The World Health Organization’s 1998 Boffetta paper into passive smoking

In 1998 the WHO published their report into passive smoking, heart disease and lung cancer here is the abstract. The key finding in my opinion as that it proved no proof that passive smoking is harmful.


Background: An association between exposure to environmental tobacco smoke (ETS) and lung cancer risk has been suggested. To evaluate this possible association better, researchers need more precise estimates of risk, the relative contribution of different sources of ETS, and the effect of ETS exposure on different histologic types of lung cancer. To address these issues, we have conducted a case-control study of lung cancer and exposure to ETS in 12 centers from seven European countries. Methods: A total of 650 patients with lung cancer and 1542 control subjects up to 74 years of age were interviewed about exposure to ETS. Neither case subjects nor control subjects had smoked more than 400 cigarettes in their lifetime. Results: ETS exposure during childhood was not associated with an increased risk of lung cancer (odds ratio [OR] for ever exposure = 0.78; 95% confidence interval [CI] = 0.64-0.96). The OR for ever exposure to spousal ETS was 1.16 (95% CI = 0.93-1.44). No clear dose-response relationship could be demonstrated for cumulative spousal ETS exposure. The OR for ever exposure to workplace ETS was 1.17 (95% CI = 0.94-1.45), with possible evidence of increasing risk for increasing duration of exposure. No increase in risk was detected in subjects whose exposure to spousal or workplace ETS ended more than 15 years earlier. Ever exposure to ETS from other sources was not associated with lung cancer risk. Risks from combined exposure to spousal and workplace ETS were higher for squamous cell carcinoma and small-cell carcinoma than for adenocarcinoma, but the differences were not statistically significant. Conclusions: Our results indicate no association between childhood exposure to ETS and lung cancer risk. We did find weak evidence of a dose-response relationship between risk of lung cancer and exposure to spousal and workplace ETS. There was no detectable risk after cessation of exposure. [J Natl Cancer Inst 1998;90:1440-50]

This entry was posted in Uncategorized. Bookmark the permalink.

21 Responses to The World Health Organization’s 1998 Boffetta paper into passive smoking

  1. Rollo Tommasi says:

    Dave – In the CNN debate you claimed the 1998 Boffetta paper “proved that passive smoking is harmless”.

    Please tell me which part of the Boffetta report PROVES that passive smoking is harmLESS.

    • daveatherton says:

      The Boffetta paper says that there is not a link between lung cancer and passive smoking (PS). As the confidence intervals (CI) fall either side of 1.0 it is medically and scientific fraud to say that PS causes lung cancer and heart disease. Even if the CIs were starting at 1.0 and <2.0 because of confounding issues it still scientifically should not be accepted. In the case of smoking the massive confounder is misclassification. Most studies suggest misclassification of smokers is between 1%-4% averaging out at 2.5%. So an OR of 1.25 because the RR of smoking and lung cancer is 10 the 1.25 becomes 1.0. Here is some more detail

      Relative Risk or Risk Ratio (sometimes also called Hazard Ratio or Odds Ratio, but as the meaning of odds is quite different, especially in, for example, racing circles, this is to be avoided) is at the very heart of the dispute between epidemiology and real science. If X% of people exposed to a putative cause suffer a certain effect and Y% not exposed to the cause (or alternatively the general population) suffer the same effect, the RR is X/Y. If the effect is “bad”, then a RR greater than unity denotes a “bad” cause, while an RR less than unity suggests beneficial cause (and likewise if they are both “good”). An RR of exactly unity suggests that there is no correlation. There are a number of problems in a simplistic application of RR. In particular:

      1. Even where there is no correlation, the RR is never exactly unity, since both X and Y are estimates of statistical variates, so the question arises as to how much deviation from unity should be acceptable as significant.

      2. X and Y, while inherently unrelated, might be correlated through a third factor, or indeed many others ( for example, age ). Sometimes such confounding factors might be known (or thought to be known) and (sometimes dubious) attempts are made to allow for them. Where they are not known they cannot be compensated for, by definition.

      3. Sometimes biases are inherent in the method of measurement employed.

      4. Statistical results are often subjected to a chain of manipulations and selections which (whether designed to or not) can increase the deviation of the RR from unity.

      5. Publication bias can give the impression of average RRs greater than 1.5 when there is no effect at all.

      For these reasons most scientists (which includes scientifically inclined epidemiologists) take a fairly rigorous view of RR values. In observational studies, they will not normally accept an RR of less than 3 as significant and never an RR of less than 2. Likewise, for a putative beneficial effect, they never accept an RR of greater than 0.5. Sometimes epidemiologists choose to dismiss such caution as an invention of destructive sceptics, but this is not the case. For example:

      In epidemiologic research, [increases in risk of less than 100 percent] are considered small and are usually difficult to interpret. Such increases may be due to chance, statistical bias, or the effects of confounding factors that are sometimes not evident .[Source: National Cancer Institute, Press Release, October 26, 1994.]

