More on the WHO/Boffetta paper, hat tip Klaus K

I am very grateful for Klaus K for posting this on my blog. This is the whole idea of me starting it as I want an evidence based discussion. Thanks to Rollo for stimulating debate too. It seems the real OR for the WHO/Boffetta paper is 1.02. Second hand smoke is harmless. Klaus wrote.

“Some people discussed the IARC 1998 study exactly 1 year ago on Chris Snowdon’s blog, among them Mr Tommasi:
IARC 1998 is without any doubt a no-risk study, which becomes very clear when you examine the study’s Technical report with the raw figures – page 220 for workplace SHS:
Controlling for educational level and type of residence more than halved the SHS-workplace risk from unadjusted Odds Ratio 1.17 in the article to OR 1.08 (page 220, row 1, column 3). Furthermore: Eliminating 7 portuguese cases (out of 650 cases in total) with incomplete data reduces the ratio to OR 1.02 – (page 220, row 3, column 3).
More in the comments on Snowdon’s blog:”

This entry was posted in Uncategorized. Bookmark the permalink.

9 Responses to More on the WHO/Boffetta paper, hat tip Klaus K

  1. Rollo Tommasi says:

    Dave – And thanks to you for hosting the debate. Such coincidental timing!!! I’ve just posted a response to KlausK in the other thread on your site and see you’ve posted this. Here are the comments I made there, for ease of reference…..

    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!!!

    • Klaus K says:

      Mr. Tommasi, I don’t understand your use of the expression “cherry-picking”. I gave you the quotes from the IARC Technical Report one year ago because you at that time claimed that the official risk figures in IARC 1998 (fx. OR 1,17 for workplace SHS-exposure) were controlled for confounders – i.e. you claimed that the figures were “realistic” adjusted risk figures.

      The Technical Report shows that your claim was not true. The relative risk estimate for workplace SHS-exposure was OR 1,08 (0,83 – 1,39) when the authors controlled for residence type and educational level. (page 220, All centers, column 3).

      It is a serious weakness in IARC 1998 that more than half of the total risk disapperas when 20% of the cases with no data on these confounders are controlled for. Controlling also for a small handful of cases with no dietry data put the relative risk at OR 1,02. (page 220, All centers with dietry data, column 3).

      In other words: The small official risk of passive smoking in IARC 1998 exists only because of the figures from 3 study centers in Sweden and Portugal with insufficient data and some unusually high relative risks (however still insignificant). In the other 9 study centers (77% of the cases) with complete data, the risk was zero.

      Which is why I took the liberty to call IARC 1998 a “no-risk” study. As I have written to you earlier on Snowdon’s blog this is not surprising since there does not seem to be any health risks from passive smoking in European workplaces. The studies made here show no sigificant risk, even all of them (seven) pooled together: 1.13 (0.96-1.34) – including the IARC study. Please go check yourself: Surgeon General 2006, chapter 7 – page 436.

      • Rollo Tommasi says:

        KlausK – You are indulging in ridiculous cherry-picking!!!

        That 1.08 figure does not tell the actual risk! If you want to work out what the effect of educational status and type of residence is as a confounding factor, you need to look at the figures in the row “All Centers with Educational Level and Type of Residence”. You then compare the figure in the OR1 column (before adjusting for these potential confounders) with the figure in the OR3 column (after adjusting for these possible confounders). What you find is that the effect of adjustment is actually to INCREASE the OR slightly, from 1.06 to 1.08. What that shows is that these factors actually DO NOT confound the results. All of which means that the original OR of 1.17 is a reasonable reflection of the actual risk in these studies from passive smoking, with the results from Portugal and Sweden included – which is exactly what Boffetta et al concluded!

        1.17 is a reasonable reflection of the actual risk in these studies from passive smoking – which is exactly what Boffetta et al concluded!

        Your ludicrous cherry-picking then continues in that you choose to disregard the findings from Portugal and Sweden. Why? Simply because you don’t like their results! It’s certainly not because they’re affected by confounders – because results from the other studies show that the effects of these potential confounders is minimal – in fact, if anything they have the effect of increasing the OR!

        So your claim that the risk is “zero”, either in the other 9 cases or in all 12 cases undertaken, is completely wrong.

        And then, if that’s not enough cherry-picking, you indulge in even more in the US Surgeon General’s report. You select only the figures for Europe, for no other reason than you like its result most. There is no reason to ignore the results from the USA, Canada and Asia. Nor is there any reason for disregarding the results about the effects of workplace exposure on non-smoking women or non-smokers collectively. ALL of these results show a STATISTICALLY SIGNIFICANT increased risk of lung cancer from workplace exposure. The only other result without a statistically significant increased risk was for non-smoking men.

        So here’s a word to the wise. In future, try to read reports a bit more objectively – even if you don’t like the results.

