You are here

Article Text

Article menu

Download PDFPDF

Reviews
Environmental tobacco smoke and the risk of pancreatic cancer among non-smokers: a meta-analysis
  1. Jiachen Zhou1,
  2. Gregory A Wellenius1,
  3. Dominique S Michaud1,2
  1. 1Department of Epidemiology, Brown Public Health Program, Brown University, Providence, Rhode Island, USA
  2. 2Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, UK
  1. Correspondence to Dr Dominique S Michaud, Department of Epidemiology, Brown Public Health Program, Brown University, Box G-S121-2, Providence RI 02912, USA; Dominique_Michaud{at}brown.edu

Abstract

Background Experimental studies have linked exposure to tobacco-specific nitrosamines with pancreatic carcinogenesis. A number of epidemiological studies have examined the association between environmental tobacco smoke (ETS) and risk of pancreatic cancer but they have not yet been jointly summarised.

Objectives To investigate the association between exposure to ETS and risk of pancreatic cancer by systematically reviewing and synthesising the available evidence.

Methods We conducted a comprehensive literature search using MEDLINE and EMBASE and manual searching of the reference lists of the relevant research reports and review articles to identify full texts and abstracts published through October 2011. We used the random-effects model to pool summary relative risks (RR) comparing the highest category of exposure to ETS to people who had never been exposed.

Results Exposure to ETS during childhood was not associated with risk of pancreatic cancer (three prospective and two retrospective studies; summary RR 1.12;  95% CI 0.89 to 1.43). In addition, no association was found with exposure to ETS during adulthood at home (five prospective and three retrospective studies; summary RR 1.23; 95% CI 0.86 to 1.77) or at work (one prospective and two retrospective studies; summary RR 0.94; 95% CI 0.67 to 1.33).

Conclusions This meta-analysis does not provide evidence for an association between exposure to ETS and risk of pancreatic cancer.

https://doi.org/10.1136/oemed-2012-100844

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Pancreatic cancer is the fourth leading cause of cancer-related deaths for men and women in the USA.1 The annual age-adjusted incidence rate of pancreatic cancer was 13.3 per 100 000 for men and 10.2 per 100 000 for women from 2003 to 2007, while the annual age-adjusted death rate was 12.3 per 100 000 for men and 9.4 per 100 000 for women during the same time period.1 The 5-year survival rate of pancreatic cancer is only about 5.6%, the lowest rate among all cancer sites.1 ,2 Lack of effective therapies to treat this cancer underscores the importance of identifying high-risk populations and implementing primary prevention programmes. Tobacco smoking is the most common known risk factor, which contributes to about 20–25% of all pancreatic cancer cases.2–6 A recent meta-analysis showed that active tobacco smoking is associated with a 75% increase in the risk of pancreatic cancer compared to never smokers, and the elevated risk persists for at least 10 years after smoking cessation.7 Increasing prevalence of smoking in developing countries and increasing population longevity may elevate the worldwide burden of pancreatic cancer in the near future.2 Other important risk factors for pancreatic cancer include diabetes, chronic pancreatitis, obesity, genetic predisposition and non-O blood types.2 ,3

    Environmental tobacco smoke (ETS), also known as second-hand smoke or passive smoking, contains many of the same carcinogenic compounds found in the mainstream cigarette smoke. The exposure of non-smokers to tobacco-specific carcinogens has been confirmed by detection and quantitation of the metabolites in the urine of people who had been exposed to ETS.8 ,9 Although uptake of tobacco-specific carcinogens among non-smokers exposed to ETS was only about one-seventieth that of active smokers, long term of intensive exposure to ETS could also lead to adverse effects.10

    Epidemiological studies on the relation between ETS and risk of pancreatic cancer among non-smokers are inconsistent and have not yet been jointly summarised. Therefore, we systematically reviewed and synthesised the existing epidemiological studies to evaluate whether exposure to ETS during childhood or adulthood is related to the risk of pancreatic cancer.

