Tobacco Documents Online

ASSESSING EXPOSURES TO ENVIRONMENTAL TOBACCO SMOKE AS IT PERTAINS TO: "HEALTH EFFECTS OF PASSIVE SMOKING: ASSESSMENT OF LUNG CANCER IN ADULTS AND RESPIRATORY DISORDERS IN CHILDREN" Comments Prepared by: Howard Goodfellow, Ph.D., P.Eng. Susan Eyre, MHSc. August, 1990 We have been asked by The Tobacco Institute to comment on a draft EPA report entitled "Health Effects of Passive Smoking: Assessment of Lung Cancer in Adults and Respiratory Disorders in Children." Relevant aspects of our backgrounds follow. A copy of a curriculum vitae for each of us is attached. Howard D. Goodfellow is president of Goodfellow Consultants Inc., a Toronto-based consulting firm specializing in environmental engineering and occupational health and safety. He received a Ph.D. from the Department of Chemical Engineering and Applied Chemistry at the University of Toronto and has been responsible for specialized consulting and engineering design services for over 300 industrial and government clients in the environmental field. Dr. Goodfellow is the author of a scientific textbook entitled "Advanced Design of Ventilation Systems for Contaminant Control" and has presented or published over 40 technical papers. He has been

- 2 - active with the Occupational Hygiene Association of Ontario, which he served as President from 1989-1990. Dr. Goodfellow was recently awarded the 2T5 Meritorious Service Medal of the Engineering Alumni Association of the University of Toronto. The award was for his contributions as an engineer,,teacher, researcher, author and administrator in the field of ventilation and occupational health at the University of Toronto, with recognition for achievements in the advancement of environmental consulting. Susan J. Eyre is a Technical Specialist/Occupational Hygienist for Goodfellow Consultants Inc. She graduated with' an MHSc from the Faculty of Medicine at the University of Toronto and has been providing specialized consulting services in the environmental field for several years. She co-authored a paper entitled, "Assessing Exposure to ETS" and has been involved in numerous national indoor air quality surveys. 1.0 INTRODUCTION Since the 1986 Surgeon General's Report (1), environmental tobacco smoke ("ETS") or passive smoking has been linked with adverse health effects in adults and children. The current EPA document (2) uses subsequent epidemiologic evidence to examine the association between ETS and (1) lung cancer in nonsmoking adults and (2) respiratory disease and pulmonary effects in children. It concludes that ETS--particularly during infancy--is associated with increased

- 3 - prevalence of acute lower respiratory tract infections, respiratory symptoms of irritation, and middle ear effusions and also is a cause of lung cancer among nonsmokers. 2.0 BACKGROUND For discussion purposes, it is important first to define mainstream smoke, sidestream smoke, exposure and dose. The definitions are as follows: "The burning of tobacco products leads to the formation of mainstream smoke (MS) and sidestream smoke (SS). MS from cigarettes and cigars is generated during puff drawing in the burning cone and hot - zones; it travels through the tobacco column and exits from the mouth piece. SS is formed in between puff drawing and is emitted freely from the smoldering tobacco product into the ambient air" (3). Exposure is "the concentration at the boundary of the body" whereas dose is the concentration "once the exposure crosses the boundary" (4). Accurate assessment of exposure is fundamental to any credible epidemiological study of ETS health effects. It is well known that measuring exposure to ETS is a complicated task, in part because of the complex nature of ETS itself. ETS is a diverse, dynamic mixture consisting of several thousand constituents. "At present about 4000 single compounds are known in mainstream tobacco smoke. Referring to ~~ the emission of the single cigarette the amounts of these ~ .i1 ,a,

4 compounds range from milligrams down to the scope of picogrammes" (5). Some of these constituents have been used as surrogates for the ETS mixture as a whole or for potentially significant portions of the mixture. The accurate detection of these surrogates is often difficult beqause many are present in extremely low concentrations in air. Further, many substances found in ETS also emanate from other indoor sources, such as building materials and furnishings, cooking and heating fuels, aerosol propellants and cleaning compounds. Thus there are substantial problems in attempting to trace the existence of these substances back to ETS. Finally, the physical and chemical complexity of ETS is increased by its propensity to change over very brief periods of time. Such change can occur as a result of reactions among constituents or between constituents and other surfaces or chemical sources. "Fresh unaged tobacco smoke is an extremely unstable system. Immediately after leaving the coal cone, reactions are starting between the individual compounds" (5). This dynamic quality of changes in physical and chemical composition further complicates the measurement of ETS. Given these difficulties, it is not surprising that there is no universally accepted and standardized method to quantify the exposure of an individual to ETS. A large number of studies have based estimates of exposure on retrospective data obtained from questionnaires, but problems of recall and bias associated with these subjective responses have led

- 5 - investigators to search for a more objective method for measuring exposure. To date, investigators have principally utilized two methods for exposure evaluation. The first is to measure levels of ETS constituents in blood, urine or saliva,(i.e., biological monitoring). The second is to measure levels of ETS constituents in samples of air collected either in the breathing zone or at a fixed sampling location. As will be discussed in the following section, none of these methods are entirely reliable. There are many sources of bias that limit both the qualitative and quantitative estimation of ETS exposure. These limitations make it difficult to associate ETS exposure with lung cancer or respiratory illnesses as the EPA document attempts to do. The scientific legitimacy of the epidemiological evidence used to associate ETS exposure with lung cancer in non-smoking adults and respiratory disease and pulmonary effects in children has been evaluated addressing the following five areas of concern: 1) General methodological considerations. 2) The definition of "exposed" and "unexposed" individuals as well as "background" levels of ETS exposure as the basis for subject cate- gorization and raw data collection. 3) The assessment of ETS exposure with respect to ~ environmental conditions. 4) The applicability of ETS surrogates. ~ CJT 5) The crude classification of ETS exposure in ~p dose-response risk assessment studies. 0