      • Klaus K says:

        I am sorry to disappoint you, Mr. Tommasi, but the IARC 1998 study could find no lung cancer risk from passive smoking. No matter the careful words of the authors, that is what the cool figures show. The insignificant OR of 1,17 does not represent anything different from zero risk.

        The OR should be put in perspective of the true never-smoker risk of lung cancer, which is 10 to 100.000 in real life. The study thus suggests that many years of passive smoking could lead to an extra 1,7 lung cancer cases per 100.000. Is it possible to detect a tiny risk like this in a questionnaire case-control study with 650 patients? You tell me.

        Also, leaving aside all problems with measuring pasttime smoke in questionnaires, the OR 1,17 is still an unadjusted figure. It is uncontrolled for app. 30 competing lung cancer risk factors. It is not a “real” risk as you suggest.

        Removing a few cases with missing data on some of these competing risk factors more than halved the risk to OR 1.08, as you have shown yourself. Removing all 23% cases with missing data yielded zero risk.

        Removing 23 cases that were not confirmed lung cancer cases at all yielded a risk for the confirmed 627 cases at OR 1,11 according to the article.

        There is a very clear pattern emerging from the full information of the study: Everytime you remove a small handful of cases with doubtful or no data, you end up very close to zero risk.

        It is obvious that the OR 1,17 exists because a small high risk group of patients with low-quality data inflates the figures – some of them might not even be lung cancer patients. Without this group of patients with missing data the OR was not different from 1.0 – i.e. zero relative risk.

  2. Junican says:

    Rollo said (a couple of days ago):

    “”It takes a particularly callous kind of pro-smoker to argue that a risk which is responsible [for] an estimated 6,600 premature deaths in the UK each year from lung cancer and heart disease is “negligible”.””
    Out comes the word ‘premature’……….The word ‘premature’ seems to be very important – Lansley used it on TV and a letter from DOH to a complaint of mine used it. I suppose that you could say that anyone who has ever smoked or been exposed to SHS must, by definition, die prematurely. ‘Estimated’ 6,600?
    Here is a table from 2009 ONS stats for Eng & Wales:

    To 34………….55………….120
    85 +…………8800………25580
    (Myocardial is ‘heart attack’).
    (Ischaemic is ‘failure of blood supply to heart muscle’)

    A calculation based upon ‘life years’ gives an average ‘years lived (age)’ at death of about 85 years for Ischaemic heart disease, and yet Isch…. is by far the biggest cause of death – some 72,000. Nearest next is cerebrovascular (brain blood vessels – stroke) with some 43,000. Total deaths, 500,000.

    So, one must ask, at what age do we discount deaths for prematurity purposes? 65? 75? 85? And how premature is premature? A week? A month? A year? And when you have done that, you have to discount every other possible cause – including actual smoking itself.

    Oh, and there is another very, very peculiar thing. I mentioned Cerebrovascular deaths. In 2009, 43,000 people died thus, of which only 3000 were under 65; of those, 17,000 were male and 26,500 were female. Erm…….Just a minute………..In’t it true that an awful lot of the men (at that age) are likely to have been smokers and that most of the women are likely to have been non-smokers (and therefore SHS exposed)? So are we saying that vastly more SHS ‘suffers’ died from cardiovascular disease than full smokers?

    Oh I know that my reasoning is far from sound, but it is a damn sight sounder that the Health Department’s propaganda!


  3. Rollo Tommasi says:

    Junican: “Prematurely” means that these people die earlier than they would have, had they not been exposed to passive smoking.

    How much earlier? Obviously the answer to that differs from individual to individual. The key point is that this should be an irrelevant factor to you or any smoker who would happily expose these people to SHS. Their life expectancy is not yours to bargain over, just so can indulge in your craving whenever and wherever you want.

    As for SHS cerebrovascular disease, there is now quite a bit of evidence suggesting a link between the two. But you’ll find most scientists have been cautious about concluding that passive smoking does cause strokes. Certainly the 6,600 premature deaths figure I mentioned covers lung cancer and heart disease only; it excludes diseases like stroke and respiratory diseases even though there is some evidence linking these with passive smoking.

    All of which gives the lie to your claims about “propaganda”.

    It takes a particularly callous kind of pro-smoker to conclude on the basis of so little evidence that passive smoking must be “harmless”.

  4. Junican says:

    Well no, Rollo.
    You said:
    “”“Prematurely” means that these people die earlier than they would have, had they not been exposed to passive smoking””
    It follows, from that, that everyone who has been exposed to SHS dies prematurely.

    And so we see where the DOH gets the figure of 100,000 deaths as a result smoking from. 500,000 people per an die. 20% of people are smokers. 20% of 500,000 is 100,000. Therefore 100,000 people die from smoking. Sorry…..die PREMATURELY because of smoking. Funny, though, that Lansley, on the TV said 80,000, whereas his spokesperson from the DOH in the letter I received said 100,000. That is a very big difference. Which is the true figure? But also I remember Patricia Hewitt (maybe still a director (or something like that) of Boots Chemists) saying that the SHS deaths figure was 3,000. Funny how it has now become 6,600. Think of any figure you like then double it. What does it matter whether it is true or not, provided that it fits in with the propaganda?

