NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

IARC Working Group on the Evaluation of Carcinogenic Risk to Humans. Tobacco Smoke and Involuntary Smoking. Lyon (FR): International Agency for Research on Cancer; 2004. (IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, No. 83.)

Cover of Tobacco Smoke and Involuntary Smoking

Tobacco Smoke and Involuntary Smoking.

Show details

2Studies of Cancer in Humans

2.1. Lung cancer

The section summarizes the results of the relevant cohort studies and case–control studies of the association between lung cancer and exposure to secondhand smoke. They are ordered by source of exposure, i.e. secondhand smoke from partners at home, at the workplace, during childhood and from other sources. For each type of study, the results are presented first without differentiation according to the levels of exposure and then the exposure–response relationship is described.

The most commonly used measure of exposure to secondhand smoke has been from the spouse. This is because it is well defined and has been validated using cotinine studies of never-smokers who do or do not live with smokers. Spousal exposure is also a marker of exposure to tobacco smoke in general because people who live with smokers tend to mix with smokers outside the home. Other measures of exposure, at the workplace or during childhood, are not so well validated. It is more difficult to quantify exposure at the workplace than spousal exposure; the extent of exposure may vary considerably between different working environments (exposure from the spouse is clearly defined and fairly consistent); people are more likely to change jobs than to remarry or divorce and, in studies based on people who have died from lung cancer, it may be more difficult for the next of kin or other respondent to know whether or not the subject had been exposed to secondhand smoke at work. Exposure during childhood has not been validated and, in studies of exposure to secondhand smoke, the relative risk for lung cancer associated with exposure during childhood should be stratified according to spousal exposure. Few studies have done this, and, even when they have, the number of lung cancer cases has been too small to enable robust conclusions to be drawn.

2.1.1. Cohort studies

There have been eight cohort studies of nonsmokers who were followed for several years to determine the risk for lung cancer (these are described in Table 2.1). Six of these studies (Garfinkel, 1981; Hirayama, 1984; Butler, 1988; Cardenas et al., 1997; Jee et al., 1999; Nishino et al., 2001) reported the risk of lung cancer associated with exposure to secondhand smoke from the spouse. All six studies found that the risk for nonsmoking women with partners who smoked was higher than that for those whose partner did not smoke (see Table 2.2). In both cohort studies that reported on the effect in nonsmoking men whose wives smoked, the relative risk was increased (Hirayama 1984; Cardenas et al., 1997). The two other cohort studies, which were based on general exposure to secondhand smoke (deWaard et al., 1995; Speizer et al., 1999), obtained similar results.

Table 2.1. Cohort studies of secondhand smoke and lung cancer.

Table 2.1

Cohort studies of secondhand smoke and lung cancer.

Table 2.2. Epidemiological studies of the risk for lung cancer in lifelong nonsmokers whose spouses smoked relative to the risk in those whose spouses did not smoke.

Table 2.2

Epidemiological studies of the risk for lung cancer in lifelong nonsmokers whose spouses smoked relative to the risk in those whose spouses did not smoke.

Exposure–response relationships

The analysis of exposure–response relationships provides critical evidence for or against a causal relationship between exposure to secondhand smoke and the development of lung cancer.

In the study by Garfinkel (1981), the relative risk did not increase with increasing exposure levels.

In the study by Hirayama (1984), the relative risks for women were 1.4, 1.4, 1.6 and 1.9 when their husbands were ex-smokers, and when they smoked 1–14, 15–19 or ≥20 cigarettes/day, respectively (p value for trend test, 0.002). Similarly the relative risk for nonsmoking men increased with exposure level: it was 2.1 when the wives smoked 1–19 cigarettes/day and 2.3 when they smoked ≥20 cigarettes/day (p value for trend test, 0.02).

The study by Cardenas et al. (1997) also found a significant exposure–response relationship. When the husbands smoked 1–19, 20–39 and ≥40 cigarettes/day, the relative risks for women exposed to secondhand smoke were 1.1, 1.2 and 1.9 respectively (p value for trend test, 0.03). There was no evidence of an association between risk and the length of time the couples had been married. A similar analysis for nonsmoking men exposed to secondhand smoke would not be robust because the number of cases was too small. The particular strengths of this study were the near complete data on cause of death (97%), direct questioning of both partners about their smoking habits, and the taking into account of numerous potential confounders such as previous lung disease, occupational exposure to asbestos, dietary habits and education.

Taken together, the three large cohort studies demonstrate an increased incidence of deaths from lung cancer associated with exposure to secondhand smoke from the spouse. The increase in deaths from lung cancer in the study by Hirayama (1984) is significant, and this study and that by Cardenas et al. (1997) also reported a significant exposure–response relationship.

2.1.2. Case–control studies

Many case–control studies have been undertaken in several countries (mostly China and the USA) (described in Table 2.3). In these studies lung cancer cases were ascertained and matched with controls (usually for age and other factors). The controls were selected from either the general population or the hospital in which the patients with lung cancer were diagnosed. Details of the smoking habits of the partners of both cases and controls were obtained either by interview or questionnaire. In some instances the next of kin provided the relevant information. These studies were based on various measures of exposure to secondhand smoke, including exposure from the partner, at the workplace, during childhood or exposure from other sources. The following section describes only large, relevant studies in which all cases and controls were interviewed in person and no information from the next of kin was used to reconstruct exposure.

Table 2.3. Main study design characteristics of case–control studies on exposure to secondhand smoke and lung cancer.

Table 2.3

Main study design characteristics of case–control studies on exposure to secondhand smoke and lung cancer.

(a) Description of studies

The study of Lam et al. (1987) included 199 cases and 335 controls from Hong Kong, Special Administrative Region of China. The study is characterized by good data on exposure (from various sources) to secondhand smoke, valid classification of the smoking habits of the husband and a consideration of potential confounders (including education, place of birth, duration of residence and marital status).

The study from China by Gao et al. (1987) included 246 cases and 375 controls. The cases were identified using a system built upon the Shanghai Cancer Registry. The analyses were controlled for age and education.

The study of Wu-Williams et al. (1990), also from China, included 417 cases and 602 controls. A limitation of this study is that it was not able to control for important indoor sources of exposure to potential lung carcinogens such as those produced during burning of coal and frying in oil.

The study of Brownson et al. (1992) in the USA included 431 cases and 1166 controls. It is characterized by good documentation of data on exposure in the home and the workplace, and took into account potential confounders (age, sex and socioeconomic status).

In the study of Fontham et al. (1994) in the USA, 651 cases and 1253 controls were interviewed. Possible misclassification of smokers and potential confounding by age, occupational exposure to known carcinogens, eating habits, familial history of lung cancer and education were taken into account. The smoking status was verified by means of cotinine determination to minimize the misclassification of smokers as nonsmokers.

The study from China by Sun et al. (1996) included 230 cases and 230 controls. The study controlled for age and education, but not for burning of coal and frying in oil.

The study of Lee et al. (2000) from China (Province of Taiwan) included 268 cases and 445 controls and was an extension of the study of Ko et al. (1997). Detailed information on exposure to secondhand smoke was collected, and nonsmoking status was verified by household members. Potential confounding by age, education, occupation, cooking fuels and other factors was allowed for.

The participants in a European multicentre study included 650 cases and 1542 controls from 12 centres in seven countries. Potential confounders such as occupational exposure, socioeconomic status and intake of fruits and vegetables were taken into account. The main publication was by Boffetta et al. (1998), but additional analyses were made of effects of secondhand smoke from cigars, cigarillos and pipes (Boffetta et al., 1999b) and of exposure to secondhand smoke and diet (Brennan et al., 2000). In addition, the data from some centres on specific aspects have been published separately and in some cases with additional data (Germany: Jöckel et al., 1998a,b; Kreuzer et al., 2000, 2001, 2002; Sweden: Nyberg et al., 1998).

The study of Zaridze et al. (1998) from Russia included 189 cases and 358 controls. Information on exposure to secondhand smoke in the family and from colleagues at work was obtained, and potential confounders (age and education) were considered.

(b) Exposure to secondhand smoke from the partner

Table 2.2 shows the relative risk for lung cancer associated with exposure to secondhand smoke from the spouse. Taking the crude relative risks, or the adjusted estimates when the crude ones are not available (in any event, the crude and adjusted estimates are similar) 25 of the 40 case–control studies of nonsmoking women showed an increased risk; the results of seven of the 25 studies were statistically significant (Trichopolous et al., 1983; Lam, 1985; Lam et al., 1987; Geng et al., 1988; Fontham et al., 1994; Zaridze et al., 1998; Lee et al., 2000). In studies of nonsmoking men, five of the nine studies showed an increased risk, although none were statistically significant.

Exposure–response relationships

Several studies reported the risk of lung cancer associated with increasing levels of exposure, in particular, the number of cigarettes smoked by the spouse per day, the number of years of living with a smoker and pack–years; these studies are listed in Table 2.4. Because most of these studies were relatively small, they would not have had sufficient statistical power to find an exposure–response relationship. Eight studies found a statistically significant trend (p value < 0.05) between lung cancer risk and the number of cigarettes smoked by the spouse (Trichopolous et al., 1983; Hirayama, 1984; Garfinkel et al., 1985; Lam et al., 1987; Geng et al., 1988; Inoue & Hirayama, 1988; Liu et al., 1993; Cardenas et al., 1997) and one other found an almost statistically significant trend (Akiba et al., 1986; p = 0.06). Six studies found a statistically significant trend (p value < 0.05) for lung cancer risk and the number of years of marriage to a smoker (Gao et al., 1987; Geng et al., 1988; Stockwell et al., 1992; Fontham et al., 1994; Cardenas et al., 1997; Jee et al., 1999) and the results of two others were almost significant (Kalandidi et al., 1990; Zaridze et al., 1998; p value = 0.07 in both).

Table 2.4. Relative risk of lung cancer in lifelong nonsmoking women comparing those with the highest exposure to secondhand smoke from a smoking partner to women with nonsmoking partners (the relative risks are ranked in ascending order for each type of exposure).

Table 2.4

Relative risk of lung cancer in lifelong nonsmoking women comparing those with the highest exposure to secondhand smoke from a smoking partner to women with nonsmoking partners (the relative risks are ranked in ascending order for each type of exposure). (more...)

Table 2.4 shows the increase in risk in nonsmoking women who have the highest level of exposure according to each measure. All 20 of the studies that reported on the number of cigarettes smoked showed an increased risk in the highest exposure group, and seven of the studies reported a doubling of risk or more. Similarly, of the 18 studies that looked at the number of years of marriage to a smoker, all but three showed an increased risk in the highest exposure group; six reported a relative risk of at least 2.0.

In summary, there is evidence of an exposure–response relationship, thus providing further support for a causal relationship between the development of lung cancer and exposure to secondhand smoke from partners.

(c) Exposure to secondhand smoke at the workplace

In total, 23 studies have been published on exposure to secondhand smoke at the workplace (Table 2.5). The results from these studies are mixed with some showing a positive association and others not. Only one study reported a statistically significant association between exposure to secondhand smoke at the workplace and risk for lung cancer (Reynolds et al., 1996). Many of the studies assessed only recent workplace exposure to secondhand smoke; this is likely to result in a serious misclassification of exposure because past exposure is more likely to be etiologically relevant.

Table 2.5. The relative risk for lung cancer in nonsmokers exposed to secondhand smoke at the workplace compared with nonsmokers who were not.

Table 2.5

The relative risk for lung cancer in nonsmokers exposed to secondhand smoke at the workplace compared with nonsmokers who were not.

Exposure–response relationships

Two studies found no statistically significant exposure–response relationship (Kalandidi et al., 1990; Kabat et al., 1995).

In the study by Reynolds et al. (1996) in the USA, the risk for lung cancer in women who were exposed to secondhand smoke at work was significantly increased to 1.6 (95% CI, 1.2–2.0). For women who had been exposed to secondhand smoke for 1–15, 16–30 or > 30 years, the relative risk for developing lung cancer increased significantly (p < 0.001) with the length of the exposure period: 1.5 (95% CI, 1.1–1.9), 1.6 (95% CI, 1.1–2.2) and 2.1 (95% CI, 1.4–3.2), respectively.

In the European multicentre study (Boffetta et al. 1998), the relative risk for lung cancer after exposure to secondhand smoke at work was 1.2 (95% CI, 0.9–1.5). No exposure–response relationship was seen when the data were analysed according to duration of exposure but a significant trend was observed after analysis of weighted exposure, which is most likely a better index of exposure than duration. A significant relative risk of 2.1 (95% CI, 1.3–3.2) was observed in the group with the highest weighted exposure.

Two studies from Germany which are included in part in Boffetta et al. (1998) also showed an increased risk in the highest exposure group of 1.9 (95% CI, 1.1–3.6) and 2.5 (women only, 95% CI, 1.1–5.7) (Jöckel et al., 1998b; Kreuzer et al., 2000).

The study of Rapiti et al. (1999) reported increasing relative risks with increasing duration of exposure, but the trend did not reach statistical significance.

In the study of Zhong et al. (1999), women ever exposed to secondhand smoke at work showed an odds ratio of 1.7 (95% CI, 1.3–2.3). There was a statistically significant (p < 0.001) increase in risk associated with the number of hours of exposure per day at work with odds ratios of 1.0 (95% CI, 0.6–1.7), 1.6 (95% CI, 1.0–2.5) and 2.9 (95% CI, 1.8–4.7) for 1–2, 3–4 and > 4 h per day. When the number of co-workers who smoked was considered there was again a statistically significant trend (p < 0.001) with odds ratios of 1.0 (95% CI, 0.6–1.6), 1.7 (95% CI, 1.1–2.8) and 3.0 (95% CI, 1.8–4.9) for 1–2, 3–4 and > 4 co-workers who smoked, whereas there was no increase in relative risk with increasing numbers of years of exposure to secondhand smoke. Risk estimates were not affected when analyses were restricted to personal interviews excluding proxy interviews.

In summary, the studies in which exposure–response relationships were analysed generally revealed an increase in the relative risk for lung cancer associated with exposure to secondhand smoke at work and statistically significant increases in relative risk in those groups with the highest level of exposure. The associations are stronger in studies with better assessment of exposure and other aspects of study design.

(d) Exposure during childhood

The studies on exposure to secondhand smoke during childhood are summarized in Table 2.6. The results of these studies have been somewhat contradictory. Out of 23 studies, only three studies of exposure from the mother reported a significantly increased relative risk (Brownson et al., 1992; Sun et al., 1996; Rapiti et al., 1999) and two studies reported a significant increase in relative risk related to exposure from the father or either parent (Sun et al., 1996; Rapiti et al., 1999). One study found a significant inverse association with exposure from the father or either parent (Boffetta et al., 1998).

Table 2.6. The relative risk for lung cancer in nonsmokers exposed to secondhand smoke during childhood compared with that in nonsmokers who were not.

Table 2.6

The relative risk for lung cancer in nonsmokers exposed to secondhand smoke during childhood compared with that in nonsmokers who were not.

Exposure–response relationships

The study of Wang et al. (2000) observed a significant trend (p < 0.01) with increasing pack–years of childhood exposure to secondhand smoke with odds ratios for men and women combined of 1.0, 1.4 (95% CI, 1.0–2.1), 1.8 (95% CI, 1.0–3.3) and 3.0 (95% CI, 1.0–8.9) for < 1, 1–9, 10–19 and ≥ 20 pack–years. In contrast, the study of Boffetta et al. (1998) suggested a negative trend for cumulative exposure, which was statitistically significant for all subjects combined (p = 0.02).

In summary, there is no clear indication that lung cancer risk in later life is associated with exposure to secondhand smoke in childhood. However, an important problem in interpreting these studies is the very poor quality of the assessment of exposure that occurred 50 or more years in the past.

(e) Exposure from other sources

Few studies have addressed exposure to secondhand smoke from other sources. Kreuzer et al. (2000) reported a significantly increased relative risk of 2.6 (95% CI, 1.3–5.4) for exposure in vehicles in the highest category of weighted duration of exposure.

Other studies have either not addressed these other sources of exposure or have considered them only as part of a cumulative exposure from all sources.

In summary, insufficient data are available to evaluate the risk from exposure to secondhand smoke from other sources.

(f) Bias and confounding

There are two sources of bias (misclassification bias and bias resulting from exposure to secondhand smoke in the reference group) and several potential confounders (e.g. dietary confounding) that can result in the relative risk being overestimated or underestimated in the studies of the association between lung cancer and exposure to secondhand smoke described above.

(i) Misclassification bias

Misclassification bias occurs when some of the subjects recorded as never-smokers who are included in the studies are in fact current or former smokers who have misreported their smoking status. Their true smoking status makes these subjects more likely to develop lung cancer and because smokers tend to live with smokers, this bias will over-estimate the true risk for lung cancer from exposure to secondhand smoke from the spouse. There has been much discussion in the literature on this bias, and it is the main factor proposed as partly or fully explaining the increased risk for lung cancer observed in epidemiological studies. The bias has four determinants:

  • The prevalence of smoking in a particular population. This can be obtained directly from some of the studies or from national statistics.
  • The aggregation ratio (the extent to which a smoker is more likely to live with another smoker rather than a nonsmoker). It is generally accepted to be between 2 and 4 (Wald et al., 1986; US Environmental Protection Agency, 1992; Lee, 1992; Hackshaw et al., 1997).
  • The relative risk for lung cancer in current and former smokers misclassified as never-smokers. Some meta-analyses have assumed that the risk for lung cancer in misclassified smokers is the same as that in all reported smokers (US Environmental Protection Agency, 1992; Lee, 1992, 1998). However, misclassified current smokers tend to be light smokers and misclassified former smokers have usually given up smoking many years before the study, so the risk in both groups will be less than the average risk in all current or former smokers. The overall relative risk for lung cancer in misclassified ever smokers has been estimated to be about 3 (Hackshaw et al., 1997).
  • The percentage of current and former smokers misclassified as never-smokers. The percentage of misclassified current smokers can be estimated by comparing self-reported smoking status with serum, urine or salivary cotinine levels; current smokers who report themselves to be never-smokers would tend to have high concentrations (for example, a urinary cotinine concentration > 50 ng/mg creatinine). Wells et al. (1998) combined the results of 13 studies, seven of which were used in the US Environmental Protection Agency (1992) report, and concluded that the rates of misclassification of smokers are low; 1.6% of Caucasian women who were current smokers reported themselves as never-smokers. The estimate was higher, though still low, for women from a minority background (4.9%). Similar conclusions had been drawn from a review of six studies on cotinine and nicotine (two of which were included in the review by Wells et al., 1998) in which it was estimated that 3.1% of ever smokers were current smokers who reported themselves as never-smokers (Hackshaw et al., 1997). Two of the case–control studies on secondhand smoke and risk of lung cancer in female never-smokers (Table 2.2) measured urinary cotinine in the subjects and compared this with their reported smoking status. The percentage of reported never-smoking women with urinary cotinine concentrations > 50 ng/mg creatinine was 3.5% in the study by Riboli et al. (1995) (included in Boffetta et al., 1998 in Table 2.2) and 3.1% of patients with lung cancer and 5.0% of controls in the study by Fontham et al. (1994).