      "As a general rule of thumb, we are looking for a relative risk of 3 or more before accepting a paper for publication." – Marcia Angell, editor of the New England Journal of Medicine"

      "My basic rule is if the relative risk isn't at least 3 or 4, forget it." – Robert Temple, director of drug evaluation at the Food and Drug Administration.

      "An association is generally considered weak if the odds ratio [relative risk] is under 3.0 and particularly when it is under 2.0, as is the case in the relationship of ETS and lung cancer." – Dr. Kabat, IAQC epidemiologist

      This strict view of RRs may be relaxed somewhat in special circumstances; for example in a fully randomised double blind trial, as opposed to an observational study, which produces a result with a high level of significance."

      • Rollo Tommasi says:

        Dave – I say this with all respect, but your comments show that your understanding of the issue is really floundering. First of all, it seems you don’t know the difference between “proof that something is harmless” and “insufficient proof that something is harmful”. You argued the former in the CNN debate but your only comments here are attempts to argue the latter. You are saying absolutely nothing which backs up your claim that passive smoking is PROVEN to be HARMLESS.

        Secondly, your understanding about Relative Risk is completely wrong! A relative risk of 2.0 or more is NOT required in all cases. It is ONLY expected in order to draw definitive conclusions from a single report. In other words, where the results of a study are reproduced in other studies, then a smaller RR can be perfectly acceptable.

        To prove my point, let’s take your quotes in turn. The National Cancer Institute quote comes from a press release about a study into the links between abortion and breast cancer. The release does say “relative risks of less than 2 are considered small and are usually difficult to interpret.” But the quote is made in the context of a single study which lacked supporting evidence. Far from stating the results of the study were irrelevant, the National Cancer Institute states “…the findings are not conclusive. Further research is needed to interpret the results”. In other words, the research was valid, but not conclusive in itself without further evidence. Here’s the release:

        The Robert Temple and Marcia Angell quotes relate to a discussion about the relative risk required before a “single epidemiologic study is persuasive by itself” (

        In fact, Numberwatch cynically cuts off a critical part of Marcia Angell’s quote. What she actually said was (my emphasis): “As a general rule of thumb, we are looking for a relative risk of 3 or more before accepting a paper for publication, PARTICULARLY IF IT IS BIOLOGICALLY IMPLAUSIBLE OR IF IT’S A BRAND-NEW FINDING”.

        To further prove my point, while Ms Angell was executive editor at the New England Journal of Medicine, it published a report which showed that “nonsmokers exposed to environmental smoke had a relative risk of coronary heart disease of 1.25 (95 percent confidence interval, 1.17 to 1.32) as compared with nonsmokers not exposed to smoke” and concluded that “Given the high prevalence of cigarette smoking, the public health consequences of passive smoking with regard to coronary heart disease may be important”? Surely if your claim about relative risk was right, she wouldn’t have accepted that study for publication.

        I don’t know where the IARC or Kabat quotes came from. But I know that IARC will similarly have considered this test only appropriate for ISOLATED studies, or else they would not have produced the monographs they have on the risks of passive smoking. Nor would they have published claims relating to other health risks which have nothing to do with smoking, such as a 30-50% increased risk of stomach cancer for smelter workers exposed to lead, and sufficient evidence about the risks of breast cancer from contraceptive use (again with studies typically showing a 30-50% excess risk). Similarly, if Kabat actually thought that a RR of 2 or more was ALWAYS required, the Enstrom and Kabat study would have said that clearly, which of course it did not.

    • daveatherton says:

      Rollo, let me give you a good example of confounding problems with epidemiology. Cervical cancer (CC) was always associated with smoking and even I would at a first look at these numbers would think that smoking is a risk for CC. “OR for current smoking habit 20+ per day = 2.57; 95% Cl = 1.49-4.45). This paper comes from 2000.

      Since then scientists have discovered the sole reason for CC is the Human papillomavirus types 16 and 18, genital warts to you and me. This is a quote from Cancer Research’s website.

      “For cigarette smoking there is a strong dose-response relationship. The risk of CIN 3 for women who were HPV positive and smoking 20 or more cigarettes a day was two and a half times that of women who had never smoked. The authors concluded that even though smoking was not a risk factor for HPV, smoking acted with HPV to cause cervical neoplasia.”

      I will post a few things more on relative risk and odds ratios later.

