Tobacco Documents Online

-1- Environmental Tobacco Smoke and Lung Cancer AQproaches to risk assessment P.N.LEE, M.A. (Oxon) Independent Consultant in Statistics and Epidemiology 17 Cedar Road Sutton, Surrey, SM2 5DA England Abstract Based on epidemiological data relating marriage to a smoker to risk of lung cancer, the US Environmental Protection Agency recently concluded that ETS exposure results in approximately 3,000 lung cancer deaths annually among US nonsmokers. Such an estimate is over two orders of magnitude greater than estimates derived by linear extrapolation from data on lung cancer risk in smokers, and relative particulate matter exposure in smokers and nonsmokers. This disparity does not undermine linear extrapolation as a technique, though there must be doubts both about its appropriateness and its accuracy. The disparity reflects more the unscientific nature of the EPA's estimate and their misinterpretation of the epidemiological reports relating lung cancer risk to ETS exposure. When one takes into account the lack of relationship of lung cancer risk in never smokers to workplace or to childhood ETS exposure, the inconsistency of the evidence on histological type of lung cancer, the serious weaknesses in design evident in some studies, and the possibilities of bias due to confounding by other risk factors, . N N O N i A ~ N A O

_2_ misclassi£ication of active smoking status, misdiagnosis of lung cancer, and the failure to publish negative studies, it is clear that ETS exposure has not been shown to cause lung cancer. ~ 0 s N ~ A ~ N A

-3- 1. Introduction Since the publication in 1981 of reports from Japan [1] and Greece [2] of an increased risk of lung cancer in lifelong nonsmokers associated with to marriage to a smoker, there has been increasing concern that exposure to environmental tobacco smoke (ETS) may cause lung cancer. A number of authorities [3-8] have concluded that it does, the most recent, by the US Environmental Protection Agency (EPA) estimating ETS is "responsible for . approximately 3,000 lung cancer deaths annually in U.S. nonsmokers". In this paper I discuss various approaches to the risk assessment of ETS and demonstrate that differing and plausible assumptions can result in such great variation in the estimated number of deaths that EPA's figure of 3,060 (2,000 in never smokers and 1,060 in former smokers) has no valid scientific basis. Indeed I show Zhat there is actually no certainty that anv lung cancer deaths arise as a result of ETS exposure. There are, at least, four basic methods by which one might attempt to carry out risk assessment of ETS. 4igarette ecuivalent avnroach. In this approach, considered in section 2, risk of lung cancer in active smokers is assumed to be adequately quantified by epidemiological studies, and risk of lung cancer in relation to ETS exposure is estimated by extrapolation, based on the equivalent number of cigarettes to which nonsmokers are s exposed.

-4- Pose of carcinogen approach. Cigarette smoke contains more than 3,800 constituents, over 40 of which have been classified by the IARC as showing sufficient evidence of carcinogenicity, in some cases only in animals (5). In theory one might use data on exposure levels to each carcinogen to estimate risk in humans. This approach has never been used for three reasons. Firstly, for many carcinogens, the chemical data relate only to mainstream smoke (MS) and one is not able to quantify levels resulting from typical ETS exposure. Secondly, the observed increased risk of lung cancer in smokers has never been satisfactorily explained by the presence of known amounts of carcinogens in MS, which gives little reason for hope that this approach could adequately quantify risk of lung cancer in relation to ETS exposure. Thirdly, it fails to take into account the possibility of interactions between different chemicals, both synergistic and antagonistic. Animal extrapolation anvroach. Were there good toxicological data demonstrating that exposure of animals to ETS by inhalation resulted in an increased risk of .lung cancer then one could use standard approaches to estimate risk to humans. However, such data do not exist, so this approach cannot be pursued. Enidemioloeical approach. The final approach, considered in section 3, is to apply available epidemiological data relating lung cancer rislc to ETS exposure to a defined population. It is this approach _ that was used by the EPA to estimate their figure of 3,060 lung cancer deaths per year in the US.

-5- 2. Qgarette equivalent approach Sefore attempting.to quantify the extent of risk of lung cancer from ETS by a cigarette equivalent approach, it is important first to consider whether such an approach can actually conclusively demonstrate the existence of anv risk. In section 4 of their report the EPA (8] conclude that ETS can be categorized as a group A (human) carcinogen even in the absence of direct epidemiological data on ETS. They concluded that the epidemiological evidence on active smoking and lung cancer, with no evidence of a threshold level of exposure, the "qualitatively similar" nature of the composition of mainstream smoke and ETS, and the evidence of detectable uptake of tobacco smoke constituents in nonsmokers, taken together, are sufficient for ETS to be classified as group A. In considering this conclusion a number of points should be made: (i) If it were true, it could have been made many years ago. In 1979, for example, there was already extensive evidence available on active cigarette smoking and it was clear that nonsmokers had some exposure to tobacco smoke constituents, even if at a much lower level than smokers. And yet, the US Surgeon-General, in a 41 page chapter on "Involuntary smoking" in an extensive report on Smoking and Health [9], gave no consideration at all to the possibility that ETS might cause lung cancer. (ii) Other reports [e.g. 10] which considered that ETS exposure did cause lung cancer, based their conclusion mainly on the a epidemiological evidence on ETS and lung cancer, and merely used evidence of the type considered by EPA in their section 4