6 3.0 GENERAL METHODOLOGICAL CONSIDERATIONS A 1983 U.S. Department of Health and Human Services Report best discusses the pertinent methodological considerations for conducting epidemiological studies. The report states that "the relatively small differences'in the effects found in various studies . . . may be real and represent true differences among the various communities studied in the measurable effect of involuntary smoke exposure. Such differences may be caused by regional and geographical variations in levels of indoor air pollution that might result from differences in housing (e.g. well insulated- versus poorly insulated houses) and lifestyle (e.g. predominantly indoor living versus predominantly outdoor living). On the other hand, the differences may also be due to methodologic differences in data collection and/or analysis and in the way in which potentially confounding variables have been handled" (6). The report also discusses the difficulty of controlling potentially confounding variables that often bias the results of epidemiological studies. These variables include: "1) Unvented combustion products from different kinds of stoves used for both heating and cooling, e.g., gas, wood and kerosene; 2) other indoor pollutants such as formaldehyde and respirable particulate matter; 3) indoor pollutants of organic origin such as pollens, molds, mites, other allergens and infectious organisms; 4) characteristics

- 7 - of indoor environments such as temperature, humidity and frequency of air-exchanges; 5) social-economic status, culture (ethnic), and such factors as crowding, number of siblings, household conditions, child care, reporting biases, etc; 6) demographic and medical characteristics of the study population such as age, sex, marital status, the presence of underlying respiratory conditions, atopy, infections, disability and/or co-morbidity; 7) parental symptoms such as productive cough which will affect reporting; 8) maternal smoking during pregnancy; 9) annoyance responses and other psychological or social responses to tobacco smoking in a • nonsmoker" (6). Extensive as this list of potentially confounding variables may be, the importance of taking them into consideration in the study design and analysis cannot be overemphasized. "Many of the differences among the many population studies which have looked at the effect of active and passive-smoking on the lung function may be attributed to exposure and/or dose. The logistical difficulties in adequately monitoring these variables are recognized as is the need to develop techniques which are able to measure the biological burden of tobacco smoke. In the future it is likely that considerably less attention will be paid to indirect measures of exposure such as area and personal samplers and more attention paid to biological markers of exposure" (-6). 3.1 Subject Categorization and Background Exposure The EPA document addresses the question of whether any direct evidence exists for the relationship between ETS exposure and lung cancer. Selected epidemiological studies

- 8 - were reviewed and analyzed in an attempt to answer this question. All of "these studies purportedly compare individuals with higher ETS exposures to those with lower exposures. Typically the study subjects are married women who have never-smoked but are married to a smoker (higher exposure) and those married to a nonsmoker (lower exposure). [TJhe higher and lower exposed persons are referred to as "exposed" and "unexposed." (2) Twenty-one case control studies were analyzed and listed in Table 3-1 of the review draft. Tables 3-3 and 3-4 look at the sources of ETS and measures of exposure to ETS ~ respectively. The typical measures of ETS exposure are the number of cigarettes per day, total years and total cigarettes (a combination of the previous two categories). Other exposure categories include hours per day, pack per years and person per years. Each of these measures is subjective in nature. There are considerable problems of recall and bias associated with such responses. Different criteria were also used to categorize individuals as "exposed" or "unexposed" from within a particular exposure category (Table 3-4). For example, in the cigarettes per day category an individual in the INOU experiment would be defined as exposed if someone at home smoked five or more cigarettes per day. In comparative studies (LEE, Table 3-4), an individual exposed to one or more cigarettes per day would be similarly categorized. Such

9 discrepancies in categorization will almost certainly lead to misclassification of subjects. These variations also will cause the levels of "background" exposure to differ between groups. Background exposure is an extremely subjective term that is influenced by many factors. As previously mentioned, substances found in ETS also emanate from indoor materials such as building materials and furnishings, cooking and heating fuels, aerosol propellants and cleaning compounds. These alternate sources are significant contributors to the background levels of certain surrogates and will cause them to vary immensely depending on the environment. Only nine of the twenty-one studies referenced attempted to control background exposures. Background levels of exposure will be different for each individual. Without quantitative measures, it is feasible that a subject categorized as "unexposed" could actually be exposed at a higher level than an "exposed" subject. As the EPA document itself states, the "objective is to provide an assessment of exposure to ETS from all sources that is more inclusive and quantitative than might be available from studies based on spousal smoking" (2,4-19). None of the studies examined for the purposes of the document(2) used the more quantitative methods of exposure assessment. "In future epidemiological studies, some parameters will have to be obtained by measurement or

modelling to validate the questionnaire, especially for the comparison of results of studies performed in different countries" (7). As with the ETS and lung cancer studies, the categorization in the children's respiratory health studies of subjects and their respective background exposures have been subjective. As the report states, "very young children experience prolonged exposure since they spend much of their time at home, and they are not exposed socially or occupationally to ETS or potential confounding agents"(2). But, contrary to the claim made in the report, the mere fact ~ that one parent of a child is a smoker may or may not be reflected in "prolonged exposure" of the child to ETS. The intensity and duration of an individual child's exposure will depend upon a variety of factors that have not been investigated in the studies completed to date. Moreover, the vast majority of factors affecting "background" exposure levels and the categorization of "exposed" and "unexposed" individuals are equally applicable to the respiratory disorder studies. - In summary, the "differences between studies that affect exposure levels and the influence of host-related variables affecting measured responses that cannot be fully controlled, make it difficult to assess the overall evidence from studies on pulmonary function" (2,6-31).