      • Rollo Tommasi says:

        Dave – Your own quote contradicts you! HPV16 and 18 cannot be “the sole reason for CC” when smokers who are HPV+ are 2.5 times more likely to acquire CIN 3 than non-smokers.

      • Frank says:

        Do you fully realise what you are saying, Mr/Mrs/Miss/Ms. Tommasi? Surely you’re not this silly or maybe just slightly desperate? For the avoidance of doubt, Dave Atheton is stating that, in this report, there is no evidence for SHS and CC or direct smoking and CC. The ‘2.5 times’ relates to direct smoking in people already diagnosed with HPV, not SHS!

        I’ve seen your comments before on other sites. You are reluctant to accept any form of RR or CI when attached to this issue of smoking.

        If you wish to be taken seriously – and we’re all in favour of other arguments – I suggest you start to make proper points and not carp.

    • Jan says:

      Why on earth should Dave have to prove that something is harmless ?
      Surely it is the job of the antis to ‘prove’ that something is harmful.
      That is something that they have so far failed to do.
      There is no study on SHS that PROVES that it is harmful.

      • Rollo Tommasi says:

        Jan – He needs to prove it is harmless because he said on national television that the 1998 Boffetta paper “proved that passive smoking is harmless”.

        Dave chose to make that claim. He needs to show it was not an empty claim. So far Dave has been completely unable to do that.

  2. daveatherton says:

    This is a textbook on epidemiology Rollo.

    “The threshold for concluding that an agent was more likely than not the cause of an individual’s disease is a relative risk greater than 2.0. Recall that a relative risk of 1.0 means that the agent has no effect on the incidence of disease.

    When the relative risk reaches 2.0, the agent is responsible for an equal number
    of cases of disease as all other background causes. Thus, a relative risk of 2.0
    (with certain qualifications noted below) implies a 50% likelihood that an exposed
    individual’s disease was caused by the agent. A relative risk greater than 2.0 would permit an inference that an individual plaintiff’s disease was more likely than not caused by the implicated agent.139 A substantial number of courts in a variety of toxic substances cases have accepted this reasoning.140

    also Steve Gold, Note, Causation in Toxic Torts: Burdens of Proof, Standards of Persuasion”$file/sciman06.pdf

    • Rollo Tommasi says:

      Once again Dave, your understanding is wrong. What you’re referring to is guidance for law courts, not for scientists.

      Your quote relates to what a court needs in order to find “individual causation” – i.e. that the substance caused the result in this particular case.

      Courts consider the question of whether passive smoking is harmful in general as part of their consideration of “general causation”. You will find that discussed in Section V of the manual. And in Section V you will find absolutely no reference is made to minimum RR thresholds, whether of 2.0, 3.0 or any other number.

  3. Bradley George says:

    “He needs to prove it is harmless because he said on national television that the 1998 Boffetta paper “proved that passive smoking is harmless”. ”

    I think you’ll have to concede to Rollo that, under the bright lights and pressured environment of a TV studio on a strict timetable, this was worded erroneously. That’s all he seems to need – a little victory in the detail to comfort his apparent fragile sense of self. The actual argument of whether or not ETS represents a genuine health risk to the public seems to be largely irrelevant to him.

  4. Fantastic! You’ve flushed Rollo out, you must be rattling some big cages, Dave.

    Great work. Keep it up, sunner. 🙂

  5. John Gray says:

    I am still waiting for Rollo Whoever and any of his cronies to prove that passive smoking is dangerous
    as none of the evidence offered so far stands up to close scrutiny. Also, I am still perplexed as to why he places any great store by this argument, and I quote from above:

    “Secondly, your understanding about Relative Risk is completely wrong! A relative risk of 2.0 or more is NOT required in all cases. It is ONLY expected in order to draw definitive conclusions from a single report. In other words, where the results of a study are reproduced in other studies, then a smaller RR can be perfectly acceptable.”

    If we switch studies to another subject area, let’s say seismic activity in London over the next 20 years, and 50 different studies are conducted which show there would be little risk of any major earthquake during that period, then the general conclusion that would be drawn would be that of little risk of major earthquake in London in the next 20 years. If we follow Rollo’s reasoning, the conclusion that should be drawn, contraversially, is that we should think there should be some significant risk purely on the basis of a number of studies showing low risk. Am I the only person who finds that baffling?

    • Rollo Tommasi says:

      John – The issue is not just the RR. It is also the actual risk of harm. The threat of passive smoking would be much, much less if the chances of a non-smoker getting lung cancer or heart disease were the same as an earthquake hitting London.