to support their argument. Put the other way round, the evidence in section 4 is normally considered by the authorities as indicating an effect is plausible, not that it definitely exists. (iii) It is true that the epidemiological evidence on active smoking does not demonstrate the existence of a threshold dose. Typically [9], the major studies report an increased risk in the lowest grouping of cigarettes/day smoked, which is certainly statistically significant when the data are considered as a whole. However, the lowest grouping usually has a consumption of 1-5 or 5-10 cigarettes per day, and the evidence neither establishes nor excludes the existence of a threshold at much lower levels of exposure. There is no good evidence that one cigarette a day increases risk, let alone that 0.1, 0.01 or 0.001 cigarettes a day does. (iv) Epidemiologists often state that there is no safe dose of a carcinogen, a view contrary to the views of many toxicologists, brought up on the views of Paracelsus. Others at this conference will distinguish between situations in which thresholds are or are not likely to apply. I will merely observe three points. Firstly, one needs to know the mechanism involved before one can predict whether a threshold is likely to exist or not and in the case of tobacco-associated lung cancer we do not know the mechanism. Secondly, it cannot be . assumed that the presence of known mutagens (genotoxins) in ETS points to there being no threshold in relation to cancer risk. This is clear in relation to formaldehyde, which is

-7- mutagenic, but causes nasal cancer by a non-genotoxic mechanism with a very clear threshold [11]. Thirdly, it was clear that the US Surgeon-General did- not accept the "no-threshold", "one molecule causes cancer" theory, from statements in his 1979 report [9], viz. "The effect of chronic exposure to very low levels of this carcinogen (benzo[a]pyrene) has not been established" and "It is also not established that nitrosamines can act as carcinogens at these levels delivered by inhalation". (v) There are a number of major differences between active smoking and exposure to ETS [12]. In contrast to smokers, ETS exposed nonsmokers breathe in aged tobacco smoke. In vitro tests suggest that aged ETS is less cytotoxic than fresh MS, as inhaled by the smoker. ETS particles are of smaller mean size (0.1-0.2 pg) than MS particles (0.2-0.4 µg), and differences in inhalation patterns between smokers and nonsmokers lead to - a much lower rate of particle deposition in the lungs in the case of ETS exposure (114) as compared to that for smokers (50-90%). Also, the intact clearing mechanism of the respiratory tract of nonsmokers removes particles more effectively than does that of smokers, which may be damaged by smoking. ttf Taking all these points into account, it is clear that one must 0 have strong reservations about the validity of the EPA's argument in i chapter 4. Indeed it is interesting to note that Dr Morton Lippmann, N P the Chairman of the EPA's own Scientific Advisory Board, made it CD

-s- clear at the open meeting in Washington discussing the final draft, that the argument was not a valid one and should not be used. It is surely a matter of serious concern that the EPA chose to ignore the views of its own scientific advisers. It is clear that any attempt at risk estimation using the cigarette equivalent approach must be speculative, and subject to a number of unverifiable assumptions. However, although this is probably true for most, if not all, risk assessments, it is still of interest to see what risk this approach produces. Apart from the problems cited above, there are two particular difficulties in conducting the risk estimation. The first lies in, the form of dose response relationship to assume, even assuming there is no threshold dose. Some of the epidemiological data on active smoking and lung cancer suggests a reasonable fit to a linear relationship between risk and number of cigarettes per day [9]. However others [13] have suggested that inclusion of a quadratic term provides a better fit, rendering linear extrapolation likely to somewhat overestimate risk at lower doses. On the other side of the coin, exposure to ETS may have occurred since birth, whereas the smoking habit is not normally taken up until age 15 or so. Although there is evidence (14] that prolonging the period of exposure to ETS has little effect on the risk, ignoring duration might lead to some und.prestimation of risk. Taking these counterbalancing points in combination suggests that linear extrapolation might not be an inappropriate procedure. Repace [15] has suggested a N ~ a N ~ A 01 N A 4

