Intended for healthcare professionals


Passive smoking: history repeats itself

BMJ 1997; 315 doi: (Published 18 October 1997) Cite this as: BMJ 1997;315:961

Strong public health action is long overdue

  1. Ronald M Davis (rdavis1{at}, Editor, Tobacco Controla
  1. a Director, Center for Health Promotion and Disease Prevention, Henry Ford Health System, One Ford Place, 5C, Detroit, Michigan 48202-3450, USA

    In 1962 and 1964 the Royal College of Physicians in London and the surgeon general of the United States released landmark reports documenting the causal relation between smoking and lung cancer.1 2 During the next quarter century, extensive research confirmed that smoking affects virtually every organ system. By 1990, the surgeon general concluded that “smoking represents the most extensively documented cause of disease ever investigated in the history of biomedical research.”3

    The history of research on passive smoking followed a parallel course. The 1982 surgeon general's report on smoking and cancer reviewed the first three epidemiological studies published on the relation between passive smoking and lung cancer. Each showed an increased risk of lung cancer in non-smoking women whose husbands smoked. But because the evidence was not yet abundant, the report's conclusions were cautious.4

    In 1986 the surgeon general devoted an entire report to the topic of involuntary (passive) smoking.5 It reviewed 13 “spousal studies” on passive smoking and lung cancer, 11 of which showed a positive association. The surgeon general was now able to conclude that “involuntary smoking is a cause of disease, including lung cancer, in healthy non-smokers.” Also in 1986, four other reports from authoritative bodies in the United States, Britain, France, and Australia came to similar conclusions.6 7 8 9

    The next watershed was publication of a comprehensive report on environmental tobacco smoke by the US Environmental Protection Agency in January 1993.10 By that time, 30 epidemiological studies on passive smoking and lung cancer had been published from eight countries, 24 of which showed a positive association. The Environmental Protection Agency classified environmental tobacco smoke as a known human carcinogen, to which it attributed 3000 lung cancer deaths annually in American non-smokers. The agency also documented causal associations between exposure to environmental tobacco smoke and lower respiratory tract infections such as pneumonia and bronchitis, middle ear disease, and exacerbations of asthma in children.

    The agency calculated extremely low probabilities that the epidemiological findings had occurred by chance: a one in 10 000 probability that 24 of 30 studies would show a positive association between passive smoking and lung cancer; a one in 10 million probability that 17 out of 17 studies characterised by exposure level would show an increased risk at the highest exposure level; and a one in a billion probability that 14 out of 14 studies would show positive dose-response trends.11

    An equally impressive report on environmental tobacco smoke was published last month by the California Environmental Protection Agency.12 That report, like its federal counterpart, was the subject of extensive peer review, expert committee review, public comment, and revision. It affirmed the findings of the US Environmental Protection Agency on the link between environmental tobacco smoke and lung cancer and respiratory illness. It also concluded that passive smoking is a cause of heart disease mortality, acute and chronic heart disease morbidity, retardation of fetal growth, sudden infant death syndrome, nasal sinus cancer, and induction of asthma in children.

    Two important studies from the Wolfson Institute of Preventive Medicine in London, published in this week's BMJ, comprise the latest chapter in the history of passive smoking research. Hackshaw et al (p 980) conducted a meta-analysis of the epidemiological studies on passive smoking and lung cancer, which have now reached 37 in number.13 After careful adjustment for bias and dietary confounding, they determined that marriage to a smoker increased the risk of lung cancer by 26% (95% confidence interval 8% to 49%), a conclusion bolstered by strong evidence of a dose-response relation and by linear extrapolation of risk in smokers.

    Law et al (p 973) conducted a meta-analysis of 19 epidemiological studies of environmental tobacco smoke and ischaemic heart disease.14 After adjusting for dietary confounding, they determined that environmental tobacco smoke caused a 23% increase in risk of ischaemic heart disease (95% confidence interval 14% to 33%). They found confirmatory evidence from studies of active smoking, which showed a substantial effect on risk of ischaemic heart disease at low dose. The authors reviewed human and animal studies of exposure to environmental tobacco smoke in relation to arterial atheromatous disease, platelet aggregation, and infarct size after experimental occlusion of a coronary artery, which help to explain the biological plausibility of a low dose effect of environmental tobacco smoke on the risk of ischaemic heart disease.

    The systematic reviews from the Wolfson Institute,13 14 the California Environmental Protection Agency,12 and the US Environmental Protection Agency,10 and the five reports released in 19865 6 7 8 9 make it clear that exposure to environmental tobacco smoke is a cause of lung cancer, heart disease, and other serious illnesses. In the United States alone, it is responsible each year for 3000 deaths from lung cancer, 35 000 to 62 000 deaths from ischaemic heart disease, 150 000 to 300 000 cases of bronchitis or pneumonia in infants and children aged 18 months and younger (causing 136 to 212 deaths), 8000 to 26 000 new cases of asthma, exacerbation of asthma in 400 000 to 1 million children, 700 000 to 1.6 million visits to physician offices for middle ear infection, 9700 to 18 600 cases of low birth weight, and 1900 to 2700 sudden infant deaths.12 Those figures make passive smoking one of the leading preventable causes of premature death in the United States.15

    History repeats itself not only in research on active and passive smoking, but in the actions of the tobacco industry to deny and obfuscate the findings of that research. The latest example, which compares the hazards of second-hand smoke with the “risks” of drinking milk and eating biscuits,16 is as inane as were the industry's denials of the hazards of active smoking in past decades. Their public pronouncements are particularly cynical in the light of contradictory statements in their internal documents,17 and their recent settlement (for $300 m) of the class action lawsuit in Florida on behalf of flight attendants harmed by second-hand smoke (p 968).

    Public health action to eliminate exposure to environmental tobacco smoke is long overdue. The minimum acceptable standard for indoor facilities is to allow smoking only in physically separated and separately ventilated areas.18 19 A total ban on smoking is preferred on three grounds: it provides maximum protection of non-smokers, it avoids exposing smokers to extremely high levels of environmental tobacco smoke in designated smoking areas,20 and it avoids the costs of constructing separately ventilated smoking areas. Health advocates should pursue all strategies that would help accomplish that goal, including education, legislation, regulation, and litigation.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.