    Methods

    Study selection

    We carried out a comprehensive literature search using major electronic database (MEDLINE search first, followed by an EMBASE search) and manual searching of the reference lists of the relevant research reports and review articles to identify full text and abstracts published to date. Text words used for searching included ETS (the exploded index terms second-hand smoking, passive smoking, ETS) in combination with pancreatic cancer (the exploded index terms pancreatic cancer, pancreatic neoplasm or pancreas tumour). We also reviewed the reference lists of relevant articles.

    We limited our review to observational studies that involved human subjects, investigated the association between exposure to ETS and risk of pancreatic cancer among non-smokers, and were published in peer-reviewed scientific journals in English language only. The first round of selection was based on the examination of the titles and/or abstracts among all the identified articles. The full texts of all the eligible articles were further reviewed. Studies were finally included in the meta-analysis if they provided age-adjusted OR, or relative risks (RR) with 95% CI of the association between ETS and risk of pancreatic cancer.

    The systematic search on electronic database identified 20 studies from MEDLINE, 153 studies from EMBASE and two studies from hand searching of the reference lists of related publications. After exclusion based on previously mentioned criteria, nine eligible articles were selected into the systematic review and meta-analysis.

    Data extraction

    We extracted information from the nine eligible publications on study design, participant characteristics, assessment of exposure to ETS, assessment of diagnosis of pancreatic cancer, variables adjusted as confounders, RR or OR and 95% CI, according to the Meta-analysis of Observational Studies in Epidemiology studies.11 Studies are regarded as prospective if exposure to ETS was assessed before the diagnosis of pancreatic cancer and as retrospective if exposure was assessed after the cancer diagnosis.

    Considering exposure to ETS during childhood and adulthood may impact on the risk of pancreatic cancer through different physiological mechanisms, we abstracted RR estimates, as well as their 95% CI according to age of exposure. All the selected studies used the group with no exposure to ETS as the reference group for analysis. For studies that presented multiple levels of ETS exposure, only the RR estimates for the highest as compared with the reference group were used. For studies that presented RR estimates adjusted for different combinations of potential confounders, the ones that adjusted for the greatest number of potential confounders were used.

    Two included studies examined the relationship between childhood ETS exposure and risk of pancreatic cancer within the European Prospective Investigation into Cancer and Nutrition cohort.12 ,13 We chose to include only the study of Chuang et al12 when summarising the evidence for childhood ETS, because this study was conducted with a larger sample size and adjusted for greater number of potential confounders.

    One study reported the RR of pancreatic cancer separately for ETS exposures originating from mothers and fathers during childhood, but did not report a combined estimate.14 We contacted the first author of the article and obtained the combined estimate.

    Statistical analysis

    The summary RR estimates were estimated using both random-effects and fixed-effects models, but primary assessments were based on DerSimonian and Laird random-effects to account for variations between the results of individual studies.15 We evaluated heterogeneity across studies using the Cochrane Q statistic.16 To investigate the influence of a single study on the summary estimate, we computed the summary estimates omitting one study at a time. To evaluate the influence of study quality on the results of the meta-analysis, separate analysis for subgroups of prospective and retrospective studies was conducted whenever there were enough number of studies.

    Publication bias was assessed by performing the by the Begg and the Egger tests and visual inspection of funnel plots.17 ,18 All statistical analysis was conducted using STATA V.10 (StataCorp, College Station, Texas, USA). Statistical tests with a two-sided p value of <0.05 were considered statistically significant.

    Results

    Study characteristics

    Of the nine published articles identified for this review and meta-analysis: six had information on ETS exposure during childhood and eight studies had information on adult ETS exposure. All included studies are summarised in table 1.

