Tobacco Documents Online

Comments To: ETS Comments Indoor Air Division (ANR-445) EPA 401 M St., S.W. Washington, D.C. 20460 Comments In Regard To: Draft EPA Documents Entitled (1) "Health Effects of Passive Smoking: Assessment of Lung Cancer in Adults and Respiratory Disorders in Children" (2) Environmental Tobacco Smoke: A Guide to Workplace Smoking Policies Comments By: Larry C. Holcomb Ph.D. Holcomb Environmental Services 17375 Garfield Road Olivet, MI 49076 USA Telephone 616-763-9442 August, 1990

My Name is Larry C. Holcomb. I hold a Ph.D. in Zoology from Michigan State University. As an independent consultant on matters of environmental toxicology, I am asked to review and comment on a variety of issues. These issues include the health impact of exposure to chemicals associated with toxic waste sites, air and water pollution and occupational exposures. From 1981-86 I served as Executive Secretary of the Michigan Toxic Substance Control Commission. One of the greatest challenges of that position was to determine whether a potential toxic substance problem was a real problem that needed attention. It was my responsibility to determine the scientific validity of evidence that was presented and then make recommendations based on an evaluation of the strength of the data base. I have included a copy of my curriculum vitae for your further reference. I have reviewed the documents at the request of the Tobacco Institute. I am speaking as a scientist representing my own evaluation of the literature on health effects of environmental tobacco smoke (ETS) or second hand smoke. The EPA documents purport to demonstrate that environmental tobacco smoke exposure (ETS) causes disease in nonsmokers. The authors assert that exposure to ETS is responsible for a certain number of lung cancer deaths in nonsmokers per year. They also conclude that ETS is a Class A or known human carcinogen. Furthermore, the authors draw conclusions about the association between a variety of respiratory disorders in children and exposure to ETS. These comments will respond to the issues raised specifically to the EPA document with regard to "Health Effects." However, these comments carry over to the Guide to workplace Smoking Policies. There are many erroneous statements in the "Guide" due to assumptions carried over from the "Health Effects" document.

The preface to- the "Health Effects" document mentions two other issues not addressed in this draft: (1) the possible synergistic lung cancer effect of exposure to ETS and radon and (2) the relative lung cancer hazards from exposure to ETS at home and the workplace. I will not address these two issues here, but should the EPA documents be revised to address these issues, I would appreciate the opportunity to respond. There have been a number of studies performed to determine the impact of ETS on respiratory health in children and adults. The review of any potential health problem from air pollutants requires the use of scientifically acceptable methodology. It is no different for ETS than it would be for any other chemical compound or mixture of compounds. To determine if a health hazard exists, a scientist should make the following determinations: A. What chemical compounds are involved? What is the relative toxicity of those compounds, i.e., non-toxic, moderately toxic, highly toxic? B. What is the concentration of the chemical to which people are exposed? C. What is the route of exposure (dermal, inhalation, ingestion)? D. What is the duration of exposure (24 hours per day, a few hours or a few minutes per day)? E. What is the evidence that an association exists between exposure and health impact? By following these procedures one may ultimately draw conclusions about whether exposure to ETS poses any risk to health; if there is a risk, one may determine what the significance of that risk may be. 3

Sources of Indoor Air Substances There are a multitude of sources of indoor air substances. Some of these are listed below: Cooking Heating Outdoor Air Intake Building Materials Environmental Tobacco Smoke Fabrics Cosmetics Heating and Air Conditioning Ductwork Other possible relevant exposures: Microorganisms such as fungi and bacteria Allergens Herbicides, pesticides Household chemicals Insect and rodent products N02, S02, ozone, formaldehyde Respirable dusts Radon See Figure III for sources received via ingestion, skin and air. Each of the sources, and exposures listed above may present a different type of risk. For example, if a scientist is trying to determine the impact of exposure to formaldehyde contained in ETS, it would be very important to establish if there are other sources of exposure to formaldehyde and what portion of risk is attributable to each source. EVALUATION OF STUDIES ON RESPIRATORY IMPACT ON CHILDREN A number of studies have been performed to investigate the possible impact of ETS on the respiratory health of children. The examination of any potential health problem, no matter how provocative or politically controversial, should involve, of course, the use of standard scientific methodology. The situation in that regard is no different for ETS than it would be for any 4