(ii) Bias resulting from exposure to secondhand smoke in the reference group

Studies of the risk for lung cancer and exposure to secondhand smoke have defined the reference group as never-smoking women with husbands who are nonsmokers. However, these women, although not exposed at home, may be exposed to secondhand smoke outside the home. This bias will tend to underestimate the true relative risk.

(iii) Dietary confounding

Several potential confounders have been proposed that may partly or fully explain the increased risk of lung cancer associated with exposure to secondhand smoke from the spouse. None of these potential confounders have been established as having a causal link with lung cancer. For example, dietary confounding (perhaps the main potential confounder) may arise because (i) nonsmokers who live with smokers tend to have similar diets, (ii) the diets of smokers tend to be poorer than those of nonsmokers (i.e. lower consumption of fruits and vegetables) and (iii) people who consume less fruits and vegetables may be more likely to develop lung cancer. Several of the observational studies listed in Table 2.2 had attempted to adjust for consumption of fruits and vegetables or other dietary factors (Dalager et al., 1986 [used data from Correa et al. (1983) and Buffler et al. (1984) in Table 2.2]; Hirayama, 1989 [used data from Hirayama (1984)]; Kalandidi et al., 1990; Alavanja et al., 1993 [used data from Brownson et al. (1992)]; Fontham et al., 1994; Mayne et al., 1994 [used data from Janerich et al. (1990)]; Cardenas et al., 1997; Boffetta et al., 1998; Zhong et al., 1999; Brennan et al., 2000; Johnson et al., 2001); they showed that the effect of dietary confounding was negligible.

2.1.3. Meta-analyses of observational studies of exposure to secondhand smoke and lung cancer in adults

(a) Introduction

Since the publication of the first epidemiological studies that reported directly on the association between exposure to secondhand smoke and the risk of lung cancer in nonsmokers (Garfinkel, 1981; Hirayama, 1981), there have been several other cohort studies and case–control studies. Most of these studies were based on a relatively small number of lung cancer cases and did not, therefore, have enough power to show a statistically significant association on their own. Meta-analyses have therefore been performed with the aim of pooling the available data and thus providing a more precise estimate of the risk. A meta-analysis is a formal statistical technique used to combine the estimates of relative risk across studies into a single estimate. Originally developed for clinical trials, it has also been applied to observational studies (see Peto, 1992, for a brief discussion of some aspects of meta-analyses of case–control and cohort studies on cancer). In spite of some concerns over the application of meta-analysis to studies of secondhand smoke and lung cancer, it is an appropriate approach for interpreting the published data collectively.

(b) Published meta-analyses

This section presents the results of selected published reports.

(i) Exposure to secondhand smoke from the spouse

Table 2.7 shows the main results of published meta-analyses on the risk for lung cancer in never-smokers associated with exposure to secondhand smoke from the spouse, including an indication of whether any adjustment was made for bias and confounding. All the pooled estimates show an increased risk (relative risks of 1.1–1.6), despite using different combinations of studies and methodology.

Table 2.7. Summary results of selected published meta-analyses of the risk for lung cancer in never-smokers exposed to secondhand smoke from the spouse.

Table 2.7

Summary results of selected published meta-analyses of the risk for lung cancer in never-smokers exposed to secondhand smoke from the spouse.

Some meta-analyses adjusted for the misclassification of ever-smokers as never-smokers (which will tend to overestimate risk). For example, in the analysis by Hackshaw et al. (1997) the relative risk was reduced from 1.24 to 1.18 after allowing for misclassification bias in 37 studies of nonsmoking women. In the analysis by Lee et al. (2001), which was based on 47 studies and used a different methodology, after allowing for misclassification bias the relative risk was reduced from 1.23 to 1.17. The effect is small.

Few meta-analyses have adjusted for background exposure to secondhand smoke from sources other than the spouse in the reference group (which will tend to under-estimate risk). Hackshaw et al. (1997) reported that the effect of such an adjustment was to increase the observed relative risk from 1.24 to 1.42.

Few reviews have attempted to adjust for diet as a potential confounder. Hackshaw et al. (1997) used pooled data from nine studies of the risk of lung cancer associated with fruit and vegetable consumption in nonsmokers and pooled data from three studies on the difference in diet between nonsmokers who did and did not live with a smoker; the relative risk for lung cancer due to exposure to secondhand smoke from the spouse was reduced from 1.24 (as observed) to 1.21 after adjusting for fruit and vegetable consumption. A similarly small effect was reported by Lee et al. (2001), after adjusting for consumption of dietary fat and education as well as consumption of fruits and vegetables, and using different methodology and a larger set of studies (for the risk of lung cancer associated with each confounder: 17 studies on consumption of fruits and vegetables, seven on dietary fat and 12 on education; for the difference between nonsmokers who do and do not live with a smoker: nine studies on consumption of fruits and vegetables, seven on dietary fat and nine on education). The relative risk for lung cancer when the husband smoked 10 cigarettes/day was reduced from 1.10 (observed) to 1.09, after allowing for these three confounders (Lee et al., 2001). In both analyses the effect of allowing for confounding was small.

Generally, the overestimation due to misclassification bias and potential confounding seems to be balanced by the underestimation due to exposure to secondhand smoke in the reference group (Hackshaw et al., 1997).

(ii) Exposure at the workplace

Interest in the risk of lung cancer associated with exposure to secondhand smoke at work has increased over the years and several meta-analyses have been published. These are listed in Table 2.8; some report no association, for example, Lee (1992) and Levois and Layard (1994), whereas others do report an association (Biggerstaff et al., 1994; Wells, 1998; Zhong et al., 2000). However, the results of some of the studies may be unreliable because they used levels of exposure reported by next of kin (who may not know the true exposure status of the case or control), and because some studies evaluated only recent exposure to secondhand smoke in the workplace. Wells et al. (1998) excluded studies that documented only recent exposure and also studies that (i) included more than 50% surrogate responses for cases, (ii) had only minimal exposure, (iii) included exposure to other respiratory carcinogens, (iv) included subjects who had smoked, and (v) did not report appropriate data to allow the confidence intervals to be checked. Based on these criteria, Wells et al. (1998) identified the following studies for inclusion in their meta-analysis: Wu et al. (1985), Shimizu et al. (1988), Kalandidi et al. (1990), Kabat et al. (1995) and Reynolds et al. (1996); the pooled risk estimate was 1.4 (1.2–1.7). Overall, there seems to be an increased risk of lung cancer in subjects exposed to secondhand smoke at the workplace.

Table 2.8. Summary of results from published meta-analyses of exposure to secondhand smoke and lung cancer in never-smokers exposed at the workplace.

Table 2.8

Summary of results from published meta-analyses of exposure to secondhand smoke and lung cancer in never-smokers exposed at the workplace.

(iii) Exposure during childhood

There have been few meta-analyses on the risk of lung cancer in adulthood following exposure to secondhand smoke during childhood; the results of three of these meta-analyses are given in Table 2.9. None suggested an association, although no stratification according to gender or exposure from the mother or father was carried out. Overall, published meta-analyses have found no evidence for an increased risk for lung cancer associated with childhood exposure to secondhand smoke.

Table 2.9. Results from published meta-analyses of exposure to secondhand smoke and lung cancer in adult never-smokers exposed during childhood.

Table 2.9

Results from published meta-analyses of exposure to secondhand smoke and lung cancer in adult never-smokers exposed during childhood.

(iv) Statistical methods and other considerations
Pooling relative risks

Different methods of combining relative risk estimates from individual studies have generally tended to give similar results. For example, in 37 studies of the risk for lung cancer of never-smoking women exposed or unexposed to secondhand smoke from the spouse, the relative risks (95% CI) using the fixed or random effects model were 1.21 (1.12–1.30) and 1.24 (1.13–1.36), respectively (Hackshaw et al., 1997) (the random effects model allows for heterogeneity between the risk estimates).

More complex approaches, such as Bayesian analysis, also do not yield materially different results. The difference between the pooled estimates obtained using a Bayesian model and those obtained using a simpler random effects model was small. Tweedie et al. (1996) pooled 40 studies of male or female never-smokers exposed to secondhand smoke from the spouse, the pooled relative risk for lung cancer was 1.20 (95% CI, 1.07–1.34) using the random effects model and 1.22 (95% CI, 1.08–1.37) using a Bayesian model (Tweedie et al., 1996).

Pooling results relating to exposure–response relationships

Several studies on the effects of exposure to secondhand smoke in never-smokers have reported the relative risk for lung cancer according to the number of cigarettes smoked by the spouse or the number of years that the nonsmoker has lived with a spouse who smokes. A few researchers, using various combinations of studies and methodology, have attempted to pool the results of epidemiological studies of exposure–response in never-smoking women. For an increase of 10 cigarettes per day smoked by the husband, the excess relative risk for lung cancer compared with never-smoking husbands was estimated to be 23% (95% CI, 14–32) by Hackshaw et al. (1997), 17% (95% CI, 12–22) by Brown (1999) and 10% (95% CI, 5–15) by Lee et al. (2000). The excess relative risk that resulted from living for 10 years with a husband who smokes compared with one who does not was estimated to be 11% (95% CI, 4–17) by Hackshaw et al. (1997) and 7% (95% CI, 4–11) by Lee et al. (2000). The estimates are reasonably consistent between different reports and all found a statistically significant increase in risk associated with increasing exposure.

Heterogeneity between the estimates of relative risk

Performing a meta-analysis when there are statistically significant differences between the estimates of relative risk may yield an incorrect pooled estimate. If heterogeneity exists, an attempt should be made to explain it. If it can be explained by a single factor (or factors), then estimates should be stratified according to that factor. The authors of several reviews of the association between exposure to secondhand smoke and lung cancer have allowed for the existence of heterogeneity between geographical regions or found evidence of it and therefore stratified the relative risk estimates according to region (for example, US Environmental Protection Agency, 1992; Lee, 1998). Lee (1998) assessed heterogeneity related to several factors including geographical region, study publication date, study type and study size and concluded that there were statistically significant differences between the relative risk estimates by almost all factors. However, this was shown to be due to a single large discrepant study that unduly influenced the assessment of heterogeneity; this may be a problem especially when there are relatively few studies in the meta-analysis. In the meta-analysis by Hackshaw et al. (1997), the test for heterogeneity based on 37 studies on nonsmoking women was almost significant (p = 0.10), although when one study was excluded the p value became 0.46. The discrepant study, from China, was large (417 cases of lung cancer) and reported an almost statistically significant reduction in the risk of lung cancer associated with exposure to secondhand smoke from the spouse (relative risk, 0.8; 95% CI, 0.6–1.0), making its results inconsistent with those of the other studies. When this study was excluded, no evidence of heterogeneity was found for several factors (Hackshaw et al., 1997; Hackshaw, 1998; Zhong et al., 2000).

Publication bias

In meta-analyses of studies of the relationship between secondhand smoke and lung cancer there is a possibility of publication bias if studies with positive results (those that show an increased risk of lung cancer) are more likely to be published than studies with negative ones (those that show a decreased risk or no difference in risk). The pooled estimate of risk would then be biased upwards. Simple methods to ascertain whether much publication bias exists suggest that there is little evidence of this, for example funnel plots (Lubin, 1999) or estimating the number of negative unpublished studies that would be required to explain the increased risk observed from epidemiological studies — about 300 (Hackshaw et al., 1997; Lee, 1998); it is implausible that there would be so many unpublished negative studies. Copas and Shi (2000) used a complex method to adjust the observed relative risk for lung cancer (reported in Hackshaw et al., 1997) for publication bias; the pooled estimate was reduced from 1.24 to 1.15, but Copas and Shi assumed that 40% of all studies are unpublished. Even with such an extreme assumption, the adjusted estimate is consistent with the reported relative risk adjusted for bias and confounding (1.26; 95% CI, 1.06–1.47). The problem with assessing publication bias is that it is difficult to determine empirically how many studies are unpublished (Bero et al., 1994).

(c) Updated meta-analyses

Several individual studies on secondhand smoke and the risk of lung cancer in nonsmokers have been published since one of the last detailed meta-analyses on the subject (Hackshaw et al., 1997). This section presents updated meta-analyses using currently available results. The selection of studies to be included is as described by Hackshaw et al. (1997), and the method of pooling the relative risk estimates is that described by Dersimonian and Laird (1986), which allows for any heterogeneity between the estimates. Some case–control studies reported only crude estimates of relative risk, some reported only adjusted estimates (adjusted for various factors such as age and diet) and others reported both crude and adjusted estimates. Consideration therefore needed to be given to which should be used in the meta-analyses. Pooled estimates were obtained based on the crude relative risks and, where these were not available, the adjusted relative risks. This reduces the effect of those studies that adjusted for factors that are not established confounders. The pooled estimate was also obtained based on the adjusted relative risks, and where these were not available, the crude relative risks to show that the two approaches yielded similar results.

Table 2.10 shows the results of the updated meta-analyses according to type of exposure to secondhand smoke and gender of the subject (for the estimates from the individual studies, see Tables 2.2, 2.5 and 2.6).

Table 2.10. Summary of the updated meta-analyses of the relative risk for lung cancer in never-smokers exposed to specified sources of secondhand smoke.

Table 2.10

Summary of the updated meta-analyses of the relative risk for lung cancer in never-smokers exposed to specified sources of secondhand smoke.

(i) Exposure from the spouse

Among nonsmoking women who lived with a spouse who smoked, the risk of lung cancer was increased by 24% (relative risk, 1.24; 95% CI, 1.14–1.3; Table 2.10). This estimate was based on the crude estimates of relative risk found in the studies and, where these were not available, the adjusted estimates. Use of the adjusted estimates and, where these were not available, the crude estimates yielded a similar relative risk of 1.27 (95% CI, 1.15–1.41). The studies came from several countries, and the test for heterogeneity between the relative risk estimates across all 46 studies just misses statistical significance (p value = 0.08). However, if the discrepant study from China by Wu-Williams et al. (1990) that reported an almost statistically significant decrease in risk due to exposure to secondhand smoke is excluded, the pooled relative risk is not materially changed (1.25; 95% CI, 1.17–1.33), but the test for heterogeneity yields a p value of 0.34. Among nonsmoking men who lived with a smoker, the risk of lung cancer was increased by 37%. The risk estimates for both nonsmoking men and women are statistically significant.

(ii) Exposure at the workplace

The increase in risk for lung cancer in nonsmoking women is about 20% (relative risk, 1.19; 95% CI, 1.09–1.30; Table 2.10). If the pooled estimate was based on the adjusted relative risks reported in the studies and, where these were not available, the crude estimates, the result was similar (relative risk, 1.21; 95% CI, 1.09–1.35). There was also an increase in risk in men (12%) though this result was not statistically significant (probably because of the smaller number of studies and fewer cases of lung cancer in the meta-analysis). There was no evidence of heterogeneity between the individual risk estimates.

(iii) Exposure during childhood

There is a statistically significant increase in risk among women exposed to secondhand smoke from the mother during childhood (50% increase in risk, but the confidence interval is wide, 4–114%). There is a lower, and non-significant increase in risk for exposure to secondhand smoke from the father (25%). However, there is significant heterogeneity between the estimates of relative risk. The results on exposure during childhood are less clear than those on exposure from the spouse or at the workplace.

Overall, the evidence from the meta-analyses is clear; adult nonsmokers exposed to secondhand smoke have a higher risk for lung cancer. Although the precise quantitative estimate of risk may vary between different measures of exposure, it is consistently raised. The data on exposure to secondhand smoke from the spouse also show that risk increases with increasing exposure. The evidence for an association between lung cancer and childhood exposure to secondhand smoke is less consistent than that for exposure in adulthood.

2.2. Breast cancer

Five prospective cohort studies (Hirayama, 1984; Jee et al., 1999; Wartenberg et al., 2000; Nishino et al., 2001; Egan et al., 2002) and 12 reports of 10 case–control studies (Sandler et al., 1985a,b; Smith et al., 1994; Morabia et al., 1996; Millikan et al., 1998; Lash & Aschengrau, 1999; Delfino et al., 2000; Johnson et al., 2000; Marcus et al., 2000; Morabia et al., 2000; Chang-Claude et al., 2002; Kropp & Chang-Claude, 2002) have examined the role of secondhand smoke in breast cancer. The cohort studies are summarized in Table 2.11 and the reports from the case–control studies are summarized in Table 2.12.

Table 2.11. Cohort studies of breast cancer and involuntary exposure to tobacco smoke.

Table 2.11

Cohort studies of breast cancer and involuntary exposure to tobacco smoke.

Table 2.12. Case–control studies of breast cancer and involuntary exposure to tobacco smoke.

Table 2.12

Case–control studies of breast cancer and involuntary exposure to tobacco smoke.