    View this table:
    Table 1

    Studies included in the systematic review

    One of the 9 selected articles included two independent studies19 so there were a total of 10 studies considered, including 6 prospective cohort studies,12 ,13 ,19–21 3 retrospective case–control studies22–24 and 1 prospective case–cohort study.25 (table 1) Four of the studies were conducted in the USA, three in western European countries, one in Canada, one in Japan and one in Egypt (the only one in a middle-income or low-income country). Seven studies included both men and women, while the remaining three studies included only women. All studies reported RR estimates adjusted for age, and all but one study21 adjusted for multiple additional potential confounding factors. The degree of adjustment for other potential confounding factors varied substantially across studies.

    One cohort study included prevalent cases diagnosed before the start of follow-up,21 while all other prospective studies clearly stated that they only included incident cases. All the prospective studies assessed exposure at baseline only. Most of the prospective studies used cancer registries to ascertain pancreatic cancer cases, and one study14 used self-report questionnaires to identify cases followed by medical record review and case adjudication. The follow-up time of the prospective studies ranged from 821 to 24 years.20 The total number of cases varied by study, ranging from 19 in a prospective study21 to 294 in a case–control study.23 Case–control studies included larger number of cases compared to prospective studies.

    Childhood ETS exposure and pancreatic cancer

    In total, four prospective studies12 ,13 ,20 ,25 and two retrospective studies23 ,24 investigated the association between childhood ETS exposure and the risk of pancreatic cancer. As two prospective studies12 ,13 were both conducted within the European Prospective Investigation into Cancer and Nutrition cohort, only one of them12 was included in the analysis (further details are provided in the Methods section). After the exclusion, RR estimates ranged from 0.90 (95% CI 0.54 to 1.50; Heinen et al25) to 2.09 (95% CI 1.14 to 3.84; Chuang et al12). Of these, only the RR estimate from the study by Chuang et al12 reached statistical significance. The summary RR estimate from a random-effect model was 1.12 (95% CI 0.89 to 1.43, figure 1) and the summary RR estimate from a fixed-effect model was 1.08 (95% CI 0.91 to 1.29). We did not observe significant heterogeneity between studies (p=0.23). Subgroup analysis of the prospective studies had a summary RR estimate (RR 1.17, 95% CI 0.77 to 1.76) very similar to that of retrospective studies (RR 1.13, 95% CI 0.77 to 1.66), but neither was statistically significant.

    Figure 1
    Figure 1

    Study-specific and summary relative risk estimates (95% CI) of pancreatic cancer by exposure to  environmental tobacco smoke smoking during childhood among non-smokers. Weights were from random-effects analysis.

    Adulthood ETS exposure and pancreatic cancer

    ETS exposure at home and pancreatic cancer

    The association between ETS exposure at home during adulthood and risk of pancreatic cancer was reported in five prospective studies,19–21 ,25 and three retrospective studies.22–24 The individual RR estimates ranged from 0.78 (95% CI 0.44 to 1.39)25 to 6.0 (95% CI 2.42 to 14.90).22 The summary estimates were 1.23 (95% CI 0.86 to 1.77, figure 2) and 1.14 (95% CI 0.91 to 1.44) from a random-effects and fixed-effects models, respectively, with evidence of heterogeneity across studies (p=0.03). The study by Lo et al22 substantially influenced the summary estimate. When this study was excluded from the meta-analysis, the summary RR estimate was 1.02 (95% CI 0.81 to 1.30), and the p value for heterogeneity was 0.94. Subgroup analysis of the prospective studies had a lower summary RR estimate (RR 0.97, 95% CI 0.71 to 1.32) compared to that of retrospective studies (RR 1.94, 95% CI 0.74 to 5.12). However, when the study by Lo et al22 was excluded, the pooled RR estimate for the retrospective studies was 1.11 (95% CI 0.77 to 1.60).

    Figure 2
    Figure 2

    Study-specific and summary relative risk estimates (95% CI) of pancreatic cancer by exposure to residential  environmental tobacco smoke during adulthood among non-smokers. Weights were from random-effects analysis.