-9- dose-relationship in which risk at low dose levels is much higher than that predicted by linear extrapolation. However, his model totally failed to fit observed data in the active smoking range and is therefore implausible [16]. The other major problem in risk estimation using the cigarette equivalent approach is to know which tobacco smoke constituent to use when computing the cigarette equivalent for ETS exposure. A number of authors have made it clear that the dose ratio for active smoking to ETS exposure depends dramatically on the constituent considered. Based on results of experimental studies in which healthy male volunteers were exposed to smoking (20 cigs/day) or to ETS exposure (8 hours/day), Scherer and his colleagues [12] estimated ratios of uptake doses for smoking as compared with ETS exposure. For particulate phase components, exposure from smoking was much higher than that from ETS, with ratios estimated as 1250-3000 for particles, 70-150 for benzo[a]pyrene, 110-1500 for cadmium, and 2300-4500 for tobacco-specific nitrosamines. For nicotine, particle-bound in. MS and a gas-phase constituent in ETS, the ratio was estimated to be 75-90. For gaseous phase components exposure was only slightly greater from smoking than from ETS, ratios being estimated as 2.7-4.2 for CO, 4-5 for formaldehyde, 1.5-2.5 for volatile nitrosamines, and 3-5 for benzene. The authors point out that the concentrations of the most frequently used ETS markers that were found in their study were 10 times higher than . those found in everyday environments where real-life exposure to ETS may occur.

-10- The rationale behind the attempts in many countries to encourage a switch from high-tar to low-tar cigarettes is based on the presumption that the tumorigenic effect of tobacco smoke is mainly attributable to the particle phase [5]. It would therefore seem most appropriate to use particulate matter as the constituent for estimation of cigarette equivalents. Arundel and his colleagues [17] carried out a detailed estimation of never smoker lung cancer risks from exposure to particulate tobacco smoke. They calculated that, in the US, current smokers have a daily retained exposure of 310 mg for men and 249 mg for women. In contrast the average never smoker was estimated to have a daily retained exposure of 0.07 mg for men and 0.03 mg for women, equivalent to an average of about 1/200th of a cigarette per day. They further estimated, based on linear extrapolation from lung cancer risks in smokers, that in the US in 1980 there would be a total of jZ lung cancer deaths among never smokers from exposure to particulate ETS: 8 in men and 4 in women. This is much lower than the EPA estimate of 2.000 lung cancer deaths among never smokers from ETS, 500 in men and 1,500 in women. (N.B. Arundel gy gl [17.] did not estimate deaths among former smokers, so comparison with the EPA estimate of 1,060 deaths is not possible.) The above estimates indicate that in males never smokers retain about 0.02% of the amount of particulate tobacco smoke retained by current smokers. For females the figure is 0.01%. A number of researchers have used cotinine, a major metabolite of nicotine, as a

a marker of relative tobacco smoke exposure. In a large study I conducted in 1985 in the UK (18), I found that the corresponding ratios were about 10 times higher, 0.27% in males and 0.13% in females. Had cotinine been used as the index of exposure in Arundel S_t al's calculations, this suggests that about 160 lung cancer deaths would have been predicted, still a full order of magnitude lower than the EPA estimates. It is doubtful, however, that estimates based on cotinine are valid. In the first place, nicotine itself is not deemed to be a carcinogen. Secondly, even if cotinine is used only as an index of smoke uptake, it suffers from the major problem that while cotinine is a marker of the lung's particulate exposure in active smokers, it is a marker of ETS gas phase exposure in nonsmokers. Comparison of cotinine levels in body fluids of smokers and nonsmokers is therefore misleading. N N d .a A O1 UI 0

-12- 3. Enidemioloeical avnroach There are by now 33 published epidemiological studies of lung cancer for which results relating to ETS exposure have been separately presented for lifelong never smokers. I have recently prepared an up-to-date assessment of the data from these studies [19], drawing partly on an earlier book [20) in which I examined 28 of these studies in detail. The 33 studies I considered included all those 30 considered by EPA, with the addition of recently published studies by Brownson [21j and Stockwell [22], and a study by Kabat [23J for which results were only presented at a conference. It is convenient first to describe how the EPA conducted their risk assessment to reach their estimate of 3,060 lung cancer deaths attributable to ETS. The main steps taken, described in full in sections 5 and 6 of their report, can be summarized as follows: (1) Estimate, for never smoking women, the relative risk of lung cancer associated with marriage to a smoker (or in some studies with living with a smoker) in each study. (2) Adjust the relative risk estimates downward to account for bias caused by a proportion of current and former smokers misrepresenting themselves as never smokers, coupled with the tendency for smokers preferentially to marry smokers. (3) Demonstrate, by combining adjusted relative risk estimates from the relevant studies, that there are statistically significant 4 increases in risk in relation to marriage to/living with a smoker in the 11 US studies, in the 5 Japanese studies, in the 4 Hong Kong studies, and in the 2 Greek studies, though not in