    But here are some more stats from 2009 for you:


    To 34………………50
    85 +…………..2150

    Look at the horror of breast cancers which killed young women, compared with the incidence of Ischaematic Heart Disease, in young people. These figures are horrific! So where is the massive, massive publicity about the causes and etymology of breast cancer? Where are the conclusive studies which show the cause of breast cancer in young women? If all the studies have failed to find the etymology of breast cancer, how dare you guys claim to have discovered the etymology of lung cancer? Or is it that there is a lot more money to be made by ratcheting up the smoking issue, funded by big pharm and the eugenicists?

    One last thing. There was a huge scare about bird flu and another about swine flu. Do you know how many of the 500,000 people who died in 2009 died of flu? A total of……………….78, of which 37 were aged over 75. How much money do big pharm make out of flu jabs? The cost of emergency call outs? The cost of visits to the doctor?

    The NHS is being taken for a very expensive ride left, right and centre.

    BUT NOBODY CAN STOP IT! There is no mechanism to stop it, just like there is no mechanism for an EU edict to be withdrawn. It cannot be done. The situation is not unlike the position of Galileo when he proposed that the Earth goes around the Sun. The bible seems to say otherwise, and the bible must, by definition, be correct, therefore Galileo must be wrong.

    God! What a mess!

  5. Rollo Tommasi says:

    Junican – I genuinely am struggling to understand your logic. Surely you are not trying to suggest that scientists should devote all of their time to investigating one disease at a time? There is a heck of a lot of evidence about the causes of breast cancer ( and a heck of a lot of research going on into breast cancer (see e.g. So don’t try to pretend that science into causes of lung cancer and heart disease is preventing research into breast cancer. By the way, both lung cancer and heart disease kill many more people each year than breast cancer. My view is that research must try to find ways of cutting down the number of deaths from all of these diseases. I hope yours is too.

    My 6,700 deaths figure comes from Jamrozik’s 2005 BMJ article, and takes account of UK deaths from lung cancer and heart disease caused by exposure to smoke in the home. If you see other figures, you also need to check what the figures relate to – e.g. which SHS-related diseases do they cover; from what kind of exposure; do they cover England only, E&W, GB or UK?

  6. Rollo Tommasi says:

    In answer to Klaus K’s latest post: You claim “The insignificant OR of 1,17 does not represent anything different from zero risk.” That is WRONG. That is a lie used by the pro-smoking lobby. A statistically non-significant result does not mean “no risk”. It means that it is not possible to draw a statistically significant conclusion from the result about whether there is added risk or not. But that result can be read next to other results. In this case, the Boffetta finding was in keeping with other results showing an increased risk of lung cancer from passive smoking – an increased risk which is statistically SIGNIFICANT after proper meta-analysis.

    You say the risk of lung cancer for non-smokers is “10 to 100.000 in real life”. Well you are right but also wrong. That is the ANNUAL risk to a non-smoker, not their LIFETIME risk. As a result, after 40 years of exposure, an annual additional risk of 2-3 per 100,000 becomes 80-120 per 100,000. And that only covers the non-smoker’s risk of lung cancer from passive smoking. It doesn’t take account of their increased risk of heart disease and possibly other diseases too.

    You say the figure of 1.17 is unadjusted. That is true. But your attempts at “adjusting” the figure are simply crude efforts to corrupt the findings. As I have shown, when like cases are compared with like, accounting for education level, residence type and diet makes no difference to the RR. Your reference to “1.08” is irrelevant to a discussion about the real risk, because you are not comparing like studies with like studies. And you continue to prefer to exclude the Portugal and Sweden results, for the entirely arbitrary reason that you do not like them. Those results are no weaker than any others; Boffetta did not need to test all the studies for potential confounders. I could quite easily cherry-pick some of the studies myself to get an even higher RR – it would be no less valid than the dubious interpretation you’re trying to pull off.

    You say the Boffetta RR is “uncontrolled for app. 30 competing lung cancer risk factors”. But you make absolutely no effort to show how these factors would bring down the RR materially. It’s argument by innuendo.

    The truth is that Boffetta et al were right to conclude that the 1.17 RR was consistent with findings from other studies which pointed to an increased risk of lung cancer from passive smoking. As you pointed out yourself, the US Surgeon General’s report demonstrates how these results have been replicated on other continents.

    If you and Dave want to argue that the true added risk is close to zero, you’re going to have to use some proper science to back up your arguments, not the pseudo-science which Big Tobacco invented and now only the confirmed pro-smokers believe.

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