaction between either the two responses or the two pollutants, the resultant response
19     could be larger than additive  (synergism) or smaller than additive (antagonism).  Interaction
20     between pollutants could  result in the production of a more or less toxic byproduct.
21     Alternatively,  the response to one pollutant could magnify the response to the other pollutant
22     or could interfere with or block the action of the other pollutant.  Binary mixture studies fall
23     into two categories, simultaneous  and sequential exposures.  In the simultaneous exposures,
24     both the  responses and the pollutants can interact. In the sequential exposures, it is primarily
25     the responses that would  interact.
26          In general, controlled human studies of O3 mixed with other pollutants show no more
27     than an additive response with symptoms or spirometry as an endpoint. This applies to O3 in
28     combination with NO2, SO2,  H2SO4, HNO3, or CO.  Indeed,  at the levels of copollutants
29     used in human exposure  studies, the responses can be attributed primarily to O3. In one
30     study, exposure to O3 increased airway responsiveness to SO2 in asthmatics.  Similarly, other
       February 1994                           9.36       DRAFT-DO NOT QUOTE OR CITE

-------
 1     pollutants that may increase airway responsiveness could augment the effect of 03 on airway
 2     responsiveness.
 3          The relatively large number of animal studies of O3 in mixture with NO2 and
 4     H2SO4 show that additivity, synergism, and antagonism can result, depending upon the
 5     exposure regimen and the endpoint studied. The numerous observations of synergism are of
 6     concern, but the interpretation of most of these studies relative to the real world is
 7     confounded by unrealistic exposure designs.  For example, often ambient concentrations of
 8     O3 were combined with levels of copollutants substantially higher than ambient, creating the
 9     possibility that mechanisms of toxicity unlikely in the real world contributed to the
10     experimental outcome.  Nevertheless, the data support  a hypothesis that coexposure to
11     pollutants, each at innocuous or low-effect levels, may result in effects of significance.
12
13
14     9.5   CONCLUSIONS
15          This section summarizes  the primary conclusions  derived from an integration of the
16     known health effects of O3 provided by animal lexicological, human clinical, and
17     epidemiological studies.
18
19     1.  What are the health effects of short-term (<8 h) exposures to ozone?
20          Acute O3 exposure of laboratory animals and humans causes  changes in pulmonary
21     function,  including tachypnea (rapid,  shallow breathing), decreased lung volumes and flows,
22     and increased  airway responsiveness to nonspecific stimuli.  Increased airway resistance
23     occurs in both humans and laboratory animals, but typically at higher exposure levels than
24     other functional endpoints.  In addition, adult human subjects experience O3-induced
25     symptoms of airway irritation  such as cough or pain on deep inspiration.  The changes in
26     pulmonary function and respiratory symptoms occur as a function of exposure concentration,
27     duration,  and level of exercise.  Recovery of pulmonary function and the absence of
28     O3-induced symptoms is usually  complete within 24 h of the end of exposure, although other
29     responses may persist somewhat longer.
30
       February  1994                           9.37       DRAFT-DO NOT QUOTE OR CITE