-13- the 4 West European studies or in the 4 Chinese studies, and that the overall evidence indicates an association. (4) Demonstrate that there is a stronger association of lung cancer risk with mhrriage to a heavy smoker (or with marriage to a smoker for a long time) than with marriage to an average smoker. (5) Consider various potential confounding factors (history of lung disease, family history of lung disease, heat sources for cooking or heating, cooking with oil, occupation, dietary factors) and conclude that none explains the association between lung cancer and ETS exposure. (6) Classify studies into four "tiers" by a quality assessment, and show that the associations generally remain statistically significant if attention is restricted to studies considered to be of a better quality. (7) Use the information in (1) to (6) to determine that there is a causal relationship, i.e. that a hazard has been identified. (8) Use an estimate of Z - 1.75 for the relative cotinine level of never smokers married to a smoker and never smokers married to a nonsmoker to adjust US relative risk estimates to a non-exposed baseline. Thus, relative to a never smoking woman unexposed to ETS the risk of a never smoking woman married to a nonsmoker is 1.34 and the risk of a never smoking woman married N tlY to a smoker is 1.59. (N.B. the ratio of risks 1.59/1.34 - 1.19 O N ~ s is the relative risk estimate from the 11 US studies, and the A ~ ratio of excess risks 0.59/0.34 - 1.75 is the Z-factor N N N assumed).

-14- (9) Take an estimate of 9.26, from the American Cancer Society Cancer Prevention Study, for the risk of current and former smokers relative to never smokers, and convert it to an estimate (of 13.8) relative to never smokers unexposed to ETS. (10) Use estimates of the total number of lung cancer deaths in US women in 1985, and estimates of the relative frequency of never smokers married to nonsmokers, never smokers married to smokers, and ever smokers, in conjunction with the relative risks of 1.34, 1.59 and 13.8 to calculate that there are 6,970 lung cancer deaths among never smokers, of which 470 are from ETS exposure from the spouse and 1,030 from other, non-spousal, sources of ETS exposure. (11) Assume estimates of the increased lung cancer risk in never smokers in relation to spousal and non-spousal ETS exposure for women apply equally to men, and calculate that there are 80 deaths from ETS exposure from the spouse and 420 from non-spousal ETS exposure. (12) Assume estimates of the increased lung cancer risk in never smokers in relation to ETS exposure for women apply to former smokers of both sexes who have given up five or more years ago, leading to an estimated further 160 female and 150 male deaths from spousal ETS exposure and 270 female and 480 male deaths from non-spousal ETS exposure. (13) Sum the numbers for men (1,130) and women (1,930) for never (2,000) and former (1,060) smokers, or for spousal (860) and non-spousal (2,200) exposure, to give a total of 3,060 deaths due to ETS, rounded to 3,000.

-15- In my recent review of the evidence I concluded that the overall evidence from, the 33 studies indicated a statistically significant relationship between lung cancer risk in never smoking women and marriage to (or living with) a smoker. Of 33 relative risk estimates, unadjusted.for covariates or misclassification of smoking status, 25 were greater than unity (p<0.01) and a "fixed effects" meta-analysis (24] gave an overall relative risk estimate of 1.17 (95% confidence interval (CI) 1.08-1.27). This estimate was only marginally changed (to 1.14, 95% CI 1.05-1_23) if one used, where available, relative risk estimates adjusted for covariates, or if one used "random effects" meta-analysis (to 1.21, 95% CI 1.09-1.36). "Fixed effects" meta-analysis only takes within-study variability into account, but "random effects" meta-analysis also considers between-study variability. An association could also be seen separately in studies in the USA, Europe and Asia, estimates (based on unadjusted relative risks and "fixed effects" meta-analysis) being respectively 1.13 (95% CI 1.00-1.28), 1.40 (95% CI 1.06-1.85), and 1.17 (95• CI 1.03-1.32). 21 studies provided data on risk in relation to extent or duration of smoking by the husband (or cohabitant). Comparing risk in the most heavily exposed group with that in the overall exposed group, the former had a higher risk in 16 studies and a lower risk in only 4, a significant (p<0.05) departure from chance s expectation. Overall the most heavily exposed group had 1.16 times

-16- the risk of the overall exposed group, which suggests that the most heavily exposed group had 1.35 (- 1.16 x 1.17) times the risk of the women not married to a smoker. So far my conclusions were broadly in line with those of the EPA summarized in points (3) and (4) above. However further examination of the data revealed a very large number of flaws in the. EPA's argument which totally overturned their conclusions. These flaws are summarized below: Failure to consider workvlace and childhood ,Qnosure. When evaluating whether an association with ETS exposure exists, it is vital to consider all indices of exposure with adequate data. Although, in 1986, when a number of the major reviews [3-5] were published, there was a worthwhile amount of data only on smoking by the spouse, this is certainly not true now. There are 14 estimates of risk in never smokers in relation to workplace ETS exposure, which, when combined, provide no evidence at all of an association with lung cancer (RR - 1.02,95E CI 0.93-1.12). Similarly there are 14 estimates in relation to childhood ETS exposure from the parents, and again there is no evidence of an association (RR - 0.94, 95% CI 0.84-1.05). There seem to be no strong reasons to believe that smoking by the spouse is a much better marker of ETS exposure than is smoking in the workplace or smoking by the parent in childhood (121. It is therefore grossly biassed to do what the EPA did, namely to conceal from the reader the results for these two ~ N O N a ~ 01 N rn N