2.2.1. Cohort studies

The first cohort study that suggested a possible association of exposure to secondhand smoke with breast cancer was reported by Hirayama in 1984. Specific details of how risk estimates for breast cancer were calculated were provided by Wells (1991). A total of 115 deaths from breast cancer were identified after 15 years of follow-up (1966–81) of over 91 000 married nonsmoking Japanese women. Women whose husbands had ever smoked had a small non-significantly increased risk of breast cancer (relative risk, 1.26; 95% CI, 0.8–2.0). [The Working Group noted this was a first prospective report with a number of limitations. For example, it reported mortality rather than incidence; there was limited assessment of risk specific to breast cancer (spouse only); there was no adjustment for potential confounders; exposure was assessed at only one time-point.]

In a study in the Republic of Korea, Jee et al. (1999) also found small non-significantly increased risks of breast cancer associated with husbands' smoking status: for former smokers the relative risk was 1.2 (95% CI, 0.8–1.8) and for current smokers the relative risk was 1.3 (95% CI, 0.9–1.8). Relative risks were adjusted for age of husbands and wives, socioeconomic status, residence, vegetable consumption and occupation of the husband. These findings were based on 138 incident and prevalent breast cancer cases in 3.5 years of follow-up (July 1994–December 1997) of a cohort of 157 436 nonsmoking Korean women. A higher risk, of borderline significance, was observed for women married to current smokers who had smoked for more than 30 years (relative risk, 1.7; 95% CI, 1.0–2.8). [The Working Group noted that this study had several limitations, i.e. prevalent cases were not excluded; limited adjustment was made for potential confounders, and the adjustment did not include reproductive or hormonal factors; assessment of exposure included only secondhand smoke from the spouse.]

Wartenberg et al. (2000) reported findings from the large American Cancer Society Cancer Prevention Study II cohort based on 12 years of follow-up (1982–94) of never-smoking women who had been married once. A total of 669 deaths from breast cancer were included and risk estimates were adjusted for year of age at baseline, race, number of years of education, history of breast cancer in mother or sister, personal history of breast cysts, age at first live birth, age at menopause, number of spontaneous abortions, use of oral contraceptives, use of estrogen replacement therapy, body mass index, alcohol intake, fat consumption, vegetable consumption, occupation and occupation of spouse. No increased risks were found for women married to current smokers (relative risk, 1.0; 95% CI, 0.8–1.2) or former smokers (relative risk, 1.0; 95% CI, 0.8–1.2) when compared with never-smokers married to nonsmoking husbands. No association was found by type of tobacco. No trend in risk was observed by years, packs per day or pack–years of spousal smoking. No significant associations were noted between breast cancer and all exposures at home (relative risk, 1.1; 95% CI, 0.9–1.3), at work (relative risk, 0.8; 95% CI; 0.6–1.0), or in other places (relative risk, 0.9; 95% CI, 0.7–1.2). When reported exposures from all sources were combined and examined according to daily hours of exposure using no exposure from any source as referent (0 hour), no trend was observed. [The Working Group considered that the strengths of this study included the large number of cases, the excellent follow-up, the thorough statistical adjustment for potential confounders and that the spouse directly reported his own tobacco use. The limitations include the use of mortality rather than incidence as the outcome and that the assessment of spousal smoking was made at only one time-point.]

A smaller cohort study that included 9675 Japanese female never-smokers over the age of 40 years accrued 67 incident cases over a 9-year follow-up period (1984–92) (Nishino et al., 2001). Relative risks were adjusted for age, study area, alcohol consumption, intake of green and yellow vegetables, intake of fruit, age at first birth, number of live births, age at menarche and body mass index. The age-adjusted relative risk for breast cancer was 0.6 (95% CI, 0.3–1.0) among women whose husbands smoked when compared to that in women married to nonsmokers. The age-adjusted risk associated with living in a household with other smokers was also below unity (relative risk, 0.4; 95% CI, 0.2–0.8) when compared with women living in households where there were no smokers. Further adjustment of these relative risks for the potential confounders listed above did not appreciably change the risk estimates, but the relative risks were no longer statistically significant after full adjustment: exposure from spouse, 0.6 (95% CI, 0.3–1.1), and other household members, 0.8 (95% CI, 0.4–1.5). [The Working Group considered that the strengths of this study include adjustment for some reproductive or hormonal and dietary factors; its limitations include the very small sample size, lack of information on marital status at baseline and inclusion of unmarried women at high risk of breast cancer as unexposed which may have reduced point estimates of relative risk.]

The Nurses' Health Study in the USA has provided the largest number of prospectively accrued breast cancer cases in never-smoking women (Egan et al., 2002). After 14 years of follow-up (1982–96), 1359 cases of invasive breast cancer were diagnosed among 35 193 never-smokers. Exposure to secondhand smoke was assessed as exposure during childhood as well as during adult life at home, at work and in other settings. Relative risks were adjusted for many variables including age, parity, age at first birth, menopausal status, age at menopause, change in weight (i.e. weight at age 18 years compared to the most recent reported weight), age at menarche, history of benign breast disease, family history of breast cancer, post-menopausal hormone treatment, alcohol intake and carotenoid intake. No statistically significant associations were found for exposure between breast cancer and exposure to secondhand smoke in adult life or in childhood, and most relative risks were near unity. No trends were apparent either for number of years lived with a smoker as an adult (p for trend = 0.87) or for a categorized index of adult exposures (p for trend = 0.97). Women who reported the highest levels of exposure to secondhand smoke during adulthood had a rate of breast cancer similar to that of women who reported no current exposure to secondhand smoke (relative risk, 1.0; 95% CI, 0.8–1.3). The findings were similar for pre- and postmenopausal women. [The Working Group noted that this study's main strength is that it is the largest and most methodologically rigorous prospective study to date. Other strengths were that exposure assessments were updated over time, incident cases rather than mortality were studied and comprehensive adjustment was made for potential confounders.]

2.2.2. Case–control studies

The first two reports (Sandler et al., 1985a,b) on involuntary smoking and breast cancer were based on a case–control study conducted in North Carolina, USA. Cases were selected from a single hospital tumour registry and included patients diagnosed between 1 July 1979 and 31 March 1981, who were between the ages of 15 and 59 years at the time of diagnosis. Approximately 60% of the controls were friends or acquaintances identified by cases and the remaining 40% were selected by systematic telephone sampling. The two control groups were combined after separate analyses of the two groups indicated similar results. The risk for breast cancer in nonsmoking women was not associated with exposure to secondhand smoke during childhood from either mother (relative risk, 0.9) or father (relative risk, 0.9) (Sandler et al., 1985a). Exposure to secondhand smoke in nonsmoking women based on husband's smoking was associated with a two-fold, nonsignificant increase in risk (relative risk, 2.0; 95% CI, 0.9–4.3) (Sandler et al., 1985b). Risk estimates of childhood exposure were adjusted for age and education, and risk estimates of exposure during adulthood were adjusted for age, race and education. Both reports included only a few lifetime nonsmokers with breast cancer (29 and 32 cases, respectively). [The Working Group noted that the limitations of this study include small sample size, lack of adjustment for reproductive factors and the potentially inappropriate control group (i.e. friends and neighbours of cases supplemented with controls selected by random digit dialling.]

Smith et al. (1994) investigated the relationship between exposure to secondhand smoke and risk for breast cancer in a sample of nonsmokers including 94 incident cases and 99 controls drawn from a larger study of breast cancer diagnosed in young women below the age of 36 years between 1982 and 1985. This study was conducted in the United Kingdom and information on exposure to secondhand smoke was collected by postal questionnaire in a sample of participants from the main study. Controls were selected randomly from the list of the case's general practitioner and matched to the case on age. Risk estimates were adjusted for age, residence, age at menarche, family history of breast cancer, history of biopsy for benign breast disease, oral contraceptive use and history of breastfeeding. Although most of the risk estimates exceeded unity as shown in Table 2.12, none were statistically significant and there was no evidence of a positive trend in risk associated with increasing exposure. When total lifetime exposure as measured in cigarette–years was considered, elevations in risk were found for all levels above zero (referent). However, the trend was not statistically significant. No effect of active smoking was found in this study. [The Working Group considered this study to have limited generalizability (cases < 36 years of age) and noted that no exposure–response relationship was observed despite comparatively high point estimates of risk.]

A population-based case–control study conducted in Switzerland by Morabia et al. (1996) was designed specifically to evaluate the role of exposure to secondhand smoke in risk of breast cancer. It included 126 cases and 620 controls who were lifetime never-smokers. Never-smokers were defined as having smoked fewer than 100 cigarettes in a lifetime. Eligible cases were women less than 75 years of age who had been diagnosed with invasive breast cancer between 1 January 1992 and 31 October 1993. Population controls were selected from the official registers of residents of Geneva and were 30–74 years of age. This study included a detailed assessment of exposure to secondhand smoke and risk estimates were adjusted for the following potential confounders: age, education, body mass index, age at menarche, age at first live birth, oral contraceptive use, breast cancer in mother or sister, history of breast biopsy, alcohol intake and saturated fat intake. The referent unexposed group in this study included women who were never regularly exposed (< 1 (h/day) × years) to either active or passive smoking (28/244 cases and 241/1032 controls). Estimates of relative risk associated with any exposure to secondhand smoke, duration of exposure to secondhand smoke, exposure to spousal secondhand smoke only and duration of spousal smoking were all approximately three and were statistically significant; however, risk estimates stratified by duration (1–50 or > 50 (h/day) × years) were virtually identical and there was no suggestion of an exposure–response relationship. [The Working Group considered that the strengths of the study were its comprehensive assessment of exposure, being population-based and the large number of potential confounders included in the analysis. Concerns included the following: magnitude of the association between cancer and passive smoking is the same as that for active smoking in same study, no exposure–response relationship was found for secondhand smoke and the very restrictive reference category used may have biased the results.]

Morabia et al. (2000) next conducted a sub-study from the above-mentioned case–control study. It was designed to evaluate the role of N-acetyltransferase 2 (NAT2) in the relationship between breast cancer and active and passive smoking. Cases believed to be alive and living in Geneva in 1996–97 were re-contacted, as were a subset of controls, and asked to provide a buccal swab for DNA extraction and NAT2 genotyping for subsequent classification as slow or fast acetylators. This sub-study included 84 cases who were never-smokers and 99 controls who were never-smokers. As in the parent study, a three-fold increase in risk of breast cancer was associated with any reported exposure to secondhand smoke (relative risk, 3.1; 95% CI, 1.5–6.0). The association between exposure to secondhand smoke and breast cancer appeared to be modified by acetylation status; breast cancer risk was higher in persons with the fast acetylation genotype (relative risk, 5.9; 95% CI, 2.0–17.4) than in slow acetylators (relative risk, 1.9; 95% CI, 0.7–4.6). [The Working Group's comments on the parent study also applied to this sub-study.]

Millikan et al. (1998) conducted a population-based case–control study in North Carolina, USA, that also examined the effect of N-acetylation genotypes (NAT1 and NAT2), exposure to secondhand smoke and breast cancer risk. Cases included women between 20 and 74 years of age who were diagnosed with invasive primary breast cancer between May 1993 and December 1996. Controls less than 65 years of age were selected from files of the North Carolina Division of Motor Vehicles and those from 65 to 74 years of age from the United States Health Care Financing Administration (HCFA) files. All African–American cases and a sample of white cases were selected. This report was based on cases and controls who provided a blood sample. Cases and controls were broadly frequency-matched on race (African–American and white) and age (age less than 50 years and 50 years and above). Relative risk estimates were adjusted for age, race, age at menarche, age at first full-term pregnancy, parity, family history of breast cancer, breast biopsies showing benign tumours and alcohol consumption. Statistically non-significant increases in risk were associated with exposure to secondhand smoke after the age of 18 years in never-smokers (relative risk, 1.3; 95% CI, 0.9–1.9). The point estimates for premenopausal (relative risk, 1.5; 95% CI, 0.8–2.8) and postmenopausal women (relative risk, 1.2; 95% CI, 0.7–2.2) were not substantially different. Stratification by menopausal status and NAT1 and NAT2 genotypes resulted in statistically non-significant relative risks for all subgroups. The point estimates for exposure to secondhand smoke after the age of 18 years were highest for pre-menopausal never-smoking women for NAT1*10 (relative risk, 1.7; 95% CI, 0.7–4.3) and NAT2rapid (relative risk, 2.3; 95% CI, 0.9–6.2).

Marcus et al. (2000) included additional cases from this North Carolina study without the requirement for a blood sample and addressed the issue of exposure to secondhand smoke before the age of 18 years. Exposure to secondhand smoke at home during childhood showed no statistically significant association with risk of breast cancer in this study (relative risk, 0.8; 95% CI, 0.6–1.1). [The Working Group considered that the strengths of this study included the large number of never-smoking cases and controls; the multiethnic study population (although no ethnicity-specific risk estimates for exposure to secondhand smoke were reported and that the study investigated possible high-risk subgroups.]

Lash and Aschengrau (1999) reported the findings of a case–control study conducted in five towns in Massachusetts, USA. The incident cases of breast cancer were diagnosed from 1983–1986. Population controls from these towns for living cases under 65 years of age were selected using random-digit dialling and for women 65 years and older from the US Health Care Financing Administration (HCFA) files. Because deceased cases were also eligible for this study, deceased controls were selected from Massachusetts Department of Vital Statistics and Research. A total of 120 cases and 406 controls were never-smokers. About one-third of the interviews relating to cases and 45% of the interviews relating to controls were with proxy respondents. Age, parity, history of breast cancer other than index diagnosis, family history of breast cancer, history of benign breast disease, and history of radiation therapy were adjusted for in the analyses. A twofold increase in risk (relative risk, 2.0; 95% CI, 1.1–3.7) was associated with any exposure to secondhand smoke; however, increasing duration was not associated with increasing risk. The relative risk estimates for exposure to secondhand smoke and for active smoking also determined in this study were similar. [The Working Group noted that the limitations of the study were that the original study was not designed to evaluate exposure to secondhand smoke; it was unclear whether controls from the parent study which included three types of cancer cases were matched to breast cancer cases in this substudy, and that this substudy included a large number of proxy respondents.]

Findings from a small clinic-based case–control study, conducted in Orange County, California, USA, were reported by Delfino et al. (2000). Three breast cancer centres were included in the study. Subjects diagnosed with a suspicious breast mass detected either clinically or radiographically who were over the age of 39 years were considered to be eligible. Information on exposure was obtained from a self-administered questionnaire completed prior to biopsy in order to minimize recall bias and interviewer bias. Among women who were never-smokers, 64 were subsequently found to have malignant tumours and comprised the case series, and 149 never-smokers with benign breast disease were classified as controls. Risk estimates for exposure to secondhand smoke in the home were adjusted for age, menopausal status, age at menarche, age at first full-term pregnancy, total months of pregnancy, lactation history, education, race/ethnicity, body mass index and family history of breast cancer. NAT2 genotype was also determined, but was not associated with risk for breast cancer in this study. No statistically significant association was found between exposure to secondhand smoke and breast cancer risk in these never-smokers (relative risk, 1.3; 95% CI, 0.7–2.5) for any exposure to secondhand smoke in the home. [The Working Group considered that the strengths of this study included the fact that exposure data were collected prior to determination of case–control status. Its limitations were that it was a small study and that information on exposure to secondhand smoke was limited to exposure in the household.]

A large Canadian study (Johnson et al., 2000) identified population-based incident cases of breast cancer aged 25–74 years at diagnosis from the National Enhanced Cancer Surveillance System beginning in April 1994 in some provinces (later in others) and continuing until July 1997. The study included 378 premenopausal and 700 postmenopausal never-smoking cases and 369 pre- and 845 postmenopausal population-based never-smoking controls. Exposure to secondhand smoke in the household during childhood and adult life as well as in the workplace were assessed. Relative risk estimates were adjusted for age, province, education, body mass index, alcohol use, physical activity, age at menarche, age at the end of first pregnancy lasting 5 months or longer, number of live births, months of breastfeeding and height. There was no evidence of an association between breast cancer and exposure to secondhand smoke during childhood or adulthood in postmenopausal women (relative risk estimates ranged from 0.9 to 1.3, none were statistically significant). However, premenopausal women had significantly elevated risks for breast cancer associated with any exposure to secondhand smoke (relative risk, 2.3; 95% CI, 1.2–4.6), exposure to secondhand smoke during adulthood (relative risk, 2.6; 95% CI, 1.1–6.0), and exposure to secondhand smoke during both childhood and adulthood (relative risk, 2.6; 95% CI, 1.2–5.5). There was evidence of a strong dose–response relationship in premenopausal women associated with duration of residential and occupational exposure (p for trend = 0.0007). [The Working Group noted that the risk associated with passive smoking was similar in magnitude to that in former active smokers (relative risk, 2.6) and was higher than that for current active smokers (relative risk, 1.9) in the same study. The limitations of the study were that information was missing on a large number of cases and controls who were excluded from this study and information on exposure to secondhand smoke was available for only 59% of never-smokers.]

