WPC" 2 ZBn X`\  PQPBibliogrphy66wBibliographyN=(N*Nxே$E Doc Init66wInitialize Document Style*Nxே$E    I. 1. A. a.(1)(a) i) a)DocumentҲa1DocumentE+N xDocument StyleP ~N=(N!*NF *  ׃  2qea2DocumentE+N xDocument StyleP ~N=(N!*N*    a3DocumentE+N xDocument StyleP ~N=(N!*N0     a4DocumentE+N xDocument StyleP ~N=(N!*N   . a5DocumentE+N xDocument StyleP ~N=(N!*N  2Ke p p ca6DocumentE+N xDocument StyleP ~N=(N!*N  a7DocumentE+N xDocument StyleP ~N=(N!*N ` ` ` a8DocumentE+N xDocument StyleP ~N=(N!*N ` ` ` Tech Init"66wInitialize Technical Style"*Nxே$E  1 .1 .1 .1 .1 .1 .1 .1 Technical2 } -^ a1Technical+N xTechnical Document Style~N=(N#*N 4!     a2Technical+N xTechnical Document Style~N=(N#*N *    a3Technical+N xTechnical Document Style~N=(N#*N'   a4Technical+N xTechnical Document Style~N=(N#*N&   28 $  .  a5Technical+N xTechnical Document Style~N=(N#*N&   . a6Technical+N xTechnical Document Style~N=(N#*N&!"  . a7Technical+N xTechnical Document Style~N=(N#*N&#$  . a8Technical+N xTechnical Document Style~N=(N#*N&%&  . 2e*j &Pleading$66wHeader for numbered pleading paper*Nxே$E'(   ,#`\  PQP# X  y*dddyy*dddy H\1 H\2 H\3 H\4 H\5 H\6 H\7 H\8 H\9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21 H22 H23 H24 H25 H26 H27 H28   Ӳa1Right Par+N xRight-Aligned Paragraph NumbersN**N8)*@   a2Right Par+N xRight-Aligned Paragraph NumbersN**NA+,@` `  ` ` ` a3Right Par+N xRight-Aligned Paragraph NumbersN**NJ-.` ` @  ` `  2Da4Right Par+N xRight-Aligned Paragraph NumbersN**NS/0` `  @  a5Right Par+N xRight-Aligned Paragraph NumbersN**N\12` `  @hh# hhh a6Right Par+N xRight-Aligned Paragraph NumbersN**Ne34` `  hh#@( hh# a7Right Par+N xRight-Aligned Paragraph NumbersN**Nn56` `  hh#(@- ( 2 0Xa8Right Par+N xRight-Aligned Paragraph NumbersN**Nw78` `  hh#(-@pp2 -ppp Ы`\  PQP)\  ` TmsRmn 10pt (F)`\  PQP)\  ` TmsRmn 10pt (F)2 3|`#L |)2#$ 2. INTRODUCTION ă  An estimated 434,000 deaths per year in the United States, or more than one of every six deaths, are attributable to tobacco use, in particular cigarette smoking (CDC, 1991a; figures for 1988). Approximately 112,000 of these smokingrelated deaths are from lung cancer, accounting for an estimated 87% of U.S. lung cancer mortality (U.S. DHHS, 1989). Cigarette smoking is also causally related to cancer at various other sites, such as the bladder, renal pelvis, pancreas, and upper respiratory and digestive tracts (IARC, 1986). Roughly 30,000 deaths per year from cancers at these sites are attributable to smoking (CDC, 1991a). Furthermore, smoking is the major cause of chronic obstructive pulmonary disease (COPD), which includes emphysema, and is thought to be responsible for approximately 61,000 COPD deaths yearly, or about 82% of COPD deaths (U.S. DHHS, 1989). Tobacco use is also a major risk factor for cardiovascular diseases, the leading cause of death in the United States. It is estimated that each year 156,000 heart disease deaths and 26,000 deaths from stroke are attributable to smoking (CDC, 1991a). In addition to this substantial mortality, the association of smoking with these conditions also involves significant morbidity. Smoking also is a risk factor for various respiratory infections, such as influenza, bronchitis, and pneumonia. An estimated 20,000 influenza and pneumonia deaths per year are attributable to smoking (CDC, 1991a). Smokers also suffer from lung function impairment and numerous other respiratory symptoms, such as cough, phlegm production, wheezing, and shortness of breath. In addition, smokers are at increased risk for a variety of other conditions, including pregnancy complications and ulcers. Although the exact mechanisms and tobacco smoke components associated with these health effects are not known with certainty, more than 40 known or suspected human carcinogens have been identified in tobacco smoke. These include, for example, benzene, nickel, polonium-210, 2napthylamine, 4aminobiphenyl, formaldehyde, various Nĩnitrosamines, benz[a]anthracene, and benzo[a]pyrene. Many other toxic agents, such as carbon monoxide, nitrogen oxides, ammonia, and hydrogen cyanide, are also found in tobacco smoke. Smokers, however, are not the only ones at risk from exposure to these tobacco smoke toxicants. In utero exposure from maternal smoking during pregnancy is known to be associated with low birthweight and increased risk of fetal and infant death (U.S. DHHS, 1989). Furthermore, nonsmokers might be at risk for smokingassociated health effects from "passive smoking," or exposure to environmental tobacco smoke (ETS). When a cigarette is smoked, approximately onehalf of the smoke generated is sidestream smoke (SS) emitted from the smoldering cigarette between puffs. This SS contains essentially all of the