-17- alternative indices which show no association at all with lung cancer, and to concentrate solely on the single index, marriage to a smoker, which does show an association_ Eailure to consider histoloeical tvoe of lune cancer. The association of lung cancer with active smoking is much stronger for squamous cell cancer than for adenocarcinoma. If, as EPA assume, ETS is merely a reduced dose of active smoking, one would expect to see effects, if any, for squamous-cell cancer. In fact the evidence regarding histological type of lung cancer is conflicting. There are four studies where the data on spousal smoking seem more consistent with a relationship with squamous (or small cell) carcinoma, four studies where the data seem more consistent with a relationship with adeno (or large cell) carcinoma, one study which found a relationship with both types, and five studies which found no relationship with either type. A major weakness of the EPA report is that it makes no attempt to compare and contrast results for the two major types of lung cancer. Consistency is a criterion that EPA cite for testing causality, but which they do not apply in this context_ Failure to take into account uronerlv the vossibilitv of confoundin¢. There are three fundamental flaws in the EPA's argument. First, they only consider confounding relevant if a single risk factor can be shown to explain the whole association between lung cancer and spousal smoking. This is clearly not sensible. More than one risk factor might confound. Second, when trying to determine whether a factor actually elevates lung cancer risk EPA

-18- restrict attention virtually completely to evidence from the ETS/lung cancer studies themselves, ignoring abundant relevant data from other sources- Third, they ignore the growing evidence [19,25] that ETS exposure is associated with exposure to dietary and other risk factors, and fail to reach the appropriate conclusion that a nonsmoker married to (or living with) a smoker is generally more exposed to other risk factors than is a nonsmoker with no smoker in the household. In a study I conducted recently with my colleagues Alison Thornton and John Fry [25], I identified 33 lifestyle "risk factors", i.e. factors generally perceived to be associated with adverse health consequences, not necessarily with increased lung cancer risk. Of the 33 factors, 27 showed a highly significantly (p<0.001) increased prevalence in smokers of 20+ cigarettes a day compared to never smokers, and only 2 a decreased prevalence. 14 of these factors were also significantly (pG0.01) increased in never smokers with a smoker in the household, and on were significantly decreased. The factors included low fresh fruit and vegetable consumption, high fried food consumption, working in an occupation with a possible cancer risk, low social class, poor eduction, and high alcohol consumption, all factors linked to an increased lung cancer risk. Clearly the question is not whether confounding exists, but what magnitude of bias it causes. Analyses presented elsewhere [19.25] suggest that confounding could explain a material part of the reported association of lung cancer with spousal smoking. It should be noted that the extent to which confounding variables were taken into account in the epidemiological studies of spousal smoking on which the EPA's estimate was based was very limited. Thus over

-19- half the studies where spouse smoking was the index of exposure failed to restrict attention to married subjects, thereby producing a serious confounding between potential effects of ETS exposure and potential effects of marital status (and its correlates). Furthermore, although many studies reported having recorded numerous risk factors, few took any account of these in analysis. Thus, 31 of the 33 studies did not adjust for dietary factors, one of these being a study [22] which actually reported a striking relationship between diet and lung cancer among never smokers in another paper [26]! F_a_ilure to correct fully for bias due to misclassification of active mok n. Adjustment for misclassification is a complex issue involving a number of assumptions that are not easily justified, and variables that are not precisely known. In the EPA report, Wells estimates the bias to be negligible, only increasing the overall relative risk estimate by a factor of about 1.02. However, as discussed in detail elsewhere [27], there are two major reasons why this analysis may understate the effect of bias. One reason lies in the error of applying a misclassification rate, estimated from data virtually all of which comes from North American, European and Australian populations, to results from lung cancer studies conducted in Asia. The fact is that in some countries, such as Japan, smoking by -women is considered socially unacceptable. Consequently misclassification rates are likely to be much higher there. The second reason lies in underestimating the extent of misclassification in the countries for which data are available. N w 0 N ~ ~ O> N UI ~