Two recent reports from a case–control study of breast cancer in German women aged 50 years and younger have used as the referent for assessing the risks of both active and involuntary smoking those women who have experienced no active and no passive exposure to tobacco smoke (lifetime non-exposed: < 1 (h/day) × year) (Kropp & Chang-Claude, 2002; Chang-Claude et al., 2002). The study included 706 cases (response rate, 70.1%) and 1381 controls (response rate, 61.2%). Data were initially collected by self-administered questionnaires for active smoking, and later, living cases and controls were re-contacted for information on involuntary exposure to tobacco smoke. Of the original participants, approximately 66% of the cases and 79% of the controls completed this part of the study. Risk estimates were adjusted for daily alcohol intake, total number of months of breastfeeding, education, first-degree family history of breast cancer, menopausal status and body mass index. For active smoking, a relative risk of 1.1 (95% CI, 0.6–2.0) was recorded, whereas never-smokers exposed to involuntary smoking had a statistically increased risk of about 60% (Kropp & Chang-Claude, 2002). In a subgroup analysis of 422 cases and 887 controls, the effect of NAT2 on the association between tobacco and breast cancer was considered (Chang-Claude et al., 2002). When compared to women who had never been exposed to any tobacco smoke, no association with active smoking was seen in rapid acetylators and a modest statistically non-significant increase in risk was observed in slow acetylators. In contrast, passive smoking was associated with a statistically non-significant risk that was higher in rapid than in slow acetylators (relative risk, 2.0; 95% CI, 1.0–4.1; and 1.2; 95% CI, 0.7–2.0, respectively). [The Working Group noted that this study has included many subgroup analyses, had reported incongruent findings related to active and involuntary smoking in the same study AND had obtained passive smoking data for only about 50% of study subjects. However, the strength of this study was the inclusion of a referent group of subjects who had not been exposed to any tobacco smoke during their lifetimes by self-report.]

2.3. Childhood cancers

Many studies have evaluated the association of cancer risk in childhood with exposure to parental smoking since this issue was considered previously in the IARC Monograph Volume 38 (IARC, 1986). These associations will be evaluated below for all cancers combined and separately, for brain tumours, leukaemias and lymphomas, and other childhood cancers.

Few studies distinguish times of exposure to tobacco smoke from parents, i.e. whether the exposure was preconception, in utero or postnatal. Exposure may have occurred in all three periods even when a study reports on only one, or exposure may also be reported as ‘ever’. Involuntary smoking during each of these time periods tends to be correlated, in particular exposure to secondhand smoke from the father because father's smoking habits are less likely to change.

2.3.1. All sites combined

Four cohort studies (Neutel & Buck, 1971; Golding et al., 1990; Pershagen et al., 1992; Klebanoff et al., 1996) and ten case–control studies (Buckley et al., 1986; McKinney et al., 1986; Stjernfeldt et al., 1986; Forsberg & Kallen, 1990; John et al., 1991; Golding et al., 1992; Sorahan et al., 1995; Ji et al., 1997; Sorahan et al., 1997a,b) (Table 2.13) have examined the role of involuntary exposure to tobacco smoke in risk for childhood cancers in general.

Table 2.13. Childhood cancers, all sites combined, and involuntary exposure to parental smoking.

Table 2.13

Childhood cancers, all sites combined, and involuntary exposure to parental smoking.

All four cohort studies specifically reported on the risk associated with cancer related to mothers' smoking during pregnancy. Neutel and Buck (1971) identified 97 deaths from childhood cancer in a cohort of 89 302 births from Ontario (Canada), and England and Wales followed from 7 to 10 years. Children with a mother who had smoked during pregnancy had a relative risk of 1.3 (95% CI, 0.8–2.2). No exposure–response relationship was apparent. [The Working Group noted several limitations of this study: no control for potential confounders; completeness of follow-up unknown, and limited assessment of exposure to secondhand smoke.]

Golding et al. (1990) followed a cohort of 16 193 births for 10 years (1970–80) and a total of 33 cancers were diagnosed. After adjustment for social class, exposure to X-rays during pregnancy, term delivery, administration of pethidine in labour and of drugs during infancy, a statistically significant increase in risk was found for children whose mothers smoked five or more cigarettes per day during the index pregnancy (relative risk, 2.5; 95% CI, 1.2–5.1). [The Working Group noted that the strength of this study was that the effect of exposure to secondhand smoke was independent of other risk factors found in this study. Its limitations were that the completeness of follow-up was unknown and that there was limited assessment of exposure to secondhand smoke.]

Pershagen et al. (1992), in Sweden, followed a large cohort of 497 051 births. In 5 years of follow-up, a total of 327 cancers that could be linked to data on maternal smoking were diagnosed. Relative risks were adjusted for year and county of birth, birth order and maternal age. No association was found for any maternal smoking during pregnancy (relative risk, 1.0; 95% CI, 0.8–1.3) and no exposure–response relationship was seen for number of cigarettes smoked during pregnancy (< 10 cigarettes per day, relative risk, 1.0; ≥ 10 cigarettes per day, relative risk, 0.9). No cancer at any of the sites evaluated individually was associated with maternal smoking. [The Working Group noted that the strengths of this study were that it was the largest cohort study, some statistical adjustment of risk estimates had been made and there was a high rate of follow-up. Its limitation was that there had been limited assessment of exposure to secondhand smoke.]

The most recent prospective study to evaluate the association between maternal smoking during pregnancy and childhood cancer is the US Collaborative Perinatal Project that included 54 795 children born from 1959–66 who were followed until the age of seven or eight years (Klebanoff et al., 1996). The hazard ratio for cancer in children whose mother smoked during pregnancy compared to those whose mother did not was 0.7 (95% CI, 0.4–1.2). Adjustment of the hazard ratio for maternal race, age, education, socioeconomic status, height and pre-pregnancy weight as well as previous pregnancies, exposure to diagnostic radiation during pregnancy, feeding of infant in hospital, sex of infant and date of delivery had only a minimal effect on the point estimates, all of which remained in the range of 0.6. [The Working Group noted that the limitations of this study were that the completeness of follow-up was unknown, but the estimates of expected incidence suggest that few cases were missed, and that assessment of exposure to secondhand smoke was limited.]

Buckley et al. (1986) conducted a case–control analysis using data from the US/Canada Children's Cancer Study Group. These investigators compared smoking by mothers and fathers of 1814 childhood cancer cases with that of parents of 720 controls selected at random from approximately the same geographical regions as cases. Smoking in the periods before and during pregnancy was assessed. No association was found between maternal smoking during pregnancy (< 10 cigarettes per day, relative risk, 1.3; 95% CI, 0.9–1.9; ≥ 10 cigarettes per day, relative risk, 1.0; 95% CI, 0.8–1.2) and no association with paternal smoking was found [relative risk not reported]. Adjustment for potential confounders such as year of birth, age of mother, illnesses during pregnancy and socioeconomic factors, did not alter findings. [The Working Group noted that the strength of this study was the large sample size and its limitations were that the report lacked details of the study and the control group was not well described.]

In a case–control study based on the Inter-Regional Epidemiological Study of Childhood Cancer in the United Kingdom, 555 cases of cancer in children < 15 years of age and 1110 controls matched for age and sex were compared for exposure to maternal smoking during pregnancy (McKinney et al., 1986). There was no evidence of an association between maternal smoking and risk for childhood cancer (1–10 cigarettes per day, relative risk, 1.1; 95% CI, 0.9–1.5; > 11 cigarettes per day, relative risk, 0.8; 95% CI, 0.7–1.1). [The Working Group noted that the strength of this study was the large sample size. The limitations were that the report provided few study details; other than matching for age and sex there was no adjustment for potential confounders, and there was limited assessment of exposure to secondhand smoke.] This dataset was recently re-evaluated (Sorahan et al., 2001). Microfilmed interview records of all study subjects were reviewed and information on parental cigarette smoking habits was re-abstracted. There was a statistically significant positive trend (p = 0.02) associated with daily paternal cigarette consumption before pregnancy for all cancers combined when cases were compared with controls selected from General Practitioners' (GPs') lists (n = 555), but no significant association was observed when cases were compared with hospital controls (n = 555). The opposite was seen for maternal smoking before pregnancy: an inverse trend (p < 0.001) was noted between daily cigarette consumption when cases were compared with hospital controls, but not when compared with GP controls. Risk estimates were adjusted for socioeconomic status, ethnicity, parental age at child's birth and other parent's smoking. [The Working Group noted that the two sets of controls produced very different results that are not easily explained.]

Stjernfeldt et al. (1986) reported the findings of a nationwide case–control study in Sweden that included 305 cases of cancer in children ≤ 16 years of age and 340 children with insulin-dependent diabetes mellitus who served as controls. Estimates of relative risk were adjusted for year of child's birth and maternal age, illness during pregnancy, occupation and place of residence. A 50% (p < 0.01) increase in risk for cancer was associated with in-utero exposure to maternal smoking. [The Working Group noted that the strengths of the study included the good response rate and the attempt to control for response bias by using children with diabetes mellitus as controls; its limitation was that the appropriateness of the control group was unknown.]

A case–control study from Sweden by Forsberg and Kallen (1990) found no association between childhood cancers and maternal smoking (relative risk, 1.1; 95% CI, 0.6–2.0) based on 69 cases and 139 controls for whom maternal smoking status was known. [The Working Group noted that the limitations of this study included the small sample size, uncertainty as to whether original case–control matching also applied to the substudy sample and the limited assessment of exposure.]

John et al. (1991) evaluated both maternal and paternal prenatal smoking histories in relation to risk for childhood cancer. The study included 223 incident cases < 15 years of age diagnosed from 1976 to 1983 in Denver, CO, USA. Controls were selected using random digit dialling and were matched to cases on age, sex and telephone exchange, and 196 controls were included in the analysis. Mothers' and fathers' smoking was highly correlated. Of the 109 children exposed to mother's smoking during the first trimester, 81% were also exposed to father's tobacco smoking while an additional 105 children were exposed to father's smoking alone. Mother's smoking during the first trimester was associated with a modest statistically nonsignificant increase in risk for childhood cancer after adjustment for father's education (relative risk, 1.3; 95% CI, 0.7–2.1). Children whose mothers did not smoke who were exposed to father's smoking also had a modest, statistically non-significant increase in risk (relative risk, 1.2; 95% CI, 0.8–2.1). [The Working Group noted that this study included a more detailed assessment of exposure to secondhand smoke than did earlier studies.]

Golding et al. (1992) conducted a case–control study in the United Kingdom to assess the association of childhood cancer with administration of intramuscular vitamin K and pethidine during labour. Data on mothers' smoking during pregnancy as a potential confounder were collected. A twofold increase in risk (relative risk, 2.0; 95% CI, 1.3–3.2) adjusted for year of delivery was found. [The Working Group noted that this study was not designed to investigate exposure to secondhand smoke, that only maternal smoking during pregnancy was recorded and that there was only minimal control for potential confounders.]

Three reports from the Oxford Survey of Childhood Cancer (OSCC) provided data from large case–control studies of childhood cancer deaths during different time periods: 1977–81 (Sorahan et al., 1995), 1953–55 (Sorahan et al., 1997a) and 1971–76 (Sorahan et al., 1997b). The first report in 1995 included 1641 cases and an equal number of controls. There was no association with prenatal maternal cigarette smoking; however, paternal smoking was associated with a statistically significant positive trend (p for trend = 0.003). When cigarette use by one or both parents was adjusted for social class, maternal age at birth and use of alcohol, the relative risk was 1.4 (95% CI, 1.1–1.7) for father's use of cigarettes and 1.4 (95% CI, 1.1–1.7) for cigarette use by both parents, whereas cigarette use by mother was not statistically significantly associated with an increased risk (relative risk, 1.2; 95% CI, 1.0–1.6). A total of 1549 deaths from childhood cancer between 1953 and 1955 and 1549 matched healthy controls were used to further investigate the earlier findings from the OSCC (Sorahan et al., 1997a). After adjustment for smoking by the spouse, social class, age of father, age of mother, birth order, and exposure to obstetric radiography, no statistically significant dose–response trend was found to be associated with maternal smoking, but maternal smoking only was associated with a 30% increased risk for childhood cancer (relative risk, 1.3; 95% CI, 1.1–1.5). At the highest level of paternal smoking (> 20 cigarettes per day), a clear trend was noted (p for trend < 0.001) with a relative risk of 1.4 (95% CI, 1.1–1.9); paternal smoking only was also associated with increased risk (relative risk, 1.7; 95% CI, 1.3–2.2). The third report (Sorahan et al., 1997b) which examined deaths from 1971 to 1976 provided very similar results to those in the first two reports, i.e. no clear association with childhood cancer was evident for maternal smoking and there was a statistically significant positive trend for paternal smoking. [The Working Group noted the very large sample sizes, the consistent findings over time, the adjustment for potential confounders and the assessment of exposure from mothers and fathers with data for trends.]

A large case–control study in China by Ji et al. (1997) also examined paternal smoking and risk for cancer in children (< 15 years of age) of nonsmoking mothers. Relative risks were adjusted for birth weight, income, paternal age, education and alcohol drinking. For all sites combined, the relative risk for ‘ever smoking’ by the father was 1.3 (95% CI, 1.0–1.7). Statistically significant trends were found for duration of paternal smoking (p for trend = 0.007) and pack–years (p for trend = 0.01), but not age of starting smoking (p for trend = 0.28) or cigarettes per day (p for trend = 0.07). [The Working Group noted the large sample size, the minimization of exposure misclassification by including only children of nonsmoking mothers, the adjustment for potential confounders and the extensive exposure assessment for fathers.]

Boffetta et al. (2000) conducted a meta-analysis of childhood cancers associated with passive exposure to smoke based on the random effects model. The relative risk estimate for maternal smoking during pregnancy for all cancers combined included all cohort studies and eight of the ten case–control studies listed in Table 2.13 (Sorahan et al. 1997b; Ji et al. 1997; were not included). The results suggest a small increase in risk for all cancers for maternal smoking during pregnancy (relative risk, 1.1; 95% CI, 1.0–1.2), but not for specific cancer sites. Results on exposure before and after pregnancy were too sparse for any conclusion to be drawn. Studies of exposure to paternal tobacco smoke and risk for all cancers combined are fewer than those addressing maternal smoking and no relative risk was reported in this meta-analysis.

2.3.2. Brain and central nervous system

Table 2.14 lists one cohort study (Pershagen et al. 1992) and 15 case–control studies (Gold et al., 1979; Preston-Martin et al., 1982; Stjernfeldt et al., 1986; Howe et al., 1989; Kuijten et al., 1990; Gold et al., 1993; Bunin et al., 1994; Cordier et al., 1994; Filippini et al., 1994; McCredie et al., 1994; Norman et al., 1996; Ji et al., 1997; Sorahan et al., 1997a,b; Filippini et al., 2000) that have examined parental smoking and risk for brain tumours or for all tumours of the central nervous system combined.

Table 2.14. Tumours of the brain and central nervous system and involuntary exposure to parental smoking.

Table 2.14

Tumours of the brain and central nervous system and involuntary exposure to parental smoking.

Only the cohort study of Pershagen et al. (1992) (see section 2.3.1) has published a relative risk for tumours of the central nervous system. No association was found between maternal smoking in pregnancy and risk for tumours of the central nervous system.

The first case–control study to examine risk for brain tumour and maternal smoking was reported by Gold et al. (1979). This study was conducted in the USA and included 84 children with brain tumours and two control groups. One control group comprised 78 children with other malignancies matched on sex, race, date and age at diagnosis, and the other, 73 children selected from the state birth certificate file and matched on sex, date of birth and race. Risk associated with maternal smoking before and during pregnancy was associated with large non-statistically significant risks for childhood brain tumour that were based on a small sample size.

Preston-Martin et al. (1982) reported the findings from a larger case–control study in the USA designed to evaluate the risk for brain tumour associated with childhood exposure to N-nitroso compounds, including those from tobacco smoke. No increased risk was associated with maternal smoking, but a relative risk of 1.5 (p = 0.03) was found for children of mothers living with a smoker during pregnancy. The small Swedish case–control study by Stjernfeldt et al. (1986) (see section 2.3.1) found no increased risk for tumours of the central nervous system associated with maternal smoking in pregnancy.

An exploratory case–control study of brain tumours in Canadian children diagnosed in Ontario between 1977 and 1983 included 74 cases and 138 age- and sex-matched controls. The study found neither maternal nor paternal smoking during pregnancy to be statistically significantly associated with risk for brain tumours (Howe et al., 1989). Similarly, a population-based case–control study in the USA of childhood astrocytomas that included 163 case–control pairs found no increased risk associated with any smoking by either mother or father (Kuijten et al., 1990).

A large population-based case–control study in the USA of childhood brain tumours examined smoking by both parents in some detail. The study included exposure assessments for the preconception period as well as the pre- and postnatal period (year of birth of child) and dose–response estimates (Gold et al., 1993). There was no statistically significant association between risk for brain tumours and any indicator of parental smoking. [The Working Group noted that this was a well-conducted study designed to examine parental smoking in detail, and having sufficient statistical power.]

Bunin et al. (1994) studied the two most common types of brain tumour, astrocytoma and primitive neuroectodermal tumour, in children less than six years of age. Controls, selected by random-digit dialling, were matched to cases on race, year of birth, and telephone area code and prefix. Estimates of relative risk for astrocytoma were adjusted for income level, but primitive neuroectodermal tumour estimates were unadjusted. No association was found between either of these types of tumour and maternal active (ever and/or during pregnancy) or passive smoking (during pregnancy) or paternal smoking (ever and/or during pregnancy).

A non-statistically significant increase in risk for brain tumours (relative risk, 1.6; 95% CI, 0.7–3.5) associated with any smoking by the mother was found in a small case–control study in France (Cordier et al. 1994). Filippini et al. (1994) in Italy assessed the risk associated with active and passive smoking by mothers during pregnancy in a case–control study with 91 cases. Active smoking by the mother during pregnancy was associated with a relative risk of 1.7 (95% CI, 0.8–3.8); no dose–response relationship was observed. Relative risks were adjusted for education level. Among nonsmoking mothers, the relative risks for light and heavy exposure to secondhand smoke were 1.7 (95% CI, 0.8–3.6) and 2.2 (95% CI, 1.1–4.5; p trend = 0.02). McCredie et al. (1994) conducted another small, population-based case–control study of brain tumours in Australia. Two controls were matched to each case by age and sex. No association was found with exposure to tobacco smoke from another member of the household, but no risk estimates were provided. [The Working Group noted that the limitations of these studies were that they lacked statistical power; there was limited adjustment for potential confounders and limited assessment of exposure.]