other chemical compound or mixture of compounds. This presentation will review the available scientific literature on whether exposure to ETS increases the risk of respiratory problems in children. In the past, several approaches to health questions involving ETS have been considered, namely, (1) extrapolation from data obtained in studies of active smoking, (2) animal toxicologic studies and (3) epidemiologic data. As suggested in the 1986 report of the National Research Council of the National Academy of Sciences (NRC/NAS;1986), extrapolation from studies of active smoking is not appropriate since ETS exposure and active smoking (i.e., inhalation of mainstream smoke) and significantly different. ETS is a mixture of sidestream smoke (emitted from the tip of the cigarette) and exhaled mainstream smoke, and it is a mixture that has been aged and markedly diluted in the indoor air. ETS and mainstream smoke differ chemically and physically, as- reflected, for example, in their respective pHs,- particle size and the distribution of their constituents among gaseous and particulate phases. Furthermore, unlike active smoking, ETS exposure is through the nose with shallow breathing. Consistent with these fundamental differences between active smoking and ETS exposure, levels of specific tobacco smoke markers (such as nicotine or its metabolite, cotinine) in body fluids of ETS-exposed individuals are usually very small (1$ or less) relative to those of active smokers. Additionally, no toxicologic study in animals has adequately addressed the possible health effects of ETS, since the available studies have involved extremely high doses, usually of mainstream smoke. Therefore, the only relevant information that is available on the possible effects of ETS on the respiratory system in children comes from epidemiologic studies. - In epidemiologic studies, information about family health and smoking history has been obtained from supervised or unsupervised 5 m ~ C~T CA N N F+

questionnaires, usually completed by parents or guardians. The end-points of these studies have been (1) the incidence of respiratory symptoms and disease and/or (2) pulmonary performance, as measured by respiratory flow rates. The incidence of respiratory symptoms or illness and pulmonary function data have been compared statistically in children classified as to ETS exposure by their parents' smoking status, with the statistical significance of an association having been regarded as suggesting a link between parental smoking and an adverse respiratory health effect in children. In all of the relevant epidemiologic reports it has been assumed that parental smoking (or the smoking of other household smokers) is an adequate surrogate for ETS exposure of the children being examined. Despite the dubious nature of that assumption, in the vast majority of the studies has ETS exposure has not been verified by measurement of a specific environmental or biological marker, such as airborne nicotine or body fluid cotinine. The epidemiologic literature pertaining to the possible effects of ETS on children was recently examined at an international symposium on ETS issues at McGill University in Montreal (Witorsch,1990). That examination pointed out that age differences were apparent in the literature on the relationship between parental (or household) smoking and respiratory effects in children. A number_ of epidemiologic -studies have reported an association between parental, usually maternal, smoking and increased risk of respiratory problems (such as cough, sneezing, asthma, bronchitis and pneumonia) in infants and children under five years of age. A few of these studies also have reported that the effects varied dose-dependently with the amount of ETS_exposure (e.g., number of cigarettes reportedly smoked by the parents and/or number of smokers). The literature on the reported statistical association between parental smoking and respiratory effects is not consistent in 6 m ~ Vt ~Jt N N N

children five years and older. An examination of 28 studies at the McGill symposium revealed eight giving negative results. In those reporting a statistical association, the relative risk was usually below 2.0 and the results were highly variable. When a statistical association between ETS exposure and a particular health effect was reported (such as with asthma, coughing, sneezing and bronchitis), there was considerable variation from one study to the next. A particular symptom or illness was confirmed usually no more than about 50 percent of the time. As with respiratory symptoms and illness data, the reported statistical association between parental smoking and pulmonary function in older children is also inconsistent. (I might note here that such studies are difficult to undertake in young children). For example, a decrement was observed for respiratory flow rates (e.g., the FEV1 or the volume of air forcibly expired in 1 second) in only 12 of 23 studies. Furthermore, the decrements observed were usually small in magnitude (e.g., <1t to 7% for FEV,) and pulmonary function was still within the normal range. Of five studies that have attempted to associate parental smoking with lung growth in children (measured as a change in FEV, over time) only two have reported a small decrement (1$ or less per year). When studies have been published on the same cohort of children at different times, the data from one study have not been confirmed by the data from the subsequent study. The notion of age-dependency suggested in the available epidemiologic literature was noted in the 1986 Reports of both the U.S. Surgeon General and the National Research council of the National Academy of Sciences. Several studies have reported that as the child ages (e.g., reaches 6 months to 2 years) the association between parental smoking and respiratory effects diminishes or even disappears (Colley gt aj~,, 1974; Fergusson g,t Al_,, 1981; Fergusson and Horwood, 1985; Chen gt Al,, 1988). 7