    ETS exposure at work and pancreatic cancer

    One prospective study25 and two retrospective studies23 ,24 evaluated the association between ETS exposure at work and the risk of pancreatic cancer independently from exposure at home. RR estimates ranged from 0.61 (95% CI 0.34 to 1.08)25 to 1.20 (95% CI 0.54 to 2.67)24 and 1.20 (95% CI 0.73 to 1.98).23 The summary RR estimates were 0.94 (95% CI 0.67 to 1.33; figure 3) and 0.95 (95% CI 0.71 to 1.37) from random-effects and fixed-effects models, respectively. Subgroup analysis was not conducted because of the small number of studies.

    Figure 3
    Figure 3

    Study-specific and summary relative risk estimates (95% CI) of pancreatic cancer by exposure to  environmental tobacco smoke at work among non-smokers. Weights were from random-effects analysis.

    Assessment of publication bias

    Neither the Begg nor the Egger test provided evidence for statistically significant publication bias for any of the above meta-analysis. However, the funnel plots seemed slightly asymmetrical for adult ETS exposure at home.

    Discussion

    After reviewing the existing literature, we did not find sufficient evidence to suggest that ETS exposure increases the risk of pancreatic cancer. Specifically, meta-analysis summary estimates for childhood exposure, adulthood residential exposure and adulthood exposure at work were statistically indistinguishable from the null hypothesis of no association.

    Pancreatic cancer is a rare disease and retrospective case–control studies are commonly used to investigate the risk factors for this disease. This approach is substantially more cost-effective and time-effective compared to prospective studies. Although, retrospective studies have numerous limitations, including the use of proxy data for exposure ascertainment, and the potential for recall and selection bias. Among the three retrospective studies included in this review, only one of them used proxy data for 24% of the cases.24 Prospective studies are generally less prone to these biases and therefore may yield more valid estimates of the association between ETS and pancreatic cancer. In our analysis, the meta-analytic summary estimates did not differ greatly between prospective and retrospective studies.

    Tobacco-specific nitrosamines are a group of carcinogens found only in tobacco products, including 3-(1-nitrosopyrrolidin-2-yl)pyridine, nicotine-derived nitrosamine ketone and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol), formed by nitrosation of nicotine and related tobacco alkaloids.10 ,26 Nitrosamines have already been proven to be potent pancreatic carcinogens in animal models.27 Most of the carcinogens detected in cigarette mainstream smoke have also been found in sidestream smoke.26 DNA adducts are formed from the exposure to tobacco-specific nitrosamines in tobacco smoke, causing mutations in the gene and increasing cell proliferation and transformation.26 In addition, tobacco smoke may trigger pancreatic cancer through chronic inflammatory responses.28 ,29

    In all included studies, information on exposure to ETS was collected by self-reported questionnaires. Ideally, the exposure to ETS should be measured in terms of the biological dose of the contaminant or its metabolites received by the target tissue. Current studies only collected the intensity and/or the duration of the exposure to ETS using questionnaires or interviews, which were unable to accurately estimate the dose of tobacco-specific carcinogens received by subjects. In addition, the amount of tobacco-relative carcinogens received from ETS exposure is influenced by a number of external factors, including the number of smokers in an enclosed area, the size of the area and the level of ventilation.30 Future studies could use the biochemical measures that have been developed for assessment of active cigarette smoking to evaluate the uptake of specific agents from exposure to ETS by the non-smokers.30

    Maternal smoking during childhood was only examined in one of the included studies by Bao et al20 (RR 1.52, 95% CI 0.97 to 2.39). As was shown in this study conducted in the USA with 37 757 never-active smoking women, exposure to maternal smoking was associated with an elevated risk of pancreatic cancer (RR 1.52, 95% CI 0.97 to 2.39), although not statistically significant. Future studies are needed to further confirm the effect of maternal smoking, including in utero exposure, on pancreatic cancer risk.