-------
 1            •  Pulmonary function decrements are generally observed in healthy subjects
 2              (8 to 45 years of age) after 1 to 3 h of exposure as a function of the level of
 3              exercise performed and the O3 concentration inhaled during the exposure.
 4              Group mean data from numerous controlled human exposure and field
 5              studies indicate that, in general, statistically significant pulmonary function
 6              decrements beyond the range of normal measurement variability (e.g., 3 to
 7              5% forFEVi) occur
 8
 9              (1) at >0.5 ppm when at rest,
10              (2) at >0.37 ppm with light exercise (slow walking),
11              (3) at >0.30 ppm with moderate exercise (brisk walking),
12              (4) at >0.18 ppm with heavy exercise (easy jogging), and
13              (5) at >0.16 ppm with very heavy exercise (running).
14
15              For a number of studies, small group mean changes (e.g., <5%) in FEVl5
16              the medical significance of which is a matter of controversy, have been
17              observed at lower O3  concentrations than those listed above. For example,
18              data from one specific study indicate that FEVj decrements occur with very
19              heavy exercise in healthy adults at 0.15 to  0.16 ppm O3, and data from two
20              studies indicate that such effects may occur in healthy adults at levels as
21              low as 0.12 ppm.  Also, pulmonary function decrements have been
22              observed in children and adolescents at concentrations of 0.12 and
23              0.14 ppm O3 with heavy exercise.  Pulmonary function  decrements were
24              observed at 0.12 ppm O3 in healthy young adults undergoing heavy exercise
25              in a recent study.  Some individuals within a study may experience FEVj
26              decrements in excess of 15 %  under these exposure conditions, even when
27              the group mean decrement is  less than 5 %.
28
29            • For exposures of healthy subjects performing moderate exercise during
30              longer duration exposures (6 to 8 h), 5% group mean decrements hi FEVj
31              were observed at
32
33              (1)  0.08 ppm O3 after 5.6 h,
34              (2)  0.10 ppm O3 after 4.6 h, and
35              (3)  0.12 ppm O3 after 3 h.
36
37              For these same subjects, 10% group mean FEVl  decrements were observed
38              at 0.12 ppm  O3 after 5.6 and 6.6 h.  As in the shorter duration studies,
39              some individuals experience changes larger than those represented by the
40              group mean changes.
41
42            • An increase in the incidence of cough has  been reported at
43               O3 concentrations  as low as 0.12 ppm hi healthy adults during 1 to 3 h of
44              exposure  with very heavy exercise. Other respiratory symptoms,  such as
45              pain on deep inspiration, shortness of breath,  and lower respiratory scores
46               (a combination of several  symptoms), have been observed at 0.16 to
47               0.18 ppm O3 with heavy and very heavy exercise.  Respiratory symptoms

        February 1994                            9-38       DRAFT-DO NOT QUOTE OR CITE

-------
 1               have also been observed following exposure to 0.08, 0.10, and 0.12 ppm
 2               O3 for 6.6 h with moderate levels of exercise.
 3
 4            •  Increases in nonspecific airway responsiveness have been observed after
 5               1 to 3 h of exposure to 0.40 ppm, but not 0.20 ppm, O3 at rest, and have
 6               been observed at concentrations as low as 0.18  ppm, but not to 0.12 ppm,
 7               O3 during exposure with very heavy exercise. Increases in nonspecific
 8               airway responsiveness during 6.6-h exposures with moderate levels of
 9               exercise have been observed at 0.08, 0.10, and 0.12 ppm O3.
10
11          Acute O3 exposure of laboratory animals and humans disrupts the barrier function of
12     the lung epithelium, permitting materials in the airspaces to enter lung tissue, allowing cells
13     and serum proteins to enter the airspaces (inflammation),  and setting off a cascade of
14     responses.
15
16            •  Increased levels of neutrophils and protein in lung lavage fluid have been
17               observed following exposure of humans to 0.20, 0.30, and 0.40 ppm with
18               very heavy exercise and have not been studied at lower concentrations for
19               1- to 3-h exposures.  Increases in protein and/or neutrophils have also been
20               observed at 0.08 and 0.10 ppm O3 during 6.6-h exposures with moderate
21               exercise; lower concentrations have not been tested.
22
23          Acute O3 exposure of laboratory animals and humans impairs alveolar macrophage
24     clearance of viable and nonviable particles from the lungs and decreases the effectiveness  of
25     host defenses against bacterial lung  infections in animals and perhaps humans.  The ability of
26     alveolar macrophages to engulf microorganisms is decreased in humans exposed to 0.08 and
27     0.10 ppm for 6.6 h with moderate exercise.
28
29     2.  What are the health effects of repeated, short-term exposures to ozone?
30          During repeated acute exposures, some of the O3-induced responses are partially or
31     completely attenuated. Over a 5-day exposure, pulmonary function changes are typically
32     greatest on the second day, but return to control levels by the fifth day of exposure. Most of
33     the inflammatory markers (e.g., neutrophil influx) also attenuate by the fifth day of exposure,
34     but markers of cell damage (e.g., lactate dehydrogenase enzyme activity) do not attenuate
35     and continue to increase.  Attenuation of lung function decrements is reversed following 7 to
36     10 days without O3. Some inflammatory markers are also reversed during this time period,
37     but others still show attenuation even after 20 days without O3.  The mechanisms and