      But non-smokers DO get lung cancer and, especially, heart disease. When a non-smoker’s chances of getting either rise by 20-30% because of exposure to SHS, then even that modest relative risk is enough to account for thousands of premature deaths in the UK each year.

  6. Rollo Tommasi says:

    KlausK: I’ve not had time to look at these threads in recent days. Since you have now made your point twice, let me now at least respond to it.

    Your comments here and on VGIF only work if you engage in arbitrary cherry-picking. Why do you choose to quote one figure (for “All centers”) on page 122 yet a different figure (for “All centers with dietary data”) on page 220? The answer is just so you are able to obtain lower RR results. Not exactly an objective approach you’re taking, is it?

    So why should the results in column 3 be lower than in column 1? If you compare individual results, it’s clear it’s because column 3 excludes any results from 3 studies from Portugal and Sweden. If you compare the individual results for columns 1 and 3 on each row, you’ll find that on page 122 as many studies show increased RRs as reduced RRs after controlling for residence type and educational level. On page 220, almost all studies show increased RRs after controlling for residence type and educational level. So there is absolutely no pattern that controlling for these things reduces the RR – if anything, it’s more likely to INCREASE the RR.

    All of which confirms the study’s findings. And renders irrelevant your claim that this is a “no risk” study.

    And of course not even what you say supports Dave’s claims that the Boffetta study proved that passive smoking is harmless!!!

  7. Belinda says:

    John … no you are not the only person who finds that baffling. Rollo answers by stating that the risk of harm makes the difference, and the fact that non-smokers get lung cancer proves his case. He might as well say right out that correlation equals causation. The fact is that it isn’t established that lung cancer is caused by passive smoking, and the fact that there’s a lot of lung cancer in non-smokers doesn’t make it any more likely that passive smoking is the cause, when there are many variables involved and probably not all of them known yet.

    The logic of Rollo’s position is that the more people study the issue and reveal a negligible risk, the more likely it is to be considered significant (rather than negligible) by the world at large. So you could actually pay people to do these studies, knowing that the world would find a significant risk when the studies’ findings are that the risk is negligible.


    • Rollo Tommasi says:

      If Belinda had read my post properly, she would realise that I am not trying to suggest that non-smokers only get lung cancer because of passive smoking. I quite clearly said that non-smokers who are exposed to SHS are 20-30% more likely to get lung cancer (or indeed heart disease) than those who are not.

      Belinda prefers to cling to an scintilla of doubt that passive smoking could be harmful than recognise that the health of people exposed to SHS matters. It takes a particularly callous kind of pro-smoker to argue that a risk which is responsible an estimated 6,600 premature deaths in the UK each year from lung cancer and heart disease is “negligible”.

  8. Belinda says:

    The logic of Rollo’s position is that the more people say the risk elevated risk *associated with smoking* is 1.25 (which however you look at it is a small *risk* that doesn’t establish causation) the bigger the risk becomes. If anything a causative association with tobacco can be made with more confidence, but it is only ever hypothetical.

    The rest of his post can be dismissed as the usual abusive insinuations made by anti-smokers about their opponents.

  9. Rollo Tommasi says:

    Belinda – I am not suggesting that the risk becomes greater. However, I am saying that, as more and more studies conclude an excess risk of lung cancer and heart disease from passive smoking, the more corroborative evidence there is that SHS is a cause of these diseases. Demonstrating causation depends on much more than the size of the relative risk. The standard test is a set of criteria for causation outlined by Bradford Hill. As Taylor et al state (International Journal of Epidemiology (2007)), the association between passive smoking and lung cancer fulfils all but one of these criteria. So the grounds for concluding that passive smoking is a cause of lung cancer and heart disease are strong.

    As for my other remarks, I make no apology for them. As I said, that supposedly “negligible” risk is enough to account for an estimated 6,600 premature deaths in the UK each year from lung cancer and heart disease. Anyone who considers that scale of deaths to be “negligible” is absolutely callous in my opinion. The cap fits, Belinda.

  10. Belinda says:

    sorry Rollo, I can’t get it on (the cap). Clearly if I thought there were grounds for believing that SHS caused so many deaths I would be concerned about it, just as I am concerned about the side effects of Champix that anti-smokers believe is ‘worth the risk’ (people with no psychiatric history doing themselves in or suffering psychological trauma). The deaths are hypothetical and you don’t actually know what caused them. I don’t know what makes you think all these deaths from heart disease and so on are ‘premature’. What is your understanding of premature? If someone is say 94 and dies of what you consider is passive-smoking induced cardiac arrest, would you consider this premature? (by this logic of course any passive smoking induced death could be premature, or indeed almost any death – all deaths are caused by something).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s