same carcinogenic and toxic agents that have been identified in the mainstream smoke (MS) inhaled by the smoker (see Chapter 3). SS and exhaled MS are the major components of ETS. Environmental monitoring and measurements of biomarkers for ETS in the biological fluids of nonsmokers demonstrate that ETS constituents can be found at elevated levels in indoor environments where smoking occurs and that these constituents are inhaled and absorbed by nonsmokers (see Chapter 3).Twentysix percent of the U.S. adult population (CDC, 1992b), or about 50 million Americans, are smokers, and so virtually all Americans are likely to be exposed to some amount of ETS in the home, at work, or in public places. Measurements of biomarkers for ETS in nonsmokers confirm that nearly all Americans are exposed to ETS (see Chapter 3). In view of the high levels of mortality and morbidity associated with smoking, the chemical similarity between ETS and MS, and the considerable likelihood for exposure of nonsmokers to ETS, passive smoking is potentially a substantial public health concern. The objectives of this report are to assess the risk to nonsmokers for respiratory health effects from exposure to ETS (hazard identification) and to estimate the population impact (quantitative population risk assessment) of any such ETS-attributable respiratory effects. 2.1. FINDINGS OF PREVIOUS REVIEWS The first epidemiologic results associating passive smoking with lung cancer appeared in the early 1980's. Since then, two major comprehensive reviews of the health effects of passive smoking and several less extensive ones have been published. One of the major reviews was conducted by the National Research Council (NRC) in 1986. At the request of two Federal agencies, the U.S. Environmental Protection Agency and the U.S. Department of Health and Human Services, the NRC formed a committee on passive smoking to evaluate the methods for assessing exposure to ETS and to review the literature on all of the potential health consequences of exposure. The committee's report (NRC, 1986) addresses the issue of lung cancer risk in considerable detail and includes summary analyses from 10 casecontrol studies and 3 cohort (prospective) studies. The report concludes that "considering the evidence as a whole, exposure to ETS increases the incidence of lung cancer in nonsmokers." Combining the data from all the  XX studies, the committee calculated an overall observed relative risk estimate of 1.34 (95% C.I.=  X 1.18, 1.53). The NRC committee was concerned about potential bias in the study results caused by current and former smokers incorrectly selfreported as lifelong nonsmokers (neversmokers). Using plausible assumptions for misreported smoking habits, the committee determined that smoker misclassification cannot account for all of the increased risk observed in the epidemiologic studies. Furthermore, the upward bias on the relative risk of lung cancer caused by smoker misclassification is counterbalanced by the downward bias from background ETS exposure to the supposedly unexposed group. Correcting for smoker misclassification and background ETS exposure, the committee calculated an overall adjusted relative risk estimate of 1.42 (range of 1.24 to 1.61) for lung cancer in nonsmokers from exposure to ETS from spousal smoking plus background sources. The NRC committee also found evidence for noncancer respiratory effects in children exposed to ETS. It recommended that "in view of the weight of the scientific evidence that ETS exposure in children increases the frequency of pulmonary symptoms and respiratory infections, it is prudent to eliminate smoking and resultant ETS from the environments of small children." Furthermore, the committee concluded that "household exposure to ETS is linked with increased rates of chronic ear infections and middle ear effusions in young children." The NRC report also notes that "evidence has accumulated indicating that nonsmoking pregnant women exposed to ETS on a daily basis for several hours are at increased risk for producing lowbirthweight babies, through mechanisms which are, as yet, unknown." The second major review, the Surgeon General's report on the health consequences of passive smoking, also appeared in 1986 (U.S. DHHS, 1986). This review covers ETS chemistry, exposure, and various health effects, primarily lung cancer and childhood respiratory diseases. On the subject of lung cancer, the report concludes: The absence of a threshold for respiratory carcinogenesis in active smoking, the presence of the same carcinogens in mainstream and sidestream smoke, the demonstrated uptake of tobacco smoke constituents by involuntary smokers, and the