       February 1994                            9.39       DRAFT-DO NOT QUOTE OR CITE

-------
 1     impacts involved in attenuation are not known, although the underlying cell damage continues
 2     throughout the attenuation process.  In addition, attenuation may alter the normal distribution
 3     of O3 within the lung, allowing more O3 to reach sensitive regions, possibly affecting normal
 4     lung defenses (e.g., neutrophil influx in response to inhaled microorganisms).
 5
 6     3.  What are the health effects of long-term exposures to ozone
 7          Exposure to O3 for months and years causes structural changes in several regions of the
 8     respiratory tract, but effects may be of the greatest importance in the centriacinar regions
 9     where the alveoli and conducting airways  meet because this region is typically affected in
10     most chronic diseases of the human lung.  This information on O3 effects in the distal lung is
11     derived from animal lexicological studies because, to date, such data are not available in
12     humans. Epidemiological studies attempting to associate chronic health effects in humans
13     with long-term O3 exposure have yet to provide unequivocal evidence  that such a linkage
14     exists.
15          Chronic exposure of one strain of female mice to high 0$ levels  (1 ppm) caused a
16     small, but statistically significant increase  in lung tumors.  There was  no concentration-
17     response relationship and rats were not affected.   Genotoxicity data are either negative or
18     weak.   Given the nature of the database, the effects in one strain of mice cannot yet be
19     qualitatively extrapolated to humans.  Ozone did not show tumor-promoting activity in a
20     chronic rat study (at 0.5 ppm O3).
21
22     4.  What are the health effects of binary pollutant mixtures containing ozone?
23          Combined data from laboratory animal and controlled human exposure studies on
24     O3 support the hypothesis that coexposure to pollutants, each at low-effect levels, may result
25     in effects of significance.  The data from  human studies of O3 in combination with NO2,
26     SO2, sulfuric acid, nitric acid, or CO show no more than an additive response on lung
27     spirometry or respiratory symptoms.  The larger number of laboratory animal studies
28     with O3 in mixture with NO2 and sulfuric acid show that effects can be additive, synergistic,
29     or even antagonistic, depending upon the  exposure regimen and the  endpoint studied.  This
30     issue of exposure to copollutants remains  poorly understood,  especially with regard to
31     chronic effects.