The findings from a large, population-based case–control study of brain tumours in children < 15 years of age diagnosed from 1984 to 1991 provided no support for an association between brain tumour risk and maternal or paternal smoking before pregnancy or maternal smoking during pregnancy (Norman et al. 1996). Risk estimates were at or below unity and there was no evidence of a relationship between risk for brain tumours and amount or timing of exposure. [The Working Group noted that the strengths of this study were that it was large and included a relatively detailed assessment of exposure.]

Three studies discussed previously (Ji et al. 1997; Sorahan et al. 1997a,b) found no increased risk for brain tumours associated with father's smoking (Ji et al., 1997) or of tumours of the central nervous system associated with maternal or paternal smoking (Sorahan et al., 1997a,b; 2001).

Filippini et al. (2000) in northern Italy, conducted a population-based case–control study of childhood tumours of the central nervous system with cases diagnosed from 1988 to 1993. Cases from their previous study (Filippini et al., 1994) were excluded. Active smoking by parents before pregnancy was not associated with increased risk. Active smoking by mothers in early pregnancy was associated with a small increase in risk (relative risk, 1.5; 95% CI, 1.0–2.3). An increase in risk was also associated with passive smoking by nonsmoking mothers in early pregnancy (relative risk, 1.8; 95% CI, 1.2–2.6) and late pregnancy (relative risk, 1.7; 95% CI, 1.2–2.5).

The results of the meta-analysis by Boffetta et al. (2000) indicated no significant increase in risk for tumours of the central nervous system associated with maternal smoking during pregnancy (relative risk, 1.0; 95% CI, 0.9–1.2), but exposure to paternal smoking suggested an increased risk for brain tumours (relative risk, 1.1; 95% CI, 1.1–1.4). [The Working Group noted that this meta-analysis included two studies of neuroblastoma and one study of retinoblastoma with tumours of the central nervous system.]

2.3.3. Leukaemias and lymphomas

The only cohort study to report specifically on lymphatic and haematopoietic cancers (Pershagen et al., 1992) and 16 case–control studies with data on one or more of these types of malignancy are included in Table 2.15 (Manning & Carroll, 1957; Stewart et al., 1958; Van Steensel-Moll et al., 1985; Buckley et al., 1986; McKinney et al., 1986; Stjernfeldt et al., 1986; Magnani et al., 1990; John et al., 1991; Roman et al., 1993; Severson et al., 1993; Shu et al., 1996; Ji et al., 1997; Sorahan et al., 1997a,b; Brondum et al., 1999; Infante-Rivard et al., 2000).

Table 2.15. Childhood leukaemias and lymphomas and involuntary exposure to parental smoking.

Table 2.15

Childhood leukaemias and lymphomas and involuntary exposure to parental smoking.

A total of 129 lymphatic and haematopoietic cancers were diagnosed during 5 years of follow-up in the Swedish cohort (Pershagen et al., 1992). No association was observed between the development of these cancers and smoking during pregnancy or any amount of smoking by the mother.

Manning and Carroll (1957) found no difference in the proportion of mothers of children with leukaemia who smoked 10 or more cigarettes per day at the time of interview when compared to control mothers (39% versus 38%) and a somewhat lower proportion of mothers of children with lymphoma (31%) who smoked at that level. A second early study (Stewart et al., 1958) reported a very small but statistically significant increase in risk for death from leukaemia among children of mothers who had ever smoked (relative risk, 1.1; p < 0.04). [The Working Group noted that neither study was designed specifically to study the effects of involuntary smoking; only unadjusted proportions were reported.]

Van Steensel-Moll et al. (1985) found no association between maternal smoking in the year before pregnancy and risk for acute lymphocytic leukaemia in a study in the Netherlands designed to assess maternal fertility problems and this risk. [The Working Group noted that the strength of this study was the large number of cases. Its limitations are the limited assessment of exposure and the questionable time period.] The case–control study in Sweden by Stjernfeldt et al. (1986) included 157 cases of acute lymphoblastic leukaemia, 16 cases of non-Hodgkin lymphoma and 15 cases of Hodgkin disease. A statistically significant positive trend (p trend < 0.01) was found for number of cigarettes smoked per day by the mother during pregnancy and risk for acute lymphoblastic leukaemia. No statistically significant association with smoking was observed for either non-Hodgkin lymphoma or Hodgkin disease based on a very small number of cases.

McKinney et al. (1986) found no association between maternal smoking during pregnancy and risk for childhood leukaemia or lymphoma. Buckley et al. (1986) also failed to find an association between maternal smoking during pregnancy in their large study that included 742 cases of acute lymphocytic leukaemia and 169 cases of non-Hodgkin lymphoma.

Magnani et al. (1990) found no association between acute lymphocytic leukaemia, other leukaemias or non-Hodgkin lymphoma during childhood and the mother's smoking up to the time of the child's birth. The father's history of smoking was associated with a risk for non-Hodgkin lymphoma (relative risk, 6.7; 95% CI, 1.0–43.4), but not for acute lymphocytic leukaemia or other leukaemias. This Italian hospital-based case–control study included 142 cases of acute lymphocytic leukaemia, but only a small number of non-Hodgkin lymphoma (n = 19) and other types of leukaemia (n = 22). Risk estimates were adjusted for socioeconomic status only.

The case–control study in the USA reported by John et al. (1991) included 73 cases of leukaemia and 26 cases of lymphoma. Statistically significant increases in risk were associated with maternal smoking 3 months before conception for acute lymphocytic leukaemia; with smoking during the first trimester for acute lymphocytic leukaemia; and during all three trimesters for acute lymphocytic leukaemia (relative risk, 2.5; 95% CI, 1.2–5.4) and lymphoma (relative risk, 2.7; 95% CI, 1.0–7.6).

A US–Canadian case–control study of acute myeloid leukaemia found no association between risk for acute myeloid leukaemia and maternal smoking before, during or after pregnancy (Severson et al. 1993). No association was observed with smoking by the father, but this was not quantified. [The Working Group noted the reasonably detailed exposure assessment, but although relative risks were adjusted for potential confounders, the factors were not named.]

A small case–control study of leukaemia and non-Hodgkin lymphoma in the United Kingdom examined maternal smoking based on obstetric notes and by interview (Roman et al., 1993). Both relative risks were below unity. [The Working Group noted that very little information was provided, that no adjustment was made for confounders, and the small size of the sample.]

Shu et al. (1996) found that maternal smoking during pregnancy was negatively associated with risk for leukaemia (all leukaemias, acute lymphocytic leukaemia or acute myeloblastic leukaemia) in infants. Paternal smoking one month prior to pregnancy was related to an elevated risk for acute lymphocytic leukaemia (relative risk, 1.6; 95% CI, 1.0–2.4), but not acute myeloblastic leukaemia and smoking by the father during pregnancy did not lead to a statistically significant increase in risk for any type of leukaemia. [The Working Group noted that the strengths of this study included the relatively detailed exposure from mothers' and fathers' smoking, and the adjustment for some potential confounders (sex, parental age, education and alcohol consumption by the mother during pregnancy).]

The case–control study of paternal smoking and childhood cancer in China reported by Ji et al. (1997) included 166 cases of acute leukaemia and 87 of lymphoma. No statistically significant association with paternal smoking was found for leukaemia, although a borderline positive trend was found for the father's number of pack–years of smoking (trend, p = 0.06). The father's smoking was associated with a fourfold increase in risk for lymphoma (relative risk, 4.0; 95% CI, 1.3–12.5) and statistically significant positive dose–response trends for lymphoma were observed for number of years smoked preconception and pack–year history, but not for number of cigarettes smoked per day.

Sorahan et al. (1997a) reported a modest association between risk for acute lymphocytic leukaemia and maternal smoking (relative risk, 1.2; 95% CI, 1.0–1.5), but no increased risk was found for myeloid, monocytic or other types of leukaemia or lymphoma. This study found no relationship between paternal smoking and any type of leukaemia, but the risk estimate for lymphoma was 1.4 (95% CI, 1.0–1.8). No increased risks associated with parental smoking were found when cases and controls from a later time period, 1971–76, were examined (Sorahan et al. 1997b).

A large case–control study in the USA of parental cigarette smoking and risk for acute leukaemia collected detailed information on exposure to smoke from the mothers and fathers of 1842 children with acute lymphocytic leukaemia and 517 with acute myeloblastic leukaemia and controls matched on age, race, and telephone area code/exchange (Brondum et al., 1999). There was no association between risk for acute lymphocytic leukaemia and ever smoking by the father (relative risk, 1.0; 95% CI, 0.9–1.2) or mother (relative risk, 1.0; 95% CI, 0.9–1.2); similarly, no associations were observed between acute myeloblastic leukaemia and ever smoking by the father (relative risk, 0.9; 95% CI, 0.7–1.2) or the mother (relative risk, 1.0; 95% CI, 0.7–1.2). Parental smoking during or around the time of the index pregnancy was not related to risk, nor were the number of cigarettes smoked, years of smoking or pack–years. Risk estimates were adjusted for household income, mother's race and education and father's race and education. [The Working Group noted the good statistical power and the detailed histories of both parents and also that some adjustment has been made for potential confounders.]

A case–control study in Canada of acute lymphocytic leukaemia assessed the role of parental smoking and CYP1A1 genetic polymorphisms (Infante-Rivard et al., 2000). There was no statistically significant association between parents' smoking and leukaemia overall. However, a substudy that included 158 of the 491 cases suggested that the effect of parental smoking may be modified by variant alleles in the CYP1A1. CYP1A1*2Btended to decrease risks and CYP1A1*2A and CYP1A1*4 increased the risks associated with smoking in the second and third trimesters. [The Working Group noted that this was the first study to look at the interaction between parental smoking, CYP1A1 and leukaemia.]

Sorahan et al. (2001) (see Section 2.3.1) found a statistically non-significant positive association between risk for acute lymphocytic leukaemia and daily cigarette consumption by fathers before pregnancy and a statistically non-significant inverse association between risk for acute lymphocytic leukaemia and daily smoking by mothers before pregnancy.

The results of the meta-analysis for maternal smoking during pregnancy indicated that there were no statistically significant associations for all lymphatic and haematopoietic neoplasms (relative risk, 1.0; 95% CI, 0.9–1.2), for non-Hodgkin lymphoma or total lymphomas (relative risk, 1.1; 95% CI, 0.9–1.5) or for all leukaemias, acute leukaemia or acute lymphocytic leukaemia (relative risk, 1.1; 95% CI, 0.8–1.3) (Boffetta et al., 2000). The authors found evidence of publication bias for the data available on lymphomas (p = 0.04). Published studies with a small number of cases reported positive associations between exposure to tobacco smoke and childhood leukaemia, whereas larger studies showed no association. This suggests that small studies that had found no association or a negative association failed to be published. The meta-analysis for paternal smoking indicated no statistically significant association with acute lymphocytic leukaemia, but a twofold increase in risk for non-Hodgkin lymphoma (relative risk, 2.1; 95% CI, 1.1–4.0).

2.3.4. Other childhood cancers

Several other types of childhood cancer have been studied in relation to parental smoking in epidemiological investigations.

The cohort study by Pershagen et al. (1992) reported no statistically significant associations between mother's smoking during pregnancy and kidney cancer (30 cases; relative risk, 0.6; 95% CI, 0.2–1.5), eye tumours (28 cases; relative risk, 1.4; 95% CI, 0.6–2.8), endocrine tumours (13 cases; relative risk, 1.9; 95% CI, 0.6–6.0) or tumours of the connective tissue and muscle (15 cases; relative risk, 1.2; 95% CI, 0.4–3.6).

Magnani et al. (1989) conducted a hospital-based case–control study of soft-tissue sarcomas in Italy during 1983–84. The cases included 36 children with rhabdomyosarcoma and 16 cases of other soft-tissue sarcomas who were compared with 326 controls from the same hospitals. No associations were found between soft-tissue sarcoma or rhabdomyosarcoma and either mother's or father's smoking (all point estimates of relative risks were below unity). Smoking during several time periods, before, during and after birth was then looked at separately and the results were the same as for any smoking by the parents. [The Working Group noted that this was a small study, but that the exposure assessment included different time periods.]

Two studies in the USA (Holly et al., 1992; Winn et al., 1992) examined risk factors for Ewing's sarcoma. In their population-based study, Holly et al. (1992) looked at 43 cases and 193 controls selected by random digit dialling and matched to cases by sex and age. This tumour was not associated with smoking by the mother during pregnancy (relative risk, 1.1; 95% CI, 0.5–2.4) or by the father (relative risk, 0.9; 95% CI, 0.4–1.9). Risk estimates were adjusted for agricultural occupation of the father, poison or overdose of medication, area of residence, year of child's birth and income. [The Working Group noted that this was a small study that had made a detailed assessment of many factors, but less for parental smoking.] Winn et al. (1992) reported the findings of a larger case–control study that included 208 cases throughout the USA and two control groups with equal numbers of controls (sibling controls and regional controls). When cases were compared to regional controls, no significant risk estimates were found for smoking by either parent; however, parents were more likely to have smoked during pregnancy with the child with Ewing's sarcoma than during the pregnancy with the unaffected sibling; if only the mother smoked, the relative risk was 1.5 (95% CI, 0.3–9.0); if only the father smoked, the relative risk was 3.1 (95% CI, 0.7–14.0); if both parents smoked, the relative risk was 7.3 (95% CI, 1.3–41.6).

Two case–control studies in the USA evaluated prenatal drug consumption by the mother and risk for neuroblastoma (Kramer et al., 1987; Schwartzbaum, 1992). The first study was population-based and included 104 cases diagnosed from 1970 to 1979, a first group of 104 controls matched on date of birth, race and the first five digits of case's telephone number and a second group of controls comprising siblings of the index case. No significant increase in risk was associated with maternal smoking during pregnancy when cases were compared to either control group. The second study compared 101 newly diagnosed cases of neuroblastoma and 690 controls diagnosed with other types of childhood cancer at St Jude Children's Research Hospital. Cigarette smoking by the mother during pregnancy was found to increase the risk for neuroblastoma (relative risk, 1.9; 95% CI, 1.1–3.2). [The Working Group noted the questionable appropriateness of the control group in the study by Schwartzbaum and the limited exposure assessments in both studies.]

Olshan et al. (1993) reported findings from the National Wilms Tumour Study, a case–control study from a national collaborative clinical trial group in the USA. The study was conducted using interviews with 200 cases and 233 matched controls identified by random-digit dialling. No association was found for mother's smoking during pregnancy and risk for Wilms tumour (relative risk for smoking ten or more cigarettes per day, 0.7; 95% CI, 0.4–1.3).

2.4. Other cancers

2.4.1. All cancer sites combined

Hirayama (1984) reported a statistically significant association (p for trend < 0.001) between husband's smoking and cancer mortality in wives for all sites combined in the Japanese cohort (relative risk for former smoker: 1–19 cigarettes per day, 1.1; 95% CI, 1.0–1.2; relative risk for ≥ 20 cigarettes per day, 1.2; 95% CI, 1.1–1.4).

Sandler et al. (1985b), in their study previously described in detail (Section 2.2.2), found an increased risk of all cancers combined among nonsmokers passively exposed to cigarette smoke in adulthood (relative risk, 2.1; 95% CI, 1.4–3.0). Risk did not differ according to race (white or non-white), but was statistically significant only among women aged 30–49 years.

Miller (1990) reported the findings from a case–control study in the USA of cancer deaths among nonsmoking women in which next-of-kins were interviewed by telephone. Data on 906 nonsmoking wives were included in this report. The cases were women who had died of any type of cancer and the controls were nonsmoking wives who had died of cardiovascular, respiratory, kidney and other non-cancer diseases, excluding trauma. A nonsmoker was defined as a person who had smoked fewer than 20 packs of cigarettes during her lifetime. The percentage of deaths from cancer among non-exposed, non-employed wives was 2.2%; for exposed, non-employed wives, 18.9%, and for employed wives, 34.3% (p < 0.001). [The Working Group noted that the study used a questionable comparison group and a non-standard definition of a nonsmoker.]

2.4.2. Cervical cancer

Three Asian cohort studies described in Section 2.1 also reported on involuntary smoking and risk for cancer of the cervix. Risk for cervical cancer associated with involuntary exposure to smoking in nonsmokers was examined in a Japanese cohort study that found no significant increase in risk associated with husbands' smoking (Hirayama, 1984). A second cohort study also considered exposure to husbands' smoking and risk for cervical cancer in nonsmoking Korean women (Jee et al., 1999). The relative risk based on 203 cases of cervical cancer in nonsmokers was 0.9 (95% CI, 0.6–1.3) for women married to former and 0.9 (95% CI, 0.6–1.2) for women married to current smokers when compared with women married to nonsmokers. The cohort study by Nishino et al. (2001) included 11 incident cases of cervical cancer. Again, no association with husband's smoking status was observed (relative risk, 1.1; 95% CI, 0.3–4.5). [The Working Group noted that these cohort studies consistently indicated no association between exposure to secondhand smoke and cervical cancer.]

The case–control study from the USA reported by Sandler et al. (1985b; see Section 2.2.2) found an increased risk of cervical cancer associated with spousal smoking (relative risk, 2.1; 95% CI, 1.2–3.9). A second case–control study in the USA was conducted from 1984 to 1987 in Utah where a large percentage of the population are members of the Church of Jesus Christ of the Latter-day Saints which proscribes tobacco smoking (Slattery et al., 1989). The cases were population-based and controls were selected by random-digit dialling and matched to cases on age and county of residence. The response rates for cases and controls were 66% and 76%, respectively. Nonsmokers involuntarily exposed for 3 hours or more per day to secondhand smoke were found to have an increased risk for cervical cancer (relative risk, 3.4; 95% CI, 1.2–9.5). Self-characterized exposure to ‘a lot’ of secondhand smoke was also associated with increased risk (in-home relative risk, 2.9; 95% CI, 1.1–7.9; outside the home relative risk, 1.6; 95% CI, 0.6–4.5). [The Working Group noted that a statistically non-significant increase in risk was also observed in active smokers exposed to smoking by others.]