demonstration of an increased lung cancer risk in some populations with exposures to ETS leads to the conclusion that involuntary smoking is a cause of lung cancer. With respect to respiratory disorders in children, the Surgeon General's report determined that "the children of parents who smoke, compared with the children of nonsmoking parents, have an increased frequency of respiratory infections, increased respiratory symptoms, and slightly smaller rates of increase in lung function as the lung matures." In 1987, a committee of the International Agency for Research on Cancer (IARC) issued a report on methods of analysis and exposure measurement related to passive smoking (IARC, 1987a). The committee reviewed the physicochemical properties of ETS, the toxicological basis for lung cancer, and methods of assessing and monitoring exposure to ETS. The report borrows the summary statement on passive smoking from a previous IARC document that dealt mainly with tobacco smoking (IARC, 1986). The working group that produced the 1986 report had found that the epidemiologic evidence then available on passive smoking was compatible with either the presence or the absence of a lung cancer risk; however, based on other considerations related to biological plausibility, it concluded that passive smoking gives rise to some risk of cancer. Specifically, the 1986 IARC report states: Knowledge of the nature of sidestream and mainstream smoke, of the materials absorbed during "passive smoking," and of the quantitative relationships between dose and effect that are commonly observed from exposure to carcinogens . . . leads to the conclusion that passive smoking gives rise to some risk of lung cancer. More recently, the Working Group on Passive Smoking, an independent international panel of scientists supported in part by RJR Reynolds Nabisco, reported the findings of its comprehensive "bestevidence synthesis" of over 2,900 articles on the health effects of passive smoking (Spitzer et al., 1990). The group concluded that "the weight of evidence is compatible with a positive association between residential exposure to environmental tobacco smoke (primarily from spousal smoking) and the risk of lung cancer." It also found "strong evidence that children exposed in the home to environmental tobacco smoke have higher rates of hospitalization (50% to 100%) for severe respiratory illness" and that the "evidence strongly supports a relationship between exposure to environmental tobacco smoke and asthma among children." In addition, the working group reported that there is evidence for associations between home ETS exposure and many chronic and acute respiratory illnesses, as well as small decreases in physiologic measures of respiratory function, in both children and adults. Evidence demonstrating an increased prevalence of otitis media (inflammation of the middle ear) in children exposed to ETS at home was also noted. With respect to in utero exposure, the group concluded that active maternal smoking is associated with reduced birthweight and with increased infant mortality. A recent review of the health effects associated with adult workplace exposure to ETS conducted by the National Institute for Occupational Safety and Health (NIOSH, 1991) determined that "the collective weight of evidence (i.e., that from the Surgeon General's reports, the similarities in composition of MS and ETS, and the recent epidemiologic studies) is sufficient to conclude that ETS poses an increased risk of lung cancer and possibly heart disease to occupationally exposed workers." Furthermore: Although these data were not gathered in an occupational setting, ETS meets the criteria of the Occupational Safety and Health Administration (OSHA) for classification as a potential occupational carcinogen [Title 29 of the Code of Federal Regulations, Part 1990]. NIOSH therefore recommends that exposures be reduced to the lowest feasible concentration. The classification of "potential occupational carcinogen" is NIOSH's category of strongest evidence for carcinogenicity.  