       February 1994                            9-40       DRAFT-DO NOT QUOTE OR CITE

-------
 1     5.  What population groups are at-risk as a result of exposure to ozone?
 2           Identification of population groups that may show increased  susceptibility to O3 are
 3     based on their (1) biological responses to O3; (2) physiological status (e.g., preexisting lung
 4     disease); (3) activity patterns; (4) personal exposure history; and (5) personal factors (e.g.,
 5     age, nutritional status).
 6           The predominant information on the health effects of O3 noted above comes from
 7     studies on healthy, nonsmoking, exercising subjects, 8 to 45 years of age. These studies
 8     demonstrate that among this group, there is a large variation in sensitivity and responsiveness
 9     to O3, with at least a 10-fold difference between the most and least responsive individuals.
10     Individual sensitivity to O3 may also vary throughout the year, related to seasonal variations
11     in ambient O3 exposure.  The specific factors that contribute to this large intersubject
12     variability, however, remain undefined. Although differences may be due to the dosimetry
13     of O3 in the respiratory tract, available data show little effect on O3 deposition after
14     inhalation  through the nose or mouth.
15           Controlled  studies on mild asthmatics suggest that they have similar lung volume
16     responses  but greater airway resistance changes to 0$ than nonasthmatics. Furthermore,
17     limited data from studies  of moderate asthmatics suggest that they may have greater lung
18     volume responses than nonasthmatics.  Daily life studies reporting an exacerbation of asthma
19     and decrease in peak expiratory flow rates, particularly in asthmatic children, appear to
20     support the controlled studies; however, those studies are confounded by concurrent
21     temperature, particle or aeroallergen exposure, and asthma severity of the subjects or their
22     medication use.  In addition, field  studies of summertime daily hospital admissions for
23     respiratory causes show a consistent relationship between asthma and ambient levels of O3 in
24     various locations in the northeastern United States, even after controlling for independent
25     contributing factors.
26           Other population groups with preexisting limitations in pulmonary function and exercise
27     capacity (e.g., chronic obstructive  pulmonary disease, chronic bronchitis, ischemic heart
28     disease) would be of primary concern in evaluating the health effects of O^.   Unfortunately,
29     not enough is known about the  responses of these individuals to make definitive conclusions
30     regarding  their relative responsiveness to O3. Indeed, functional effects in these individuals
        February 1994                             9_41      DRAFT-DO NOT QUOTE OR CITE

-------
 1     with reduced lung function may have greater clinical significance than comparable changes in
 2     healthy individuals.

 3          Currently available data on personal factors or personal exposure history known or
 4     suspected of influencing responses to O3 are the following.
 5

 6            •  Human studies have identified a decrease in pulmonary function
 7               responsiveness to O3 with increasing age,  although symptom rates remain
 8               similar.  Toxicological studies are not easily interpreted but suggest that
 9               young animals are not more responsive than adults.
10
11            •  Available lexicological and human data have not conclusively demonstrated
12               that males and females respond differently to O3.  If gender differences
13               exist for lung function responsiveness to O3, they are not based on
14               differences in baseline pulmonary function.
15
16            •  There is no compelling evidence to date to suggest that any ethnic or racial
17               group has  a different distribution of responsiveness to O3.  However,  data
18               are not adequate to rule out the possibility of such differences.
19
20            •  Information derived from O3 exposure of  smokers is limited.  The general
21               trend is that smokers are less responsive than nonsmokers.  This reduced
22               responsiveness may wane after smoking cessation.
23
24            •  Although nutritional status (e.g., vitamin E deficiency)  makes laboratory
25               rats more susceptible to O3-induced effects, it is not clear if vitamin E
26               supplementation has an effect in human populations.  Such supplementation
27               has no or minimal effects in animals. The role of such antioxidant vitamins
28               in O3 responsiveness, especially their deficiency, has not been well studied.
29
30
31          Based on information presented in this document, the population groups that have

32     demonstrated responsiveness to ambient concentrations of O3 consist of exercising healthy
33     and asthmatic individuals, including children, adolescents, and adults.
34
        February 1994                            9.42      DRAFT-DO NOT QUOTE OR CITE