    The study conducted in Japan by Nishino et al21 did not adjust for confounders other than age. The most recent study examining exposure to childhood ETS and cancer risk by Chuang et al12 adjusted for age, sex, study centre, education level, alcohol consumption, body mass index, physical activity, vegetable and fruit consumption, energy intake, adulthood passive smoking.12 None of the other studies examining childhood exposure adjusted for adulthood exposure, and vice versa. The association between childhood ETS exposure and cancer risk may also be confounded by parents’ socioeconomic status, which was not adjusted in any of the current studies.

    A limitation of this meta-analysis is that different studies used slightly different population as the reference group. In the study by Villeneuve et al,24 for instance, the reference group consisted of those who had never been exposed to ETS, while in the studies by Bao et al20 and Heinen et al,25 the reference groups consisted of those who had not been exposed to smoke during the specific time period only.20 ,25 Therefore, the RR estimates could be biased towards the null if the individuals in the reference group were exposed to some ETS during some other time periods or from some other sources.

    In summary, in this meta-analysis with current data, we did not observe an increase in the risk of pancreatic cancer with exposure to ETS among non-smokers. In light of the fact that only 10 studies were included in our analysis, future studies need to more accurately quantify the detailed duration and intensity of ETS exposure to further investigate the association of ETS exposure in the aetiology of pancreatic cancer.

    Key messages

    • The first systematic review of the association between exposure to environmental tobacco smoke and risk of pancreatic cancer.

    • No evidence for an association was found; however, results should be interpreted with care because there are a limited number of studies.

    Acknowledgments

    This work was supported by National Institutes of Health Grant P01 CA055075.