-20- Elsewhere [19) I present the results of analyses adjusting the meta-analysis relative risk for the US studies for various plausible values of the two key parameters, the misclassification rate and the concordance ratio (the measure of the extent of the association between the smoking habits of husband and wife). Using what I regarded as the most plausible astimates (2.58 misclassification, concordance ratio of 3.0) reduced an unadjusted relative risk of 1.13 (95% CI 1.00-1.28) to 0.96 (95% CI 0.84-1.09). Even assuming only a 1.0% misclassification rate reduces the relative risk estimate to 1_06 (95% CI 0.94-1.20), halving the estimated excess risk and making it become non-significant. Failure to address oublication bias. It is well known that any meta-analysis should consider the possibilities of bias due to failure to publish null studies, but EPA do not consider this issue at all. In fact there is some evidence of this, with a tendency for relative risk estimates to decrease with increasing sample size. Thus, in the 11 studies with less than 50 lung cancer cases, the relative risk estimate was 1.41 (95% CI 1.02-1.93); in the 12 studies with 50 to 100 cases it was 1.33 (95% CI 1.12-1.58); and in the 10 studies with more than 100 cases it was 1.10 (95% CI 1.00-1.21). Failure to test for effects of study weaknesses. lteta-analysis is conventionally used to combine results from similarly designed randomized controlled trials conducted in different populations. It is much more open to question when, as here, it is applied to

-21- non-randomized epidemiological studies of varying design. Before accepting an overall estimate in such circumstances it seems prudent to see whether relative risk estimates vary systematically by different aspects of study design. The EPA did not conduct such analyses but elsewhere {19j I have done so. One clear conclusion was that studies classified as seriously weak on one or more of six criteria (fewer than 10 lung cancer cases, cases and controls from different hospitals, cases and controls interviewed in different places, next-of-kin used to supply data for a much higher proportion of cases than controls, controls and cases unmatched on vital status, no details provided at all on the controls) had much higher relative risk estimates than those with no such weaknesses. Indeed the 16 seriously weak studies included the 12 studies (of 33) with the highest estimates (p<0.001) on a rank test. My analysis also detected one other factor strongly associated with relative risk, namely year of publication of the study. Studies published after 1988 reported no increase in lung cancer risk in relation to marriage to (or living with) a smoker (RR - 1.02, 95% CI 0.91-1.15), while a higher relative risk was reported in studies published in 1981-85 (RR - 1.29, 95% CI 1.09-1.52), or in 1986-88 (RR - 1.42, 95% CI 1.20-1.68). Failure to consider bias due to inaccuracy of diagnosis. While bias due to selection of indices of exposure showing an association at the expense of those that do not, to confounding, to . misclassification of smoking status, to failure to publish null studies, and to poor study design, would tend to result in N N O N ~ ~ N ~ O

-22- overestimation of the risk associated with ETS exposure, bias due to misclassification of diagnosis would be expected to result in some underestimation of risk. However, since there is no significant difference in relative risk between the 16 studies where all or virtually all the diagnoses were histologically confirmed (RR - 1.26, 95% CI 1.10-1.43) and the 17 studies where this was not the case (RR - 1.11, 95% CI 1.00-1.24), this does not seem to be a major factor. Overinteroretation of the dose-resoonse data. The EPA noted that 10 of 14 studies testing for upward trend reported a statistically significant (p<0.05) association and that "this evidence of dose response is very supportive of a causal relationship because it would be an unlikely result of any operative sources of bias or confounding". The EPA's interpretation of this is misleading for a number of reasons. Firstly, their trend analyses included the non-exposed group, many studies being cited as having a significant trend actually showing little or no variation in risk between the exposed groups. Second, studies that report an association between spouse smoking and lung cancer risk are more likely to present dose-response data than those that do not. Third, in some studies authors, faced with a choice of indices of exposure, present detailed results for the index showing the strongest dose-response relationship [28]. Finally, a number of the sources of bias, inFluding misclassification of active smoking status and confounding by diet, would in fact be expected to produce a spurious dose-response relationship [19]. N N N ~ A W N 6f ~

-23- If one takes all the above considerations into account it is clear that the epidemiological evidence does not provide convincing support to the notion that ETS exposure causes lung cancer. There are a number of sources of bias which together could easily explain the observed association between lung cancer and marriage to (or living with) a smoker, which is weak and marginally significant. Clearly under these circumstances it is not sensible to attempt to estimate deaths occurring annually "as a result of" ETS exposure. However, some additional comments should be made on certain aspects of the EPA's risk estimate. Use of dubious Z-factor estimates. For a given relative risk estimate in relation to marriage to a smoker, the proportion of deaths attributed to ETS exposure reduces sharply as the estimated value of the Z-factor (the ratio of ETS exposure in never smokers married to a smoker to that in never smokers married to a nonsmoker) increases. In the 1986 NRC report (4), a Z-factor of 3.0 was used. Given a relative risk estimate of 1.19, this would imply that 24% of lung cancer deaths in never smokers married to a smoker are attributable to ETS exposure, and that 10% of deaths in those married to a nonsmoker are. For the Z-factor of 1.75 used by EPA, these percentages rise to 37% and 25%. It is in fact very doubtful whether 1.75 is an appropriate estimate, for reasons discussed in detail by Layard [29J and Sears [30], who cite data from a number of US studies showing substantially higher Z-factors, exceeding 4.0. a N N O N ~ ~ ~ N QI N