Coker et al. (1992) examined the risk of exposure to secondhand smoke in a case–control study of cervical intraepithelial neoplasma (CIN) of grades II (n = 40) and III (n = 63) in the USA. No statistically significant association was found between exposure to secondhand smoke and CIN II/III in nonsmokers, after adjustment for age, race, education, number of partners, contraceptive use, history of sexually transmitted disease and history of Pap smear. Another case–control study conducted in the USA compared 582 women with abnormal Pap smears (class 2–4) with 1866 controls with normal cytology (Scholes et al., 1999). Nonsmokers exposed to secondhand smoke from spouses, partners or other household members were found to have a borderline increase in risk for abnormal cervical cytology compared to nonsmokers who were not exposed to these sources of secondhand smoke (relative risk, 1.4; 95% CI, 1.0–2.0). Risk estimates were adjusted for age, age at first sexual intercourse, and number of sexual partners during lifetime.

2.4.3. Gastrointestinal cancers

The incidence of colorectal cancer in relation to passive exposure to smoke, which was defined as having lived with a person who smoked, was examined in a 12-year prospective cohort study in Washington County, MD, USA (Sandler et al., 1988). A statistically significant reduction in risk for colorectal cancer was observed for nonsmoking women who were involuntarily exposed to smoking (relative risk, 0.7; 95% CI, 0.6–1.0), but an increased risk for this cancer was found for nonsmoking men exposed to secondhand smoke in the household (relative risk, 3.0; 95% CI, 1.8–5.0).

In a Swedish population-based case–control study, Gerhardsson de Verdier et al. (1992) found an increased risk for colon cancer in women (relative risk, 1.8; 95% CI, 1.2–2.8) and rectal cancer in men (relative risk, 1.9; 95% CI, 1.0–3.6) in association with passive smoking after adjustment for numerous potential confounders. [The Working Group noted that it is unclear whether the analysis was restricted to never-smokers.]

A large Canadian case–control study of 1171 patients newly diagnosed with histologically confirmed stomach cancer and 2207 population controls evaluated the risk associated with active and passive smoking (Mao et al., 2002). Response rates of approximately 65% were obtained for both cases and controls. The analysis of passive smoking was conducted in male never-smokers (132 cases, 343 controls). Questionnaires were mailed to respondents and provided information on lifetime exposure to secondhand smoke through residential and occupational histories and also looked at source, intensity, and duration of exposure. Risk estimates for passive smoking were adjusted for 10-year age group, province of residence, education, social class, total consumption of meat and total consumption of vegetables, fruits and juices. A positive trend (p = 0.03) in risk for cancer of the gastric cardia was associated with lifetime exposure to secondhand smoke (sum of years of residential plus occupational exposure) in male never-smokers. At the highest level of exposure (≥ 43 years), the relative risk was 5.8 (95% CI, 1.2–27.5). No increased risks or trends were associated with risk for distal gastric cancer. Risks assessed by subsite (cardia and distal), were similar for active and passive smoking.

2.4.4. Nasopharyngeal and nasal sinus cavity cancer

The relationship between involuntary exposure of nonsmokers to secondhand smoke and risk for these rare cancers of the upper respiratory tract has been examined in one cohort study (Hirayama, 1984) and four case–controls studies (Fukuda & Shibata, 1990; Zheng et al., 1993; Cheng et al., 1999; Yuan et al., 2000). A positive association was found in most of these studies.

Hirayama (1984) found an increased risk of nasal sinus cancer in women (histology not noted) associated with increasing numbers of cigarettes smoked by husbands of nonsmoking women. When compared with nonsmoking women married to nonsmokers, wives whose husbands smoked had a relative risk of 1.7 (95% CI, 0.7–4.2) for 1–14 cigarettes per day, 2.0 (95% CI, 0.6–6.3) for 15–19 cigarettes per day and 2.55 (95% CI, 1.0–6.3) for ≥ 20 cigarettes per day (p for trend = 0.03).

Fukuda and Shibata (1990) reported the results of the first Japanese case–control study based on 169 cases of squamous-cell carcinoma of the maxillary sinus and 338 controls matched on sex, age and residence in Hokkaido, Japan. Among nonsmoking women, a relative risk of 5.4 (p < 0.05) was associated with exposure in the household to secondhand smoke from one or more smokers. Active smoking was associated with an increased risk for squamous-cell carcinoma in men in the same study.

Zheng et al. (1993) used data from the 1986 US National Mortality Followback Survey to assess risk for cancer of the nasal cavity and sinuses in relation to exposure to secondhand smoke in white men. Atotal of 147 deaths from cancer of the nasal cavity and sinuses were compared to 449 controls who had died from one of a variety of causes (excluding any causes strongly linked to alcohol and/or tobacco use). Data were obtained from postal questionnaires completed by next-of-kins. Among nonsmokers, patients with nasal cancer were more likely to have a spouse who smoked cigarettes (relative risk, 3.0; 95% CI, 1.0–8.9) after adjustment for age and alcohol use. When the analysis of cases was restricted to those with cancer of the maxillary sinus, the risk was somewhat higher (relative risk, 4.8; 95% CI, 0.9–24.7). The risks reported for active and for involuntary smoking were of similar magnitude in this study.

Neither involuntary exposure to tobacco smoke during childhood nor exposure during adult life were positively associated with an increased risk for nasopharyngeal cancer in a study in China (Province of Taiwan) (Cheng et al., 1999). Although histological type was not specified, all cases were histologically confirmed. Among never-smokers, the risk estimates for cumulative exposure to passive smoking (pack–person–years) in childhood declined as exposure increased (p for trend = 0.05); a similar but non-significant inverse relationship was found for exposure during adulthood. Significant elevations in risk of nasopharyngeal cancer were observed for active smokers in this study. [The Working Group noted that the exposure assessment was relatively detailed and that the estimates of relative risk were adjusted for age, sex, education and family history of nasopharyngeal cancer.]

A large population-based case–control study conducted in Shanghai, China, included 935 cases of nasopharyngeal carcinoma and 1032 population controls randomly selected from a population-registry and frequency-matched by sex and 5-year age group (Yuan et al., 2000). All cases were histologically confirmed, but the cell type was not specified. The study subjects were interviewed face to face, and the response rates were 84% for cases and 99% for controls. In female never-smokers, a consistent increase in risk related to exposure to secondhand smoke during childhood was noted. If the mother smoked, the relative risk was 3.4 (95% CI, 1.4–8.1); if the father smoked, the relative risk was 3.0 (95% CI, 1.4–6.2); if another household member smoked, the relative risk was 2.7 (95% CI, 1.1–6.9), and if any household member smoked, the relative risk was 3.0 (95% CI, 1.4–6.2). Risks associated with exposure to secondhand smoke during adulthood in women were also statistically significantly increased. For male never-smokers, the associations were weaker and were not statistically significant for exposure during childhood and adulthood. Gender-specific risk estimates were adjusted for age, level of education, consumption of preserved foods, oranges and tangerines, exposure to rapeseed oil, exposure to burning coal during cooking, occupational exposure to chemical fumes, history of chronic ear and nose conditions and family history of nasopharyngeal cancer. [The Working Group noted that this was a large, well-conducted study that included a detailed exposure assessment and adjustment for numerous potential confounders.]

2.4.5. Tumours of the brain and central nervous system

A population-based case–control study of patients with incident primary brain tumours diagnosed from 1987 through 1990 in Adelaide, Australia, was reported by Ryan et al. (1992). Controls were selected from the Australian electoral rolls which cover 95% of the population. Response rates of 90% and 63% were obtained for cases and controls, respectively. The study included 110 histologically confirmed cases of glioma, 60 meningioma cases and 417 controls. An increased risk of meningioma was associated with involuntary exposure to tobacco from the spouse, particularly among women (relative risk, 2.7; 95% CI, 1.2–6.1). No statistically significant association was found between active smoking and either glioma or meningioma in this study.