2.2. DEVELOPMENT OF EPA REPORT 2.2.1. Scope   Due to the serious health concerns that have arisen regarding ETS, a virtually ubiquitous indoor air pollutant, and the wealth of new information that has become available since the extensive 1986 reviews, the EPA has performed its own analytical hazard identification and population risk assessment for the respiratory health effects of passive smoking, based on a critical review of the data currently available, with an emphasis on the abundant epidemiologic evidence. The number of lung cancer studies analyzed in this document is more than double the number reviewed in 1986 (31 vs. 13), with a total of about 3,000 lung cancer cases in female nonsmokers now reported in casecontrol studies and almost 300,000 female nonsmokers followed by cohort studies. Furthermore, the database on passive smoking and respiratory disorders in children contains more than 50 new studies, including 9 additional studies on acute lower respiratory tract illnesses, 10on acute and chronic middle ear diseases, 18 on respiratory symptoms, 10 on asthma, and 8 on lung function. This report also discusses six recent studies of the effects of passive smoking on adult respiratory symptoms and lung function. Finally, eight studies of maternal smoking and sudden infant death syndrome (SIDS), which was not addressed in the NRC report or the Surgeon General's report, are reviewed. (Although the cause of SIDS is unknown, the most widely accepted hypotheses suggest that some form of respiratory pathogenesis is usually involved.) First, this report reviews information on the nature of ETS and human exposures. Then, in accordance with the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986a), it critically analyzes human, animal, and genotoxicity data to establish the weight of evidence for the hazard identification of ETS as a human lung carcinogen and to characterize the U.S. population risk. Similarly, it reviews studies of passive smoking and noncancer respiratory disorders, particularly in children, and provides both hazard identification and population risk estimates for some of these effects. While this report restricts analysis to ETSassociated respiratory effects because of time and resource considerations, several recent studies have also linked passive smoking with an increased risk of heart disease or cancers at sites other than the lung. For cancers of other sites, the available evidence is quite limited (e.g., Hirayama, 1984; Sandler et al., 1985), but three recent analyses, examining over 15 epidemiologic studies and various supporting mechanistic studies, suggest that ETS is an important risk factor for heart disease, accounting for as many as 35,000 to 40,000 deaths annually (Wells, 1988; Glantz and Parmley, 1991; Steenland, 1992). This report takes no position on ETS and heart disease. Other health effects of active smoking may also have passive smoking correlates of public health concern. Maternal smoking during pregnancy, for example, is known to affect fetal development. Studies on passive smoking during pregnancy are far fewer but have demonstrated an apparent association with low birthweight (e.g., Martin and Bracken, 1986). Furthermore, passive exposure to tobacco smoke products both in utero (during pregnancy) and postnatally (after birth) may result in other nonrespiratory developmental effects in childrenfor example, decrements in neurological development (Makin et al., 1991). Again, this report takes no position on these potential nonrespiratory effects.  2.2.2. Use of EPA's Guidelines  The lung cancer hazard identification and risk characterization for ETS are conducted in accordance with the EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986a). In fact, tobacco smoke is a mixture of more than 4,000 compounds and could be evaluated according to the Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S. EPA, 1986b). Such a highly complex mixture, however, is not easily characterized with respect to chemical composition, levels of exposure, and toxicity of constituents. Furthermore, the effects and mechanisms of interactions among chemicals are insufficiently understood. The Guidelines for the Health Risk Assessment of Chemical Mixtures acknowledges these inherent uncertainties and recommends various assessment approaches, depending on the nature and quality of the data. When adequate data are available on health effects and exposure for the actual mixture of concern, as is the case with both MS and ETS, the preferred approach, according to the mixtures guidelines, is to adopt the procedures used for single compounds described by the Guidelines for Carcinogen Risk Assessment, as is done here. The EPA also has used this strategy for assessments of diesel exhausts, PCBs, and unleaded gasoline. The compilation of health effects and exposure information for all the mixture components of interest is considered optional. In the case of tobacco smoke, compiling this information would be highly impractical due to the large number of components and the highly complex and changing nature of this mixture. It is also considered unnecessary, given the abundant epidemiologic data on ETS and lung cancer.The Guidelines for Carcinogen Risk Assessment provide a general framework for the analysis of carcinogenic risk, while permitting "sufficient flexibility to accommodate new knowledge and new assessment methods as they emerge" (U.S. EPA, 1986a). According to the guidelines, a qualitative risk assessment, or hazard identification, is performed by evaluating all of the relevant data to determine if a compound has carcinogenic potential. Then, a doseresponse assessment is made by using mathematical models to extrapolate from high experimental or occupational exposures, where risks are usually detected, to lower environmental exposure levels. Finally, the doseresponse assessment and an exposure assessment are integrated into a risk characterization, providing risk estimates for exposed populations. The risk characterization also describes the assumptions and uncertainties in the estimate. The enormous databases on active and passive smoking provide more than sufficient human evidence on which to base a hazard identification of ETS. The use of human evidence eliminates the uncertainties that normally arise when one has to base hazard identification on the results of highdose animal experiments. Furthermore, the epidemiologic data on passive smoking provide direct evidence from environmental exposure levels, obviating the need for a dose- response extrapolation from high to low doses. These lowlevel environmental exposures, however, are associated with low relative risks that can only be detected in welldesigned studies of sufficiently large size. For this reason, new assessment methods are used to categorize studies on the basis of quality criteria and to combine studies to increase the statistical power. Combining studies also provides a means for incorporating both positive and nonpositive study results into the statistical analysis. As an alternative to using actual epidemiologic data on ETS, an ETS risk assessment could have used "cigaretteequivalents" to correlate ETS exposure with lung cancer risk based on dose-response models from active smoking. This would have involved using measures such as cotinine or respirable suspended particles to compare smoke uptake between smokers and ETSexposed nonsmokers in order to equate passive smoking to the active smoking of some quantity of a cigarette(s). Then the carcinogenic response associated with that exposure level would be estimated from extrapolation models based on the doseresponse relationships observed for active smoking. This procedure was not used for several reasons. Although MS and ETS are qualitatively similar with respect to chemical composition (i.e., they contain most, if not all, of the same toxicants and carcinogens), the absolute and proportional quantities of the components, as well as their physical state, can differ substantially. Many tobacco smoke compounds partition preferentially into the MS component of smoke emissions; others, however, such as certain highly carcinogenic Nĩnitrosamines, are preferentially produced at lower temperatures and appear in much greater amounts in the ETS fraction. In addition, active and passive smokers have different breathing patterns, and particles in ETS are smaller than those in MS. Therefore, the distribution and deposition of smoke constituents in the respiratory tracts of active and passive smokers will not be identical. Furthermore, it is not known which of the chemicals in tobacco smoke are responsible for its carcinogenicity. Clearly, the comparison of a small number of biomarker measures cannot adequately quantify differential distributions of unknown carcinogenic compounds.Another area of uncertainty in the "cigaretteequivalents" approach relates to potential metabolic differences between active and passive smokers. Active smoking is known to induce chemical and drugmetabolizing enzymes in various tissues to levels that significantly exceed those found in nonsmokers. Thus, the doseresponse relationships for tobacco smokeassociated health effects are likely to be nonlinear. In fact, evidence suggests that a linear doseresponse extrapolation might underestimate the risk of adverse health effects from low doses of tobacco smoke (Remmer, 1987). Because of these uncertainties, the data from active smoking are more appropriate for qualitative hazard identification than for quantitative doseresponse assessment. Furthermore, at least for lung cancer and other respiratory effects, we have substantial epidemiologic data from actual exposure of nonsmokers to environmental levels of genuine ETS, which constitute a superior database from which to derive quantitative risk estimates for passive smoking, without the need for lowdose extrapolation.  