-------
 1     REFERENCES

 2     Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during continuous exercise via
 3            the effective dose concept. J. Appl. Physiol: Respir. Environ. Exercise Physiol. 51: 415-422.
 4
 5     Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative respiratory effects
 6            of ozone and ambient oxidant pollution exposure during heavy exercise. J. Air Pollut. Control Assoc.
 7            34: 804-809.
 8
 9     Dungworth, D. L. (1989) Noncarcinogenic responses of the respiratory tract to inhaled toxicants. In: McClellan,
10            R. O.; Henderson, R.  F., eds. Concepts in inhalation toxicology. New York, NY: Hemisphere
11            Publishing; pp. 273-298.
12
13     Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978) The influence of exercise on the
14            pulmonary function changes due to exposure to low concentrations of ozone. In: Folinsbee, L. J.;
15            Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F., eds. Environmental stress:
16            individual human adaptations. New York, NY:  Academic Press; pp. 125-145.
17
18     Hayes, S. R.; Rosenbaum, A.  S.; Wallsten, T. S.; Whitfield, R. G.; Winkler, R. L.; Richmond, H. (1989)
19            A health risk assessment for use in setting the U.S. primary ozone standard. In: Schneider, T.; Lee,
20            S. D.; Welters, G. J. R.; Grant, L. D.;  eds. Atmospheric ozone research and its policy implications:
21            proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
22            Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 851-867. (Studies in environmental
23            science 35).
24
25     Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
26            62: 1671-1680.
27
28     Highfill, J. W.; Costa, D. L. (1994) Statistical response-models for ozone exposure: their generality, when
29            applied to human spirometric and animal permeability functions of the lung. J. Air Waste Manage.
30            Assoc.: submitted.
31
32     Highfill, J. W.; Hatch, G. E.; Slade, R.;  Crissman, K. M.; Norwood, J.; Devlin, R. B.; Costa, D. L. (1992)
33             Concentration-time models for the effects of ozone on bronchoalveolar lavage fluid protein from rats and
34            guinea pigs. Inhalation Toxicol. 4: 1-16.
35
36     Kulle, T. J.; Sauder, L. R.; Hebel,  J. R.; Chatham, M. D. (1985) Ozone response relationships in healthy
37            nonsmokers. Am. Rev. Respir. Dis.  132: 36-41.
38
39     McDonnell, W. F.; Smith, M. V. (1994)  Prediction of acute ozone response as a function of exposure rate and
40            total inhaled dose. J. Appl.  Physiol.: in press.
41
42     McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E.  D.; Salaam, S. A.; House, D. E.
43             (1983) Pulmonary effects of ozone exposure during exercise: dose-response characteristics. J. Appl.
44            Physiol.: Respir. Environ. Exercise Physiol. 54:  1345-1352.
45
46     McDonnell, W. F.; Muller, K. E.;  Bromberg, P. A.; Shy, C.  M. (1993) Predictors of individual differences in
47            acute response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
48
49     McDonnell, W. F.; Andreoni, S.; Smith,  M. V. (1994) Proportion of moderately exercising individuals
50             responding to low-level multihour ozone exposure. Am. Rev. Respir. Dis.: submitted.
51
52
         February 1994                                9.43        DRAFT-DO NOT QUOTE OR CITE

-------
 1     Rombout, P. J. A.; Van Bree, L.; Heisterkamp, S. H.; Marra, M. (1989) The need for an eight hour ozone
 2            standard. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone
 3            research and its policy implications: proceedings of the 3rd US-Dutch international symposium; May
 4            1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers;
 5            pp. 701-710. (Studies in environmental science 35).
 6
 7     Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.; Green, J.;
 8            Raggio, L. (1993) The pulmonary response of white and black adults to six concentrations of ozone.
 9            Am. Rev. Respir. Dis.  147: 804-810.
10
11     Tepper, J. S.; Costa, D. L.; Doerfler, D.; Lehmann,  J.  R.; Stevens, M. A.; Madden, M. C. (1994) The relative
12            contribution of concentration and duration to ozone (O3)-induced lung injury. Fundam. Appl. Toxicol.:
13            submitted.
14
15     U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
16            Research Triangle Park, NC: Office of Health and Environmental  Assessment, Environmental Criteria
17            and Assessment Office; EPA report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS,
18            Springfield, VA; PB87-142949.
19
20     U.S. Environmental Protection Agency. (1989) Review of the national ambient air quality standards for ozone:
21            assessment of scientific and technical information. OAQPS staff paper. Research Triangle Park, NC:
22            Office of Air Quality Planning and Standards; EPA report no. EPA-450/2-92/001. Available from: NTIS,
23            Springfield, VA; PB92-190446.
         February 1994                                9.44       DRAFT-DO NOT QUOTE OR CITE
                                                                    4U.S. GOVERNMENT PRINTING OFFICE:  1994 - 550-064/80010

-------