    References

    1. ↵
      1. Altekruse SF,
      2. Kosary CL,
      3. Krapcho M,
      4. et al
      . SEER Cancer Statistics Review, 1975–2007. Bethesda, MD: National Cancer Institute. http://seer.cancer.gov/csr/1975_2007/ (accessed 4 Jun 2012).
    2. ↵
      1. Raimondi S,
      2. Maisonneuve P,
      3. Lowenfels AB
      . Epidemiology of pancreatic cancer: an overview. Nat Rev Gastroenterol Hepatol 2009;6:699–708.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Maisonneuve P,
      2. Lowenfels AB
      . Epidemiology of pancreatic cancer: an update. Dig Dis 2010;28:645–56.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Michaud DS
      . Epidemiology of pancreatic cancer. Minerva Chir 2004;59:99–111.
      OpenUrlPubMed
    5. ↵
      1. Lowenfels AB,
      2. Maisonneuve P
      . Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol 2006;20:197–209.
      OpenUrlCrossRefPubMed
    6. ↵
      1. Ghadirian P,
      2. Lynch HT,
      3. Krewski D
      . Epidemiology of pancreatic cancer: an overview. Cancer Detect Prev 2003;27:87–93.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Iodice S,
      2. Gandini S,
      3. Maisonneuve P,
      4. et al
      . Tobacco and the risk of pancreatic cancer: a review and meta-analysis. Langenbecks Arch Surg 2008;393:535–45.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Hecht SS,
      2. Carmella SG,
      3. Murphy SE,
      4. et al
      . A tobacco-specific lung carcinogen in the urine of men exposed to cigarette smoke. N Engl J Med 1993;329:1543–6.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Parsons WD,
      2. Carmella SG,
      3. Akerkar S,
      4. et al
      . A metabolite of the tobacco-specific lung carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in the urine of hospital workers exposed to environmental tobacco smoke. Cancer Epidemiol Biomarkers Prev 1998;7:257–60.
      OpenUrlAbstract
    10. ↵
      1. Hecht SS
      . DNA adduct formation from tobacco-specific N-nitrosamines. Mutat Res 1999;424:127–42.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Stroup DF,
      2. Berlin JA,
      3. Morton SC,
      4. et al
      . Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008–12.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Chuang SC,
      2. Gallo V,
      3. Michaud D,
      4. et al
      . Exposure to environmental tobacco smoke in childhood and incidence of cancer in adulthood in never smokers in the European prospective investigation into cancer and nutrition. Cancer Causes Control 2011;22:487–94.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Vrieling A,
      2. Bueno-de-Mesquita HB,
      3. Boshuizen HC,
      4. et al
      . Cigarette smoking, environmental tobacco smoke exposure and pancreatic cancer risk in the European prospective investigation into cancer and nutrition. Int J Cancer 2010;126:2394–403.
      OpenUrlPubMedWeb of Science
    14. ↵
      1. Bao Y,
      2. Michaud DS
      . Physical activity and pancreatic cancer risk: a systematic review. Cancer Epidemiol Biomarkers Prev 2008;17:2671–82.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. DerSimonian R,
      2. Laird N
      . Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Higgins JPT,
      2. Thompson SG
      . Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Begg CB,
      2. Mazumdar M
      . Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088–101.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Egger M,
      2. Smith GD,
      3. Schneider M,
      4. et al
      . Bias in meta-analysis detected by a simple, graphical test. Brit Med J 1997;315:629–34.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Gallicchio L,
      2. Kouzis A,
      3. Genkinger JM,
      4. et al
      . Active cigarette smoking, household passive smoke exposure, and the risk of developing pancreatic cancer. Prev Med 2006;42:200–5.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Bao Y,
      2. Giovannucci E,
      3. Fuchs CS,
      4. et al
      . Passive smoking and pancreatic cancer in women: a prospective cohort study. Cancer Epidemiol Biomarkers Prev 2009;18:2292–6.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Nishino Y,
      2. Tsubono Y,
      3. Tsuji I,
      4. et al
      . Passive smoking at home and cancer risk: a population-based prospective study in Japanese nonsmoking women. Cancer Causes Control 2001;12:797–802.
      OpenUrlPubMedWeb of Science
    22. ↵
      1. Lo AC,
      2. Soliman AS,
      3. El-Ghawalby N,
      4. et al
      . Lifestyle, occupational, and reproductive factors in relation to pancreatic cancer risk. Pancreas 2007;35:120–9.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Hassan MM,
      2. Abbruzzese JL,
      3. Bondy ML,
      4. et al
      . Passive smoking and the use of noncigarette tobacco products in association with risk for pancreatic cancer: a case–control study. Cancer 2007;109:2547–56.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Villeneuve PJ,
      2. Johnson KC,
      3. Mao Y,
      4. et al
      . Environmental tobacco smoke and the risk of pancreatic cancer: findings from a Canadian population-based case–control study. Can J Public Health 2004;95:32–7.
      OpenUrlPubMedWeb of Science
    25. ↵
      1. Heinen MM,
      2. Verhage BA,
      3. Goldbohm RA,
      4. et al
      . Active and passive smoking and the risk of pancreatic cancer in the Netherlands Cohort Study. Cancer Epidemiol Biomarkers Prev 2010;19:1612–22.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Hecht SS
      . Tobacco carcinogens, their biomarkers and tobacco-induced cancer. Nat Rev Cancer 2003;3:733–44.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Pour PM,
      2. Tomioka T
      . Tumours of the pancreas. IARC Sci Publ 1996;126:149–73.
      OpenUrlPubMed
    28. ↵
      1. Malfertheiner P,
      2. Schutte K
      . Smoking—a trigger for chronic inflammation and cancer development in the pancreas. Am J Gastroenterol 2006;101:160–2.
      OpenUrlCrossRefPubMed
    29. ↵
      1. Wittel UA,
      2. Pandey KK,
      3. Andrianifahanana M,
      4. et al
      . Chronic pancreatic inflammation induced by environmental tobacco smoke inhalation in rats. Am J Gastroenterol 2006;101:148–59.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      Committee on Passive Smoking NRC. Environmental tobacco smoke: measuring exposures and assessing health effects. Washington, DC: The National Academies Press, 1986.

    Footnotes

    • Contributors All authors have read and contributed to the article.

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.