2.2.3. Contents of This Report  ETS is chemically similar to MS, containing most, if not all, of the same toxicants and known or suspected human carcinogens. A major difference, however, is that ETS is rapidly diluted into the environment, and consequently, passive smokers are exposed to much lower concentrations of these agents than are active smokers. Therefore, in assessing potential health risks attributable to ETS, it is important to be able to measure ETS levels in the many environments where it is found and to quantify actual human ETS exposure. The physical and chemical nature of ETS and issues related to human exposure are discussed in Chapter 3. The use of marker compounds and various methods for assessing ambient ETS concentrations, as well as the use of biomarkers and questionnaires to determine human exposure, is described. Furthermore, measurements of ETS components in various indoor environments and of ETS constituents and their metabolites in nonsmokers are presented, providing evidence of actual nonsmoker exposure and uptake. Chapter 4 reviews the major evidence that conclusively establishes that the tobacco smoke inhaled from active smoking is a human lung carcinogen. Unequivocal doseresponse relationships exist between tobacco smoking and lung cancer, with no evidence of a threshold level of exposure. Supporting evidence for the carcinogenicity of tobacco smoke from animal bioassays and genotoxicity experiments is also summarized, including data from the limited animal and mutagenicity studies pertaining specifically to ETS or SS. The chemical similarity between MS and ETS and the measurable uptake of ETS constituents by nonsmokers (Chapter 3), as well as the causal doserelated association between tobacco smoking and lung cancer in humans, extending to the lowest observed doses, and the corroborative evidence for the carcinogenicity of both MS and ETS provided by animal bioassays and genotoxicity studies (Chapter 4), clearly establish the biological plausibility that ETS is also a human lung carcinogen. In fact, this evidence is sufficient in its own right to establish the weight of evidence for ETS as a Group A (known human) carcinogen under EPA guidelines. In addition to the evidence of human carcinogenicity from high exposures to tobacco smoke from active smoking, there are now more than 30 epidemiologic studies investigating lung cancer in nonsmokers exposed to actual ambient levels of ETS. The majority of these studies examine neversmoking women, with spousal smoking used as a surrogate for ETS exposure. Female exposure from spousal smoking is considered to be the single surrogate measure that is the most stable and best represents ETS exposure. Spousal smoking is, however, a crude surrogate, subject to exposure misclassification in both directions, since it actually constitutes only a varying portion of total exposure. For the purposes of the hazard identification analysis in Chapter 5, which is based primarily on the epidemiologic studies of ETS, this document extensively and critically evaluates 31 epidemiologic studies from 8 different countries, including 11 studies from the United States (Appendix A). More than half of these studies have appeared since the NRC and Surgeon General's reviews were issued in 1986. Two U.S. studies are of particular interest. The recently published fivecenter study of Fontham et al. (1991) is a welldesigned and wellconducted case-control study with 429 neversmoking female lung cancer cases and two separate sets of controls. This is the largest casecontrol study to date, and it has a high statistical power to detect the small increases in lung cancer risk that might be expected from ambient exposures. Furthermore, the Fontham et al. study is the only lung cancer study that also measured urinary cotinine levels as a biomarker of exposure. Another large U.S. casecontrol study was the recent study by Janerich et al. (1990), with 191 cases. Both of these studies were supported by the National Cancer Institute. In evaluating epidemiologic studies, potential sources of bias and confounding also must be addressed. Smoker misclassification of current and former smokers as neversmokers is the one identified source of systematic upward bias to the relative risk estimates. Therefore, prior to the analyses of the epidemiologic data that are conducted in Chapters 5 and 6, the relative risk estimates from each study are adjusted for smoker misclassification using the methodology described in Appendix B. Other potential sources of bias and confounding are discussed in Chapter 5. Chapter 5 quantitatively and qualitatively analyzes the epidemiologic data to determine the weight of evidence for the hazard identification of ETS. First, the individual studies are statistically assessed using tests for effect (i.e., association between lung cancer and ETS) and tests for exposureresponse trend. In addition, the highexposure data are analyzed alone to help minimize the effects of exposure misclassification resulting from the use of spousal smoking as a surrogate for ETS exposure. Various combining analyses also are performed to examine and compare the epidemiologic results for separate countries. Then several potential confounders and modifying factors are evaluated to determine if they affect the results. Finally, the studies are analyzed based on qualitative criteria. The studies are categorized into four tiers according to the utility of the study in terms of its likely ability to detect a possible effect, using specific criteria for evaluating the design and conduct as described in Appendix A. These tiers are integrated one at a time into statistical analyses, as an alternative method for evaluating the epidemiologic data that also takes into account qualitative considerations. Chapter 5 concludes with an overall weightofevidence determination for lung cancer based on the analyses in Chapters 3, 4, and 5. In Chapter 6, the population risk for U.S. nonsmokers is characterized by estimating the annual number of lung cancer deaths that are attributable to exposure from all sources of ETS. The overall relative risk estimate from 11 U.S. epidemiological studies of passive smoking and lung cancer in female neversmokers is adjusted upward, based on body cotinine measurements from different U.S. population studies, to correct for the systematic downward bias caused by background exposure to ETS from sources other than spousal smoke. Additional assumptions are used to extend the results from female neversmokers to male neversmokers and longterm former smokers of both sexes. Separate estimates are calculated for background (workplace and other nonhome exposures) and spousal (home) exposures, as well as for female and male never-smokers and former smokers. An alternative analysis of the population risk is performed based solely on the Fontham et al. (1991) study, the only study that provides exposurelevel measurements. Chapter 6 also discusses the sources of uncertainty and sensitivity in the lung cancer estimates. The final two chapters address passive smoking and noncancer respiratory disorders. Both the NRC and Surgeon General's reports concluded that children exposed to ETS from parental smoking are at greater risk for various respiratory illnesses and symptoms. This report confirms and extends those conclusions with analyses of more recent studies. New evidence for an association between ETS and middle ear effusion, and for a role of ETS in the cause as well as the prevalence and severity of childhood asthma, is reviewed. In addition, the evidence for an association between maternal smoking and SIDS is examined. Chapter 7 reviews and analyzes epidemiologic studies of passive smoking and noncancer respiratory disorders, mainly in children. Possible biological mechanisms, additional risk factors and modifiers, and the potential longterm significance of early effects on lung function are discussed. Then, the evidence indicating relationships between childhood exposure to ETS and acute respiratory illnesses, middle ear disease, chronic respiratory symptoms, asthma, and lung function impairment, as well as between maternal smoking and SIDS, is evaluated. Passive smoking as a risk factor for noncancer respiratory health effects in adults is also analyzed in Chapter 7. The NRC and Surgeon General's reports concluded that adults exposed to ETS may exhibit small deficits in lung function but noted that it is difficult to determine the extent to which ETS impairs respiration because so many other factors can similarly affect lung function. More recent evidence and new statistical techniques allow the demonstration of subtle effects of ETS on lung function and respiratory health in adults. Chapter 8 discusses potential confounding factors and possible sources of bias in the ETS studies that might affect the conclusions of Chapter 7. Chapter 8 also describes methodological and data considerations that limit quantitative estimation of noncancer respiratory health effects attributable to ETS exposure. Finally, the chapter develops population impact assessments for ETSattributable childhood asthma and for infant/toddler bronchitis and pneumonia. Acute respiratory illnesses are one of the leading causes of morbidity and mortality during infancy and early childhood, and an estimated 2 to 5 million children under age 18 are afflicted with asthma. Therefore, even small increases in individual risk for these illnesses can result in a substantial public health impact.