The apparent age-dependent association could reflect a declining sensitivity to ETS as the child ages and/or a change in the child's relationship with the mother &/or a variety of other factors. As noted above, parental smoking has served as the surrogate for ETS exposure in the relevant studies without verification with a specific marker. While some studies (not involving health effects) report a correlation in body fluid cotinine levels in children with the number of smokers in the family and other parameters of household smoking, the issue of whether parental smoking necessarily implies an effect of ETS, or whether other factors also correlated with parental smoking could be involved, remains to be determined. It is important to understand that no epidemiologic study can establish a cause and effect relationship. Instead, results are expressed as a statistical association or computation of relative risk between an exposed population and a population that has not been exposed. Numerous variables or confounding factors, many difficult to control, can influence the outcome of an epidemiologic study. The apparent association between parental smoking and respiratory health in young children could reflect various confounding factors most notably those related to socioeconomic status. Parental smoking has been shown to be more prevalent in low income families and positively correlated with factors that may impair respiratory health, such as outdoor air pollution, cross-infection,_gas stove usage, more family members per living space, and frequent change of address (Kerigan gt at., 1986). The data of Harlap and Davies (1974) illustrate how socioeconomic status or related factors could confound an effect attributable to parental smoking. These researchers reported that, while parental smoking was associated with increased respiratory illness in infants, it also was associated with increased hospitalizations of infants due to injury and poisonings. 8

Adjustment for socioeconomic status in epidemiologic studies involves consideration of maternal education, income, occupation and ethnicity, among other factors (Green, 1970). Most relevant studies to date have either ignored socioeconomic factors or have underestimated their impact and complexity. Finally, even the most stringent of socioeconomic adjustments may not adequately correct for such important factors as family attitudes and practices concerning fitness, stress management, and prenatal and childhood care and nutrition. I note, finally, that Rubin and Damus (1988) recently evaluated 30 studies dealing with ETS exposure and respiratory health and function in children. These papers were quantitatively rated on the basis of seven important epidemiologic criteria (such as data collection, estimates of smoke exposure, definition of illness). The authors concluded that most of the studies reviewed had significant design flaws that compromised reliance on their conclusions. ETS AND HEALTH EFFECTS ON ADULTS Some Strategies Used to Suggest ETS-related Health Effects in Adults Are: A. Extrapolation based on active smoking data. B. Chemical composition of ETS. C. Clinical and epidemiological studies examining disease endpoints as a function of ETS exposure. I will attempt in this discussion to examine some of these strategies and problems in utilizing them. Extrapolation from active smoking data to ETS data are meaningless when trying to calculate health-related effects. The two are completely different phenomena as follows: ~ ~ CA N N t11 9

Nonsmoker exposed to ETS Taken in via nose Dilution of smoke Aging affects on smoke Chemical alteration Not held in lung Direct Smoker Taken in via mouth More concentrated smoke No aging affects Little chemical alteration Smoke held in the lung Methods of Estimating ETS Exposure Several methods have been used to estimate ETS exposure as follows: A. Spousal smoking habits (Questionnaire/Interview responses). B. Hours of-ETS exposure (Questionnaire/Interview responses). C. Biological Markers Estimation of hours of ETS exposure has been attempted in very few studies and with little success. Therefore, there is nothing of value to report on this. I will discuss biological markers here and elaborate further on spousal smoking later on in the discussion. Use of Biological Markers Several attempts have been made by investigators to identify a chemical found in ETS (or one of its metabolites) that would serve as a biological marker. A biological marker that would allow estimation of exposure of nonsmokers to ETS or to separate smokers from nonsmokers in epidemiological studies, would be useful. Most biological marker studies to date have not been of value. At present, the data are not available to extrapolate from ETS to direct smoking. The methodology for sampling has been inadequate in most studies. Nicotine and cotinine are not related to carcinogenic materials. Nicotine or cotinine levels may be used to validate whether a person is a smoker or nonsmoker but they are of no value to judge long-term effects. Thus, use of biological markers is not a valid bases for measuring risk due to exposure to ETS. 10