-24- Unjustified extrapolation to males ex-smokers and non-spousal sources of ETS exnosure. Of the 3,060 lung cancer deaths attributed by EPA to ETS exposure, only 470 are attributed to spousal exposure in females. The remaining deaths attributed are in males, in ex-smokers and/or are attributed to non-spousal exposure, all calculated indirectly by extrapolation from the spousal results for females. And yet there are direct epidemiological data available on risk in males, on risk in ex-smokers, and on risk in relation to other forms of ETS exposure, all of which show no statistically significantly increased risk. While the data in males and in ax-smokers are quite limited, the data on workplace exposure is quite substantial, and shows no association with lung cancer risk. It is clearly not scientifically sensible to use data on spousal ETS exposure to estimate effects of non-spousal ETS exposure (of which workplace exposure in clearly a major part) indirectly, ignoring the direct evidence. Undue confidence in the accuracy of the estimated number of lune cancer deaths attributed to ETS. The EPA note [9] that their overall confidence in their estima te of approximately 3,000 lung cancer deaths is "medium to high" . Let us actually consider the facts about this estimate: (i) It depends on an estimate o f relative risk associated with husband's smoking, based on 11 US studies, of 1.19 which has a 95% confidence limits as wide as 1.04 to 1.35. These N U1 confidence limits, which only reflect sampling variation O N ~ a N ~ W

-25- within the studies, would at once imply that the estimated number of deaths from spousal smoking in females of 470 should have confidence limits at least as widely spread as 99 to 866. (ii) The estimate is adjusted for misclassification of smoking status, but the confidence limits do not reflect the considerable uncertainty in the parameters used in the adjustment. As noted above, using alternative, apparently more plausible assumptions, the point estimate of 470 would reduce markedly, perhaps even to zero. Certainly the confidence limits would go below zero. (iii) No adjustment is made for other sources of bias, noted above to be relevant, including confounding, publication bias and weaknesses in study design. (iv) The estimate of 470 is contingent on the Z-factor used. Using a Z-factor of 3.0 instead of 1.75 would approximately halve this estimate. (v) 72% of the total of 3,060 deaths is attributable to non-spousal ETS exposure, this being calculated by extrapolation from the data for spousal ETS exposure ignoring the evidence that workplace ETS exposure is not associated with lung cancer risk. (vi) 35% of the total deaths attributed are in ex-smokers, and 37% in males, with actual data available being ignored, and.the assumption that results for females apply to males and that results for never smokers apply to former smokers being dubious, and adding further uncertainty to the overall N N O N ~ A ~ N ~ 6 estimate of 3,060 deaths.

-26- Clearly in no sense can one have any real confidence in the estimate of 3,060 deaths, let alone the "medium to high" confidence EPA express. _

_27_ 4. Conclusions Assuming that aa linear no-threshold model applies, and that particulate matter deposited in the lung is an appropriate index of e7cposure, it has been calculated [17j, on the basis of the lung cancer risk in smokers, and the relative deposition of tobacco smoke-related particulate matter in the lung in smokers and nonsmokers, that 12 lung cancer deaths occur annually in the US as a result of ETS exposure. This estimate is over two orders of magnitude different from the recent EPA estimate of 3,060 deaths, based on epidemiological reports of an increased risk of lung cancer in never smokers married to smokers. Does this imply that the estimate produced by dose extrapolation is seriously incorrect? Certainly it is difficult to have any very great confidence in this estimate, because of doubts regarding validity of the linear no-threshold model, and of the appropriate index of exposure to use for dose-response extrapolation. However, this does not necessarily mean that the estimate is seriously incorrect. What is clear is that the epidemiologically based estimate of 3,060 deaths has no scientific justification whatsoever. Detailed examination of the evidence reveals that no effect of ETS exposure on lung cancer has been established at all. It is, of course, impossible to prove a negative, but it is clear that very much lower estimates of deaths, 30, 3 or even 0.3 are totally consistent with the data available to date. It is thus not possible to use the epidemiological data on ETS exposure as any sort of gold standard to validate estimates produced by linear extrapolation from the N Ul 0 N ~ ~ N epidemiological data relating to active smoking. M M