References

  • Akiba S., Kato H., Blot W.J. Passive smoking and lung cancer among Japanese women. Cancer Res. 1986;46:4804–4807. [PubMed: 3731126]
  • Alavanja M.C.R., Brown C.C., Swanson C., Brownson R.C. Saturated fat intake and lung cancer risk among non-smoking women in Missouri. J. nat. Cancer Inst. 1993;85:6–16. [PubMed: 8230280]
  • Bennett W.P., Alavanja M.C.R., Blomeke B., V#x00E4;häkangas K.H., Castren K., Welsh J.A., Bowman E.D., Khan M.A., Flieder D.B., Harris C.C. Environmental tobacco smoke, genetic susceptibility, and risk of lung cancer in never-smoking women. J. natl Cancer Inst. 1999;91:2009–2014. [PubMed: 10580025]
  • Berlin J.A., Longnecker M.P., Greenland S. Meta-analysis of epidemiologic doseresponse data. Epidemiology. 1993;4:218–228. [PubMed: 8512986]
  • Bero L.A., Glantz S.A., Rennie D. Publication bias and public health policy on environmental tobacco smoke. J. Am. med. Assoc. 1994;272:133–136. [PubMed: 8015124]
  • Biggerstaff B.J., Tweedie R.L., Mengersen K.L. Passive smoking in the workplace: Classical and Bayesian meta-analysis. Int. Arch. occup. environ. Health. 1994;66:269–277. [PubMed: 7843838]
  • Boffetta P. Involuntary smoking and lung cancer. Scand. J. Work Environ. Health. 2002;28 Suppl. 2:30–40. [PubMed: 12058801]
  • Boffetta P., Agudo A., Ahrens W., Benhamou E., Benhamou S., Darby S.C., Ferro G., Fortes C., Gonzales C.A., Jöckel K.-H., Krauss M., Kreienbrock L., Kreuzer M., Mendes A., Merletti F., Nyberg F., Pershagen G., Pohlabeln H., Riboli E., Schmid G., Simonato L., Tredaniel J., Whitley E., Wichmann H.-E., Winck C., Zambon P., Saracci R. Multicenter case–control study of exposure to environmental tobacco smoke and lung cancer. J. natl Cancer Inst. 1998;90:1440–1450. [PubMed: 9776409]
  • Boffetta P., Ahrens W., Nyberg F., Mukeria A., Brüske-Hohlfeld I., Fortes C., Constantinescu V., Simonato L., Batura-Gabryel H., Lea S., Gaborieau V., Benhamou S. Exposure to environmental tobacco smoke and risk of adenocarcinoma of the lung. Int. J. Cancer. 1999a;83:635–639. [PubMed: 10521800]
  • Boffetta P., Nyberg F., Agudo A., Benhamou E., Joeckel K.H., Kreuzer M., Merletti F., Pershagen G., Pohlabeln H., Simonato L., Wichmann H.E., Saracci R. Risk of lung cancer from exposure to environmental tobacco smoke from cigars, cigarillos and pipes. Int. J. Cancer. 1999b;83:805–806. [PubMed: 10597199]
  • Boffetta P., Tredaniel J., Greco A. Risk of childhood cancer and adult lung cancer after childhood exposure to passive smoke: Ameta-analysis. Environ. Health Perspect. 2000;108:73–82. [PMC free article: PMC1637845] [PubMed: 10620527]
  • Brennan P., Butler J., Agudo A., Benhamou S., Darby S., Fortes C., Joeckel K.H., Kreuzer M., Nyberg F., Pohlabeln H., Saracci R., Wichmann H.E., Boffetta P. Joint effect of diet and environmental tobacco smoke on risk of lung cancer among nonsmokers. J. natl Cancer Inst. 2000;92:426–427. [PubMed: 10699073]
  • Brondum J., Shu X.-O., Steinbuch M., Severson R.K., Potter J.D., Robison L.L. Parental cigarette smoking and the risk of acute leukemia in children. Cancer. 1999;85:1380–1388. [PubMed: 10189146]
  • Brown K.G. Lung cancer and environmental tobacco smoke: Occupational risk to nonsmokers. Environ. Health Perspect. 1999;107 Suppl. 6:885–890. [PMC free article: PMC1566198] [PubMed: 10592148]
  • Brownson R.C., Reif J.S., Keefe T.J., Ferguson S.W., Pritzl J.A. Risk factors for adenocarcinoma of the lung. Am. J. Epidemiol. 1987;125:25–34. [PubMed: 3788953]
  • Brownson R.C., Alavanja M.C., Hock E.T., Loy T.S. Passive smoking and lung cancer in nonsmoking women. Am. J. public Health. 1992;82:1525–1530. [PMC free article: PMC1694624] [PubMed: 1443304]
  • Buckley J.D., Hobbie W.L., Ruccione K., Sather H.N., Woods W.G., Hammond G.D. Maternal smoking during pregnancy and risk of childhood cancer [Letter]. Lancet. 1986;ii:519–520.
  • Buffler, P.A., Pickle, L.W., Mason, T.J. & Contant, C., (1984) The causes of lung cancer in Texas. In: Mizell, M. & Corres, P., eds, Lung Cancer: Causes and Prevention, New York, Verlag Chemie International, pp. 83–99.
  • Bunin G.R., Buckley J.D., Boesel C.P., Rorke L.B., Meadows A.T. Risk factors for astrocytic glioma and primitive neuroectodermal tumor of the brain in young children: A report from the Children's Cancer Group. Cancer Epidemiol. Biomarkers Prev. 1994;3:197–204. [PubMed: 8019366]
  • Butler, T.L. (1988) The Relationship of Passive Smoking to Various Health Outcomes Among Seventh-day Adventists in California, Los Angeles, University of California, Doctoral thesis.
  • Cardenas V.M., Thun M.J., Austin H., Lally C.A., Clark W.S., Greenberg S., Heath C.W. Environmental tobacco smoke and lung cancer mortality in the American Cancer Society's cancer prevention study II. Cancer Causes Control. 1997;8:57–64. [PubMed: 9051323]
  • Chan, W.C., & Fung, S.C. (1982) Lung cancer in non-smokers in Hong Kong. In: Grundmann, E., ed., Cancer Epidemiology (Cancer Campaign 6), Stuttgart, Gustav Fischer Verlag, pp. 199–202.
  • Chang-Claude J., Krupp S., Jager B., Bartsch H., Risch A. Differential effect of NAT2 on the association between active and passive smoke exposure and breast cancer risk. Cancer Epidemiol. Biomarkers Prev. 2002;8:698–704. [PubMed: 12163321]
  • Chappell W.R., Gratt L.B. A graphical method for pooling epidemiological studies. Am. J. public Health. 1996;88:1011–1012. [PMC free article: PMC1380494] [PubMed: 8629736]
  • Cheng Y.-J., Hildesheim A., Hsu M.-M., Chen I.-H., Brinton L.A., Levine P.H., Chen C.-J., Yang C.-S. Cigarette smoking, alcohol consumption and risk of nasopharyngeal carcinoma in Taiwan. Cancer Causes Control. 1999;10:201–207. [PubMed: 10454065]
  • Choi S.-Y., Lee K.-H., Lee T.-O. A case–control study on risk factors in lung cancer. Korean J. Epidemiol. 1989;11:66–80.
  • Coker A.L., Rosenberg A.J., McCann M.F., Hulka B.S. Active and passive cigarette smoke exposure and cervical intraepithelial neoplasia. Cancer Epidemiol. Biomarkers Prev. 1992;1:349–356. [PubMed: 1305466]
  • Copas J.B., Shi J.Q. Reanalysis of epidemiological evidence on lung cancer and passive smoking. Br. med. J. 2000;320:417–418. [PMC free article: PMC27286] [PubMed: 10669446]
  • Cordier S., Iglesias M.-J., LeGoaster C., Guyot M.-M., Mandereau L., Hemon D. Incidence and risk factors for childhood brain tumors in the Ile de France. Int. J. Cancer. 1994;59:776–782. [PubMed: 7989118]
  • Correa P., Pickle L.W., Fontham E., Lin Y., Haenszel W. Passive smoking and lung cancer. Lancet. 1983;ii:595–597. [PubMed: 6136747]
  • Dalager N.A., Pickle L.W., Mason T.J., Correa P., Fontham E.T.H., Stemhagen A., Buffler P.A., Ziegler R.G., Fraumeni J.F. Jr. The relation of passive smoking to lung cancer. Cancer Res. 1986;46:4808–4811. [PubMed: 3015399]
  • Delfino R.J., Smith C., West J.G., Lin H.J., White E., Liao S.-Y., Gim J.S.Y., Ma H.L., Butler J., Anton-Culver H. Breast cancer, passive and active cigarette smoking and N-acetyltransferase 2 genotype. Pharmacogenetics. 2000;10:461–469. [PubMed: 10898115]
  • Dersimonian R., Laird N. Meta-analysis in clinical trials. Controlled clin. Trials. 1986:177–188. [PubMed: 3802833]
  • Du Y.X., Cha Q., Chen Y.Z., Wu J.M. Exposure to environmental tobacco smoke and female lung cancer in Guangzhou, China. Proc. Indoor Air. 1993;1:511–516.
  • Egan K.M., Stampfer M.J., Hunter D., Hankinson S., Rosner B.A., Holmes M., Willett W.C., Colditz G.A. Active and passive smoking in breast cancer: Prospective results from the Nurses' Health Study. Epidemiology. 2002;13:138–145. [PubMed: 11880753]
  • Filippini G., Farinotti M., Lovicu G., Maisonneuve P., Boyle P. Mothers' active and passive smoking during pregnancy and risk of brain tumors in children. Int. J. Cancer. 1994;57:769–774. [PubMed: 8206670]
  • Filippini G., Farinotti M., Ferrarini M. Active and passive smoking during pregnancy and risk of central nervous system tumours in children. Paediatr. perinatal Epidemiol. 2000;14:78–84. [PubMed: 10703038]
  • Fleiss J.L., Gross A.J. 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;44:127–139. [PubMed: 1995774]
  • Fontham E.T.H., Correa P., Wu-Williams A.H., Reynolds P., Greenberg R.S., Buffler P.A., Chen V.W., Boyd P., Alterman T., Austin D.F., Liff J., Greenberg S.D. Lung cancer in nonsmoking women: A multicenter case–control study. Cancer Epidemiol. Biomarkers Prev. 1991;1:35–43. [PubMed: 1845167]
  • Fontham E.T.H., Correa P., Reynolds P., Wu-Williams A., Buffler P.A., Greenberg R.S., Chen V.W., Alterman T., Boyd P., Austin D.F., Liff J. Environmental tobacco smoke and lung cancer in nonsmoking women. A multicenter study. J. Am. med. Assoc. 1994;271:1752–1759. [PubMed: 8196118]
  • Forsberg J.-G., Kallen B. Pregnancy and delivery characteristics of women whose infants develop child cancer. APMIS. 1990;98:37–42. [PubMed: 2302339]
  • Fukuda K., Shibata A. Exposure–response relationships between woodworking, smoking or passive smoking and squamous cell neoplasms of the maxillary sinus. Cancer Causes Control. 1990;1:165–168. [PubMed: 2102286]
  • Gao Y.-T., Blot W.J., Zheng W., Ershow A.G., Hsu C.W., Levin L.I., Zhang R., Fraumeni J.F. Jr. Lung cancer among Chinese women. Int. J. Cancer. 1987;40:604–609. [PubMed: 2824385]
  • Garfinkel L. Time trends in lung cancer mortality among nonsmokers and a note on passive smoking. J. natl Cancer Inst. 1981;66:1061–1066. [PubMed: 6941041]
  • Garfinkel L., Auerbach O., Joubert L. Involuntary smoking and lung cancer: A case–control study. J. natl Cancer Inst. 1985;75:463–469. [PubMed: 3861899]
  • Geng, G.Y., Liang, Z.H., Zhang, A.Y. & Wu, G.L. (1988) On the relationship between smoking and female lung cancer. In: Aoki, M., Hisamichi, S. & Tominaga, S., eds, Smoking and Health, Amsterdam, Elsevier, pp. 483–486.
  • Gerhardsson de Verdier M., Plato N., Steineck G., Peters J.M. Occupational exposures and cancer of the colon and rectum. Am. J. ind. Med. 1992;22:291–303. [PubMed: 1519614]
  • Gold E., Gordis L., Tonascia J., Szklo M. Risk factors for brain tumors in children. Am. J. Epidemiol. 1979;109:309–319. [PubMed: 453168]
  • Gold E.B., Leviton A., Lopez R., Gilles F.H., Hedley-Whyte E.T., Kolonel L.N., Lyon J.L., Swanson G.M., Weiss N.S., West D., Aschenbrener C., Austin D.F. Parental smoking and risk of childhood brain tumors. Am. J. Epidemiol. 1993;137:620–628. [PubMed: 8470663]
  • Golding J., Paterson M., Kinlen L.J. Factors associated with childhood cancer in a national cohort study. Br. J. Cancer. 1990;62:3034–3038. [PMC free article: PMC1971807] [PubMed: 2386748]
  • Golding J., Greenwood R., Birmingham K., Mott M. Childhood cancer, intramuscular vitamin K, and pethidine given during labor. Br. med. J. 1992;305:341–346. [PMC free article: PMC1883000] [PubMed: 1392886]
  • Hackshaw A.K. Lung cancer and passive smoking. Stat. Meth. med. Res. 1998;7:119–136. [PubMed: 9654638]
  • Hackshaw A.K., Law M.R., Wald N.J. The accumulated evidence on lung cancer and environmental tobacco smoke. Br. med. J. 1997;315:980–988. [PMC free article: PMC2127653] [PubMed: 9365295]
  • Hackshaw A.K., Law M.R., Wald N.J. Increased risk is not disputed. Br. med. J. 2000;321:1221–1222. [PubMed: 11073522]
  • Heller W.-D., Scherer G., Sennewald E., Adlkofer F. Misclassification of smoking in a follow-up population study in southern Germany. Clin. Epidemiol. 1998;51:211–218. [PubMed: 9495686]
  • Hirayama T. Nonsmoking wives of heavy smokers have a higher risk of lung cancer: A study from Japan. Br. med. J. 1981;282:183–185. [PMC free article: PMC1503989] [PubMed: 6779940]
  • Hirayama T. Cancer mortality in nonsmoking women with smoking husbands based on a large-scale cohort study in Japan. Prev. Med. 1984;13:680–690. [PubMed: 6536942]
  • Hirayama, T. (1989) Dietary habits are of limited importance in influencing the lung cancer risk among Japanese females who never smoked. In: Bieva, C.J., Courtois, Y. & Govaerts, M., eds, Present and Future Indoor Air Quality. Proceedings of the Brussels Conference, Amsterdam, Elsevier Science.
  • Hole D.J., Gills C.R., Chopra C., Hawthorne V.M. Passive smoking and cardiorespiratory health in a general population in the west of Scotland. Br. med. J. 1989;299:423–427. [PMC free article: PMC1837285] [PubMed: 2507000]
  • Holly E.A., Aston D.A., Ahn D.K., Kristiansen J.J. Ewing's bone sarcoma, paternal occupational exposure, and other factors. Am. J. Epidemiol. 1992;135:122–129. [PubMed: 1311140]
  • Howe G.R., Burch J.D., Chiarelli A.M., Risch H., Choi B.C.K. An exploratory case–control study of brain tumors in children. Cancer Res. 1989;49:4349–4352. [PubMed: 2743324]
  • Humble C.G., Samet J.M., Pathak D.R. Marriage to a smoker and lung cancer risk. Am. J. public Health. 1987;77:598–602. [PMC free article: PMC1647031] [PubMed: 3565655]
  • IARC (1986) IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Vol. 38, Tobacco Smoking, Lyon, IARCPress. [PubMed: 3460963]
  • Infante-Rivard C., Krajinovic M., Labuda D., Sinnett D. Parental smoking, CYP1A1genetic polymorphisms and childhood leukemia (Quebec, Canada). Cancer Causes Control. 2000;11:547–553. [PubMed: 10880037]
  • Inoue, R., & Hirayama, T. (1988) Passive smoking and lung cancer in women. In: Aoki, M., Hisamichi, S. & Tominaga, S., eds, Smoking and Health, Amsterdam, Elsevier, pp. 283–285.
  • Janerich D.T., Thompson W.D., Varela L.R., Greenwald P., Chorost S., Tucci C., Zaman M.B., Melamed M., Kiely M., McKneally M.F. Lung cancer and exposure to tobacco smoke in the household. New Engl. J. Med. 1990;323:632–636. [PubMed: 2385268]
  • Jee S.H., Ohrr H., Kim I.-S. Effects of husbands' smoking on the incidence of lung cancer in Korean women. Int. J. Epidemiol. 1999;28:824–828. [PubMed: 10597977]
  • Ji B.T., Shu X.O., Linet M.S., Zheng W., Wacholder S., Gao Y.T., Ying D.M., Jin F. Paternal cigarette smoking and the risk of childhood cancer among offspring of nonsmoking mothers. J. natl Cancer Inst. 1997;89:238–244. [PubMed: 9017004]
  • Jöckel, K.-H., Krauss, M., Pohlabeln, H., Ahrens, W., Kreuzer, M., Kreienbrock, L., Möhner, M. & Wichmann, H.E. (1998a) Lungenkrebsrisiko durch berufliche Exposition –Passivrauchen. In: Wichmann, H.E., Jöckel, K.-H. & Robra, B.P., eds, Fortschritte in der Epidemiologie, Lungenkrebsrisiko durch berufliche Exposition, Landsberg, Ecomed, pp. 227–23.
  • Jöckel K.-H., Pohlabeln H., Ahrens W., Krauss M. Environmental tobacco smoke and lung cancer. Epidemiology. 1998b;9:672–675. [PubMed: 9799181]
  • John E., Savitz D.A., Sandler D.P. Prenatal exposure to parents smoking and childhood cancer. Am. J. Epidemiol. 1991;133:123–132. [PubMed: 1822074]
  • Johnson K.C., Hu J., Mao Y. The Canadian Cancer Registries Epidemiology Research Group., author Passive and active smoking and breast cancer risk in Canada, 1994–97. Cancer Causes Control. 2000;11:211–221. [PubMed: 10782655]
  • Johnson K.C., Hu J., Mao Y. Canadian Cancer Registries Epidemiology Research Group, author. Lifetime residential and workplace exposure to environmental tobacco smoke and lung cancer in never-smoking women, Canada 1994–1997. Int. J. Cancer. 2001;93:902–906. [PubMed: 11519055]
  • Kabat, G.C. (1990) Epidemiologic studies of the relationship between passive smoking and lung cancer. In: Proceedings of the 1990 Annual Winter Meeting of the Toxicology Forum, Washington DC.
  • Kabat G.C., Wynder E.L. Lung cancer in nonsmokers. Cancer. 1984;53:1214–1221. [PubMed: 6692309]
  • Kabat G.C., Stellman S.D., Wynder E.L. Relation between exposure to environmental tobacco smoke and lung cancer in lifetime nonsmokers. Am. J. Epidemiol. 1995;142:141–148. [PubMed: 7598113]
  • Kalandidi A., Trichopoulos D., Hatzakis A., Tzannes S., Saracci R. Passive smoking and chronic obstructive lung disease. Lancet. 1987;ii:1325–1326. [PubMed: 2890917]
  • Kalandidi A., Katsouyanni K., Voropoulou N., Bastas G., Saracci R., Trichopoulos D. Passive smoking and diet in the etiology of lung cancer among non-smokers. Cancer Causes Control. 1990;1:15–21. [PubMed: 1966316]
  • Katada H., Mikami R., Konishi M., Koyama Y., Narita N. Effect of passive smoking in lung cancer development in women in the Nara region. Gan No Rinsho. 1988;34:21–27. [PubMed: 2828712]
  • Keil U., Stieber J., Filipiak B., Löwel H., Wichmann H.E. On misclassification of smoking (Letter to the Editor). J. clin. Epidemiol. 1998;52:91–93. [PubMed: 9973078]
  • Klebanoff M.A., Clemens J.D., Read J.S. Maternal smoking during pregnancy and childhood cancer. Am. J. Epidemiol. 1996;144:1028–1033. [PubMed: 8942433]
  • Ko Y.-C., Lee C.-H., Chen M.-J., Huang C.-C., Chang W.-Y., Lin H.-J., Wang H.-Z., Chang P.-Y. Risk factors for primary lung cancer among non-smoking women in Taiwan. Int. J. Epidemiol. 1997;26:24–31. [PubMed: 9126500]
  • Koo L.C., Ho J.H.-C., Saw D. Active and passive smoking among female lung cancer patients in Hong Kong. J. exp. clin. Cancer Res. 1983;4:367–375.
  • Koo L.C., Ho J.H.-C., Saw D. Is passive smoking an added risk factor for lung cancer in Chinese women? J. exp. clin. Cancer Res. 1984;3:277–283.
  • Koo L.C., Ho J.H., Saw D., Ho C.Y. Measurements of passive smoking and estimates of lung cancer risk among non-smoking Chinese females. Int. J. Cancer. 1987;39:162–169. [PubMed: 3804491]
  • Koo L.C., Ho J.H.-C., Rylander R. Life-history correlates of environmental tobacco smoke: A study on nonsmoking Hong Kong Chinese wives with smoking versus nonsmoking husbands. Soc. Sci. Med. 1988;26:751–760. [PubMed: 3358146]
  • Kramer S., Ward E., Meadows A.T., Malone K.E. Medical and drug risk factors associated with neuroblastoma: A case–control study. J. natl Cancer Inst. 1987;78:797–804. [PubMed: 3471992]
  • Kreuzer M., Krauss M., Kreienbrock L., Jockel K.-H., Wichmann H.-E. Environmental tobacco smoke and lung cancer: A case–control study in Germany. Am. J. Epidemiol. 2000;151:241–250. [PubMed: 10670548]
  • Kreuzer M., Gerken M., Kreienbrock L., Wellmann J., Wichmann H.E. Lung cancer in lifetime nonsmoking men –results of a case–control study in Germany. Br. J. Cancer. 2001;84:134–140. [PMC free article: PMC2363603] [PubMed: 11139328]
  • Kreuzer M., Heinrich J., Kreienbrock L., Rosario A.S., Gerken M., Wichmann H.E. Risk factors for lung cancer among nonsmoking women. Int. J. Cancer. 2002;100:706–713. [PubMed: 12209611]
  • Kropp S., Chang-Claude J. Active and passive smoking and risk of breast cancer by age 50 years among German women. Am. J. Epidemiol. 2002;156:616–626. [PubMed: 12244030]
  • Kuijten R.R., Bunin G.R., Nass C.C., Meadows A.T. Gestational and familial risk factors for childhood astrocytoma: Results of a case–control study. Cancer Res. 1990;50:2608–2612. [PubMed: 2328486]
  • Lam, W.K. (1985) A Clinical and Epidemiological Study of Carcinoma of the Lung in Hong Kong, Hong Kong, University of Hong Kong, Doctoral thesis.
  • Lam T.H., Kung I.T., Wong C.M., Lam W.K., Kleevens J.W., Saw D., Hsu C., Seneviratne S., Lam S.Y., Lo K.K. Smoking, passive smoking and histological types in lung cancer in Hong Kong Chinese women. Br. J. Cancer. 1987;56:673–678. [PMC free article: PMC2001879] [PubMed: 3426935]
