« The DRAM, Vol. 14(2) – Who’s Got the Influence? Youth friend and family influences on DUI behavior | Main | STASH, Vol.14(2) - 🍃🍃 Is that what they’re calling marijuana now? »

Friday, February 09, 2018


Oslo, February 24 2018
Fake evidence and scientific misconduct regarding health risks at 1 cigarette a day
Erik Nord, senior researcher at the Norwegian Institute of Public Health and professor in health economics at the University of Oslo.

The report by Hackshaw et al on disproportionately high risks of CVD associated with smoking one or a few cigarettes a day, published in the BMJ on January 24, addresses an important public health issue, but is false in content and represents scientific misconduct. The media that picked up Hackshaw et al’s results and forwarded them to the public should rectify their earlier message and look at the matter from a different, but no less interesting news angle: How is it possible that completely useless research can be accepted for publicaton in the British Medical Journal?
It is true that there is some earlier evidence that risk of cardiovascular disease increases considerably already at low levels of cigarette consumption. Hackshaw et al’s extensive literature review purports to bring further evidence, focussing particularly on risk at 1 cigarette a day, and Hackshaw et al clearly give the impression that they have very extensive observations of risk at this level of consumption. But the studies reviewed by Hackshaw et al have absolutely no data on one cigarette a day. They do not even come close. The subgroups of smokers with the lowest reported levels of consumption in 50 % of the studies have a mean consumption of around 7-8 cigarettes per day and in 40 % of the studies around 11-14 CPD.
Details are reported at http://www.bmj.com/content/360/bmj.j5855/rapid-responses . I here reiterate the essential table, as this unfortunately collapsed in the online publication in the BMJ.
Distribution of 27 studies in figures 1, 3 and 4 on ‘lowest consumption category used in the study’:
Lowest interval Number of studies Midpoint used by Hackshaw et al Evidence based mean
1-4 1 2.5 3
1-7 1 4 5-6
1-9 14 5 7.5-8
1-14 6 7.5 11-12
1-19 5 10 13-14
Sum 27

To avert any misunderstanding, I stress that the point here is not that low level smoking should be thought of as having a negligible risk. From a public health point of view, but also in view of the pleasure that many derive from modest smoking and the problems many have with quitting completely, it is an important research task to establish how high the risk of ‘one or two a day’ is. The point here is that no matter what the research theme is, and hence even when the theme is consumer behaviour of which the majority of people and health politicians disapprove (here: smoking), researchers must try to remain neutral researchers and respect the rigorous rules of empirical science and evidence based medicine. Otherwise policy makers will be led to make decisions on false premises, and over time the general public’s faith in science will be compromised.
In the present case, I do not doubt Hackshaw et al’s good intentions. But their misrepresentation of evidence is – de facto - misconduct. The severity of it is amazing. The only reasonable explanation seems to be that they are on a mission. Unfortunately they are not the first ones, cfr http://www.bmj.com/rapid-response/2011/10/30/passive-smoking-non-smoking-researchers%E2%80%99-personal-values-cause-conflict-in , and will surely not be the last ones.
I conclude that the media should reopen the case.

Hi Erik,

Thank you for reading this review and raising such important concerns. In the link you provided, Hackshaw et al. made a rebuttal on 12 February, 2018 addressing several of the points you raised. We provide them below:

“…One cigarette per day was always within the observed consumption ranges used in the studies, therefore a data point of say 5 per day actually represents an average of 5 per day and thus include down to 1 per day. As such, estimating relative risks (RR) for 1 per day does not represent extrapolation, but rather interpolation. Please note that there would not be a category of 0-10 or 0-5 per day because this will mix up smokers and never-smokers; the categories are always of the form 1-5 or 1-10. Furthermore, our estimates for 1 per day, using modelling based on categories of cigarette consumption, are close to that using the CPS II study.

…To clarify, we were not providing absolute risks of cardiovascular disease in any group. The official press release reported the excess (relative) risks. Some individual news items followed this, whilst others (in addition or instead) chose to extract information from the BMJ paper on the relative risk compared to never smokers. They are different measures, and this probably explains the understandable confusion.

…it is not correct that the RR for a category of say 1-10 per day is taken to be the same as for 1 per day. A published estimate for 1-10 per day is used in a regression, producing a model which can then be used to estimate RR for 1 per day…”

All the best,
Pat Williams

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Comments are moderated, and will not appear until the author has approved them.

Your Information

(Name and email address are required. Email address will not be displayed with the comment.)