-28- 5. References 1. Hirayama T. Nonsmoking wives of heavy smokers have a higher risk of lung cancer: a study from Japan. Br Med J 1981;282:183-5. 2. Trichopoulos D, Kalandidi A, Sparros L, MacMahon B. Lung cancer and passive smoking. Int J Cancer 1981;27:1-4. 3. National Research Council. Environmental tobacco smoke. Measuring exposures and assessing health effects. Washington, National Academy Press, 1986. 4. US Surgeon-General. The health consequences of involuntary smoking; a report of the Surgeon-General. Rockville, US Department of Health and Human Services, Public Health Service, 1986, (CDC)87-8398. 5. International Agency for Research on Cancer. IARC monographs on the evaluation of the carcinogenic risk of chemicals to humans, vol 38: tobacco smoking. Switzerland, IARC, 1986. 6. Australian National Health and Medical Research Council. Effects of passive smoking on health. Australia, 1986. 7. Independent Scientific Committee. Fourth report on smoking and health. London, Her Majesty's Stationery Office, 1988. 8. US Environmental Protection Agency. Respiratory health effects of passive smoking: lung cancer and other disorders. Washington DC, 1992, EPA/600/6-90/006F. 9. US Surgeon-General. Smoking and health; a report of the Surgeon-General. Washington, US Department of Health, Education and Welfare, Public Health Service, 1979; (PHS)79-50066.

-29- 10. Wald N.T, Nanchahal K, Thompson SC, Cuckle HS. Does breathing other people's tobacco smoke cause lung cancer? Br Med J 1986;293:1217-22 11. Casanova M, Heck H d'A. The impact of DNA-protein cross-linking studies on quantitative risk assessments of formaldehyde. CIIT Activities 1991;11:1-6 12. Scherer G, Conze C, Tricker AR, Adlkofer F. Uptake of tobacco smoke constituents on exposure to environmental tobacco smoke (ETS). Clin Investig 1992;70:352-67. . 13. Doll R, Peto R. Cigarette smoking and bronchial carcinoma: dose and time relationships among regular smokers and lifelong non-smokers. J Epidemiol Community Health 1978;32:303-13. 14. Darby SC, Pike MC. Lung cancer and passive smoking: predicted effects from a mathematical model for cigarette smoking and lung cancer. Br J Cancer 1988;58:825-31. 15. Lee PN. An estimate of adult mortality in the United States from passive smoking. Further comment. Environ Int 1992;18:315-7. 16. Repace JL, Lowrey AH. estimating mortality 1991;17:386-7. 17. Arundel A, Sterling from exposure to 1987;13:409-26. T, Observational vs extrapolative models in from passive smoking. Weinkam J. Never smoker lung particulate tobacco 18. Lee PN. Passive Smoking and Lung Cancer. of Bias? Human Toxicol 1967;6:517-24. Environ Int cancer risks smoke. Environ Int 14 Association a Result O N 3 A ~ N CM W

-30- 19. Lee PN. An assessment of the epidemiological evidence relating lung cancer risk in never smokers to environmental tobacco smoke exposure. In: "Environmental Tobacco Smoke". Ed: H Kasuga. Springer-Verlag, New York 1993. 20. Lee PN. Environmental tobacco smoke and mortality. Karger, Basle, 1992. 21. Brownson RC, Alavanja HCR, Hock ET Loy TS. Passive smoking and lung cancer in nonsmoking women. Am J Public Health 1992; 82:1525-30. 22. Stockwell HG, Goldman AL, Lyman CH, Nass CI, Armstrong AW, Pinkham PA, Candelora EC, Brusa MR. Environmental tobacco smoke and lung cancer in nonsmoking women. J Natl Cancer Inst 1992;84:1417-22. 23. Kabat GC. Epidemiologic studies of the relationship between passive smoking and lung cancer. Washington, 1990 Winter Toxicology Forum, 1990:187-99. 24. Fleiss JL, Gross AJ. Meta-analysis in epidemiology, with special reference to studies of the association between exposure to environmental tobacco.smoke and lung cancer: a critique. J Clin Epidemiol 1991;44:127-39. 25. Thornton A, Lee PN, Fry JS.- Differences between smokers, ex-smokers, passive smokers and nonsmokers. Submitted to Journal of Epidemiology and Community Health 1993. 26. Candelora EC, Stockwell HG, Armstrong AW, Pinkham PA. Dietary . intake and risk of lung cancer in women who nev er smoked. Nutr Cancer 1992;17:263-70. N tly O N s A O N 01 O

-31- 27. Lee PN. An estimate of adult mortality in the United States from passive smoking. Environ Int 1993;19:91-100. 28. Lee PN. Lung cancer in nonsmoking women: a multicenter case-control study. Cancer Epidemiology, Biomarkers and Prevention 1992;1:332-3. 29. Layard MW. The background adjustment in risk assessment of environmental tobacco smoke and lung cancer. Environ Int 1992;18:453-61. 30. Sears SB. Presentation before the EPA's Science Advisory Board. Washington DC, July 21-22, 1992. S