  • Lash T.L., Aschengrau A. Active and passive cigarette smoking and the occurrence of breast cancer. Am. J. Epidemiol. 1999;149:5–12. [PubMed: 9883788]
  • Lee P.N. Lung cancer and passive smoking: Association of an artifact due to misclassification of smoking habits? Toxicol. Lett. 1987a;35:157–162. [PubMed: 3810676]
  • Lee P.N. Passive smoking and lung cancer association: A result of bias? Hum. Toxicol. 1987b;6:517–524. [PubMed: 3692498]
  • Lee, P.N. (1988) Misclassification of Smoking Habits and Passive Smoking, Berlin, Springer Verlag.
  • Lee, P.N. (1991a) Correcting Meta-analyses of the Association of Lung Cancer in Females with Spouse (or Household) Exposure for Bias due to Misclassification of Active Smoking Status (Report submitted to US Environmental Protection Agency, dated 29 November 1991)
  • Lee P.N. An estimate of adult mortality in the United States from passive smoking. Environ. int. 1991b;17:379–381.
  • Lee, P.N. (1992) Environmental Tobacco Smoke and Mortality, Basel, Karger.
  • Lee P.N. Difficulties in assessing the relationship between passive smoking and lung cancer. Stat. Meth. med. Res. 1998;7:137–163. [PubMed: 9654639]
  • Lee P.N., Chamberlain J., Alderson M.R. Relationship of passive smoking to risk of lung cancer and other smoking-associated diseases. Br. J. Cancer. 1986;54:97–105. [PMC free article: PMC2001639] [PubMed: 3730259]
  • Lee C.H., Ko Y.C., Goggins W., Huang J.J., Huang M.S., Kao E.L., Wang H.Z. Lifetime environmental exposure to tobacco smoke and primary lung cancer of non-smoking Taiwanese women. Int. J. Epidemiol. 2000;29:224–231. [PubMed: 10817117]
  • Lee P.N., Forey B., Fry J.S. Revisiting the association between environmental tobacco smoke and exposure to lung cancer risk. III. Adjustment for the biasing effect of misclassification of smoking habits. Indoor Built Environ. 2001;10:384–398.
  • Levois M.E., Layard M.W. Inconsistency between workplace and spousal studies of environmental tobacco smoke and lung cancer. Regul. Toxic. Pharmacol. 1994;19:309–316. [PubMed: 8090954]
  • Liu Z., He X., Chapman R.S. Smoking and other risk factors for lung cancer in Xuanwei, China. Int. J. Epidemiol. 1991;20:26–31. [PubMed: 2066232]
  • Liu Q., Sasco A.J., Riboli E., Hu M.X. Indoor air pollution and lung cancer in Ghuangzhou, People's Republic of China. Am. J. Epidemiol. 1993;137:145–154. [PubMed: 8452118]
  • Lubin J.H. Estimating lung cancer risk with exposure to environmental tobacco smoke. Environ. Health Perspect. 1999;107 Suppl. 6:879–883. [PMC free article: PMC1566203] [PubMed: 10592146]
  • McCredie M., Maisonneuve P., Boyle P. Perinatal and early postnatal risk factors for malignant brain tumours in New South Wales children. Int. J. Cancer. 1994;56:11–15. [PubMed: 8262665]
  • McKinney P.A., Stiller C.A. (for the IRESCC group), author. Maternal smoking during pregnancy and the risk of childhood cancer [Letter]. Lancet. 1986;ii:519. [PubMed: 2875269]
  • Magnani C., Pastore G., Luzzatto L., Carli M., Lubrano P., Terracini B. Risk factors for soft tissue sarcomas in childhood: A case–control study. Tumori. 1989;75:396–400. [PubMed: 2815346]
  • Magnani C., Pastore G., Luzzatto L., Terracini B. Parental occupation and other environmental factors in the etiology of leukemias and non-Hodgkin's lymphomas in childhood: A case–control study. Tumori. 1990;76:413–419. [PubMed: 2256184]
  • Manning M.D., Carroll B.E. Some epidemiological aspects of leukemia in children. J. natl Cancer Inst. 1957;19:1087–1094. [PubMed: 13502763]
  • Mao Y., Hu J., Semenciw R., White K. the Canadian Cancer Registries Epidemiology Research Group, author. Active and passive smoking and the risk of stomach cancer, by subsite, in Canada. Eur. J. Cancer Prev. 2002;11:27–38. [PubMed: 11917206]
  • Marcus P.M., Newman B., Millikan R.C., Moorman P.G., Baird D.D., Qaqish B. The associations of adolescent cigarette smoking, alcoholic beverage consumption, environmental tobacco smoke, and ionizing radiation with subsequent breast cancer risk (United States). Cancer Causes Control. 2000;11:271–278. [PubMed: 10782661]
  • Mayne S.T., Janerich D.T., Greenwald P., Chorost S., Tucci C., Zaman M.B., Melamed M.R., Kiely M., McKneally M.F. Dietary beta carotene and lung cancer risk in US nonsmokers. J. natl Cancer Inst. 1994;86:33–38. [PubMed: 8271280]
  • Miller G.H. The impact of passive smoking: Cancer deaths among nonsmoking women. Cancer Detect. Prev. 1990;14:497–503. [PubMed: 2224913]
  • Millikan R.C., Pittman G.S., Newman B., Tse C.-K.J., Selmin O., Rockhill B., Savitz D., Moorman P.G., Bell D.A. Cigarette smoking, N-acetyltransferases 1 and 2, and breast cancer risk. Cancer Epidemiol. Biomarkers Prev. 1998;7:371–378. [PubMed: 9610785]
  • Morabia A., Bernstein M., Heritier S., Khatchatrian N. Relation of breast cancer with passive and active exposure to tobacco smoke. Am. J. Epidemiol. 1996;143:918–928. [PubMed: 8610705]
  • Morabia A., Bernstein M.S., Bouchardy I., Kurtz J., Morris M.A. Breast cancer and active and passive smoking: The role of the N-acetyltransferase 2 genotype. Am. J. Epidemiol. 2000;152:226–232. [PubMed: 10933269]
  • National Research Council (1986) Environmental Tobacco Smoke: Measuring Exposures and Assessing Health Effects, Washington DC, National Academy Press. [PubMed: 25032469]
  • Neutel C.I., Buck C. Effect of smoking during pregnancy on the risk of cancer in children. J. natl Cancer Inst. 1971;47:59–63. [PubMed: 5570121]
  • Nishino Y., Tsubono Y., Tsuji I., Komatsu S., Kanemura S., Nakatsuka H., Fukao A., Satoh H., Hisamichi S. Passive smoking at home and cancer risk: A population-based prospective study in Japanese nonsmoking women. Cancer Causes Control. 2001;12:797–802. [PubMed: 11714107]
  • Norman M.A., Holly E.A., Ahn D.K., Preston-Martin S., Mueller B.A., Bracci P.M. Prenatal exposure to tobacco smoke and childhood brain tumors: Results from the United States West Coast Childhood Brain Tumor Study. Cancer Epidemiol. Biomarkers Prev. 1996;5:127–133. [PubMed: 8850274]
  • Nyberg F., Isaksson I., Harris J.R., Pershagen G. Misclassification of smoking status and lung cancer risk from environmental tobacco smoke in never-smokers. Epidemiology. 1997;8:304–309. [PubMed: 9115027]
  • Nyberg F., Agrenius V., Svartengren K., Svensson C., Pershagen G. Environmental tobacco smoke and lung cancer in nonsmokers: Does time since exposure play a role? Epidemiology. 1998;9:301–308. [PubMed: 9583423]
  • Olshan A.F., Breslow N.E., Falletta J.M., Grufferman S., Pendergrass T., Robison L.L., Waskerwitz M., Woods W.G., Vietti T.J., Hammond G.D. Risk factors for Wilms tumor: Report from the National Wilms Tumor Study. Cancer. 1993;72:938–944. [PubMed: 8392906]
  • Pershagen G., Hrubec Z., Svensson C. Passive smoking and lung cancer in Swedish women. Am. J. Epidemiol. 1987;125:17–24. [PubMed: 3788950]
  • Pershagen G., Ericson A., Otterblad-Olausson P. Maternal smoking in pregnancy: Does it increase the risk of childhood cancer? Int. J. Epidemiol. 1992;21:1–5. [PubMed: 1544740]
  • Peto, J. (1992) Meta-analysis of epidemiological studies of carcinogenesis. In: Vainio, H., Magee, P.N., McGregor, D.B. & McMichael, A.J., eds, Mechanisms of Carcinogenesis in Risk Identification (IARC Scientific Publications No. 116), Lyon, IARCPress, pp. 571–577.
  • Preston-Martin S., Yu M.C., Benton B., Henderson B.E. N-nitroso compounds and childhood brain tumors: A case–control study. Cancer Res. 1982;42:5240–5245. [PubMed: 7139628]
  • Rapiti E., Jindal S.K., Gupta D., Boffetta P. Passive smoking and lung cancer in Chandigarh, India. Lung Cancer. 1999;23:183–189. [PubMed: 10413195]
  • Reynolds P., Von Behren J., Fontham E.T.H., Wu A., Buffler P., Greenberg R. Occupational exposure to environmental tobacco smoke [Letter]. J. Am. med. Assoc. 1996;275:441–442. [PubMed: 8627956]
  • Riboli E., Haley N.J., Tredaniel J., Saracci R., Preston-Martin S., Trichopolous D. Misclassification of smoking status among women in relation to exposure to environmental tobacco smoke. Eur. respir. J. 1995;8:285–290. [PubMed: 7758565]
  • Roman E., Watson A., Beral V., Buckle S., Bull D., Baker K., Ryder H., Barton C. Case–control study of leukaemia and non-Hodgkin's lymphoma among children aged 0–4 years living in West Berkshire and North Hampshire health districts. Br. med. J. 1993;306:615–621. [PMC free article: PMC1676930] [PubMed: 8461811]
  • Ryan P., Lee M.W., North J.B., McMichael A.J. Risk factors for tumors of the brain and meninges: Results from the Adelaide Adult Brain Tumor study. Int. J. Cancer. 1992;51:20–27. [PubMed: 1563840]
  • Sandler D.P., Everson R.B., Wilcox A.J., Browder J.P. Cancer risk in adulthood from early life exposure to parents' smoking. Am. J. public Health. 1985a;75:487–492. [PMC free article: PMC1646286] [PubMed: 3985235]
  • Sandler D.P., Everson R.B., Wilcox A.J. Passive smoking in adulthood and cancer risk. Am. J. Epidemiol. 1985b;121:37–48. [PubMed: 3964991]
  • Sandler R.S., Sandler D.P., Comstock G.W., Helsing K.J., Shore D.L. Cigarette smoking and the risk of colorectal cancer in women. J. natl Cancer Inst. 1988;80:1329–1333. [PubMed: 3172257]
  • Scholes D., McBride C., Grothaus L., Curry S., Albright J., Ludman E. The association between cigarette smoking and low-grade cervical abnormalities in reproductive-age women. Cancer Causes Control. 1999;10:339–344. [PubMed: 10530602]
  • Schwartz A.G., Yang P., Swanson G.M. Familial risk of lung cancer among nonsmokers and their relatives. Am. J. Epidemiol. 1996;144:554–562. [PubMed: 8797515]
  • Schwartzbaum J.A. Influence of the mother's prenatal drug consumption on risk of neuroblastoma in the child. Am. J. Epidemiol. 1992;135:1358–1367. [PubMed: 1510082]
  • Severson R.K., Buckley J.D., Woods W.G., Benjamin D., Robison L.L. Cigarette smoking and alcohol consumption by parents of children with acute myeloid leukemia: An analysis within morphological subgroups—a report from the Childrens Cancer Group. Cancer Epidemiol. Biomarkers Prev. 1993;2:433–439. [PubMed: 8220087]
  • Shen X.B., Wang G.X., Zhou B.S. Relation of environmental tobacco smoke and pulmonary adenocarcinoma in non-smoking women: A case–control study in Nanjing. Oncol. Rep. 1998;5:1221–1223. [PubMed: 9683839]
  • Shimizu H., Morishita M., Mizuno K., Masuda T., Ogura Y., Santo M., Nishimura M., Kunishima K., Karasawa K., Nishiwaki K., Yamamoto M., Hisamichi S., Tominaga S. A case–control study of lung cancer in non-smoking women. Tohoku J. exp. Med. 1988;154:389–397. [PubMed: 3188004]
  • Shu X.-O., Ross J.A., Pendergrass T.W., Reaman G.H., Lampkin B., Robison L.L. Parental alcohol consumption, cigarette smoking, and risk of infant leukemia: A Childrens Cancer Group study. J. natl Cancer Inst. 1996;88:24–31. [PubMed: 8847721]
  • Slattery M.L., Robison L.M., Schuman K.L., French T.K., Abbott T.M., Overall J.C. Jr, Gardner J.W. Cigarette smoking and exposure to passive smoke are risk factors for cervical cancer. J. Am. med. Assoc. 1989;262:1593–1598. [PubMed: 2918652]
  • Smith S.J., Deacon J.M., Chilvers C.E.D. members of the UK National Case–Control Study Group., author Alcohol, smoking, passive smoking and caffeine in relation to breast cancer risk in young women. Br. J. Cancer. 1994;70:113–119. [PMC free article: PMC2033307] [PubMed: 8018520]
  • Sobue T. Association of indoor air pollution and lifestyle with lung cancer in Osaka, Japan. Int. J. Epidemiol. 1990;19 Suppl. 1:S62–S66. [PubMed: 2258278]
  • Sorahan T., Lancashire R., Prior P., Peck I., Stewart A. Childhood cancer and prenatal use of alcohol and tobacco. Ann. Epidemiol. 1995;5:354–359. [PubMed: 8653207]
  • Sorahan T., Lancashire R.J., Hulten M.A., Peck I., Stewart A.M. Childhood cancer and parental use of tobacco: Deaths from 1953 to 1955. Br. J. Cancer. 1997a;75:134–138. [PMC free article: PMC2222709] [PubMed: 9000611]
  • Sorahan T., Prior P., Lancashire R.J., Faux S.P., Hulten M.A., Peck I.M., Stewart A.M. Childhood cancer and parental use of tobacco: Deaths from 1971 to 1976. Br. J. Cancer. 1997b;76:1525–1531. [PMC free article: PMC2228172] [PubMed: 9400953]
  • Sorahan T., McKinney P.A., Mann J.R., Lancashire R.J., Stiller C.A., Birch J.M., Dodd H.E., Cartwright R.A. Childhood cancer and parental use of tobacco: Findings from the inter-regional epidemiological study of childhood cancer (IRESCC). Br. J. Cancer. 2001;84:141–148. [PMC free article: PMC2363626] [PubMed: 11139329]
  • Speizer F.E., Colditz G.A., Hunter D.J., Rosner B., Hennekens C. Prospective study of smoking, antioxidant intake, and lung cancer in middle-aged women (USA). Cancer Causes Control. 1999;10:475–482. [PubMed: 10530619]
  • Stewart A., Webb J., Hewitt D. A survey of childhood malignancies. Br. med. J. 1958;5086:1495–1508. [PMC free article: PMC2029590] [PubMed: 13546604]
  • Stjernfeldt M., Berglund K., Lindsten J., Ludvigsson J. Maternal smoking during pregnancy and risk of childhood cancer. Lancet. 1986;i:1350–1352. [PubMed: 2872471]
  • Stockwell H.G., Goldman A.L., Lyman G.H., Noss C.I., Armstrong A.W., Pinkham P.A., Candelora E.C., Brusa M.R. Environmental tobacco smoke and lung cancer risk in nonsmoking women. J. natl Cancer Inst. 1992;84:1417–1422. [PubMed: 1512793]
  • Sun X.-W., Dai X.-D., Lin C.-Y., Shi Y.-B., Ma Y.-Y., Li W. Passive smoking and lung cancer among nonsmoking women in Harbin, China. International symposium on lifestyle factors and human lung cancer, China 1994. Lung Cancer. 1996;14:237.
  • Svensson, C., Pershagen, G. & Klominek, J. (1988) Smoking and Passive Smoking in Relation to Lung Cancer in Women, Stockholm, National Institute of Environmental Medicine.
  • Svensson C., Pershagen G., Klominek J. Smoking and passive smoking in relation to lung cancer in women. Acta oncol. 1989;28:623–629. [PubMed: 2590538]
  • Trichopoulos D. Passive smoking and lung cancer. Scand. J. soc. Med. 1988;16:75–79. [PubMed: 3291102]
  • Trichopoulos D., Kalandidi A., Sparros L., MacMahon B. Lung cancer and passive smoking. Int. J. Cancer. 1981;27:1–4. [PubMed: 7251227]
  • Trichopoulos D., Kalandidi A., Sparros L. Lung cancer and passive smoking: Conclusion of Greek study. Lancet. 1983;ii:677–678. [PubMed: 6136810]
  • Tweedie R.L., Mengersen K.L. Lung cancer and passive smoking: Reconciling the biochemical and epidemiological approaches. Br. J. Cancer. 1992;66:700–705. [PMC free article: PMC1977429] [PubMed: 1419610]
  • Tweedie R.L., Scott D.J., Biggerstaff B.J., Mengersen K.L. Bayesian meta-analysis, with application to studies of ETS and lung cancer. Lung Cancer. 1996;14 Suppl. 1:S171–S194. [PubMed: 8785662]
  • US Environmental Protection Agency (1992) Respiratory Health Effects of Passive Smoking: Lung Cancer and Other Disorders (No. EPA/600/6-90/006F), Washington DC, Office of Research and Development, Office of Health and Environmental Assessment.
  • Van Steensel-Moll H.A., Valkenburg H.A., Vandenbroucke J.P., Van Zanen G.E. Are maternal fertility problems related to childhood leukaemia? Int. J. Epidemiol. 1985;14:555–559. [PubMed: 3866751]
  • de Waard F., Kemmeren J.M., van Ginkel L.A., Stolker A.A. Urinary cotinine and lung cancer risk in a female cohort. Br. J. Cancer. 1995;72:784–787. [PMC free article: PMC2033890] [PubMed: 7669595]
  • Wald N.J., Nanchahal K., Thompson S.G., Cuckle H.S. Does breathing other people's tobacco smoke cause lung cancer? Br. med. J. 1986;293:1217–1222. [PMC free article: PMC1341990] [PubMed: 3096439]
  • Wang L., Lubin J.H., Zhang S.R., Metayer C., Xia Y., Brenner A., Shang B., Wang Z., Kleinerman R.A. Lung cancer and environmental tobacco smoke in a non-industrial area of China. Int. J. Cancer. 2000;88:139–145. [PubMed: 10962452]
  • Wang T.-J., Zhou B.-S., Shi J.-P. Lung cancer in nonsmoking Chinese women: A case-control study. International symposium on lifestyle factors and human lung cancer, China 1994. Lung Cancer. 1996;14:93–98.
  • Wartenberg D., Calle E.E., Thun M.J., Heath C.W. Jr, Lally C., Woodruff T. Passive smoking exposure and female breast cancer. J. natl Cancer Inst. 2000;92:1666–1673. [PubMed: 11036112]
  • Wells A.J. Breast cancer, cigarette smoking, and passive smoking [Letter]. Am. J. Epidemiol. 1991;133:208–210. [PubMed: 1985448]
  • Wells A.J. Lung cancer from passive smoking at work. Am. J. public Health. 1998;88:1025–1029. [PMC free article: PMC1508269] [PubMed: 9663148]
  • Wells A.J., English P.B., Posner S.F., Wagenknecht L.E., Perez-Stable E.J. Misclassification rates for current smokers misclassified as nonsmokers. Am. J. public Health. 1998;88:1053–1059. [PMC free article: PMC1508460] [PubMed: 9772852]
  • Whitehead J., Whitehead A. A general parametric approach to the meta-analysis of randomised clinical trials. Stat. Med. 1991;10:1665–1677. [PubMed: 1792461]
  • Winn D.M., Li F.P., Robison L.L., Mulvihill J.J., Daigle A.E., Fraumeni J.F. Jr. A case–control study of the etiology of Ewing's sarcoma. Cancer Epidemiol. Biomarkers Prev. 1992;1:525–532. [PubMed: 1302564]
  • Wu A.H., Henderson B.E., Pike M.C., Yu M.C. Smoking and other risk factors for lung cancer in women. J. natl Cancer Inst. 1985;74:747–751. [PubMed: 3857370]
  • Wu-Williams A.H., Dai X.D., Blot W., Xu Z.Y., Sun X.W., Xiao H.P., Stone B.J., Yu S.F., Feng Y.P., Ershow A.G., Sun J., Fraumeni J.F. Jr, Henderson B.E. Lung cancer among women in north-east China. Br. J. Cancer. 1990;62:982–987. [PMC free article: PMC1971561] [PubMed: 2257230]
  • Yuan J.-M., Wang X.-L., Xiang Y.-B., Gao Y.-T., Ross R.K., Yu M.C. Non-dietary risk factors for nasopharyngeal carcinoma in Shanghai, China. Int. J. Cancer. 2000;85:364–369. [PubMed: 10652428]
  • Zaridze D.G., Zemlianaia G.M., Aitakov Z.N. [Contribution of out- and indoor air pollution to the etiology of lung cancer.] Vestn. Ross. Akad. Med. Nauk. 1995;4:6–10. (in Russian) [PubMed: 7780346]
  • Zaridze D., Maximovitch D., Zemlyana G., Aitakov Z.N., Boffetta P. Exposure to environmental tobacco smoke and risk of lung cancer in non-smoking women from Moscow, Russia. Int. J. Cancer. 1998;75:335–338. [PubMed: 9455789]
  • Zheng W., McLaughlin J.K., Chow W.-H., Co Chien H.T., Blot W.J. Risk factors for cancers of the nasal cavity and paranasal sinuses among white men in the United States. Am. J. Epidemiol. 1993;138:965–972. [PubMed: 8256781]
  • Zhong L., Goldberg M.S., Gao Y.-T., Jin F. A case–control study of lung cancer and environmental tobacco smoke among non-smoking women living in Shanghai, China. Cancer Causes Control. 1999;10:607–616. [PubMed: 10616829]
  • Zhong L., Goldberg M.S., Parent M.-E., Hanley J.A. Exposure to environmental tobacco smoke and the risk of lung cancer: A meta-analysis. Lung Cancer. 2000;27:3–18. [PubMed: 10672779]
©International Agency for Research on Cancer, 2004.
Bookshelf ID: NBK316397

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (16M)

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...