Many smoking bans in the western world were introduced at the start of the millennium on the recommendation of the UN Organization WHO (World Health Organization), who claims that indoor tobacco smoke can cause lung cancer and heart disease in non-smokers.
Since then many skeptical scientists have said that the scientific evidence for this claim is weak [ 1 , 2 , 3 , 4 , 5 ] – and now a comprehensive statistical analysis of all published studies on the relationship between passive smoking and lung cancer has found that there is no link between passive smoking and lung cancer.
The analysis shows that passive smoking cannot be a cause of lung cancer. The overall result of all of the 102 published studies indeed shows a small increase in lung cancer risk of 1.22 (corresponding to a 22% increased risk), but the 22% is reduced to zero after error corrections in the worst studies and confounder adjustments, write the British statisticians Lee, Fry, Forey, Hamling & Thornton.
Lee PN, Fry JS, Forey BA, Hamling JS, Thornton AJ. Environmental tobacco smoke exposure and lung cancer: A systematic review. World J Meta-Anal 2016; 4 (2): 10-43
The analysis was done by the statisticians in the British consulting firm PN Lee Statistics and Computing Ltd, which for 30 years has performed expert analyses of poisons, drugs and diseases for the UN-organization IARC (International Agency for Research on Cancer), and for private companies in the pharmaceutical industry, the tobacco industry and the chemical industry. One of their tasks is monitoring the published studies on passive smoking for the tobacco industry, and these studies are collected in a database with associated analyses that are available on the company website.
Inverse link between
and lung cancer
The British statisticians have corrected the studies for several error sources, including ‘smoker misclassification’ (ie the fact that some smokers declare themselves as non-smokers in surveys, and thereby distort study results). They have also adjusted the material for the known ‘confounders’, ie other risk factors for lung cancer. These sources of error push up the overall risk of passive smoking to unrealistic highs, because they are not taken into consideration in many of the 102 studies, writes the statiticians.
In their adjusted meta-analysis of the studies the overall risk result was down to 1.08 (0.99 to 1.16), which is not different from zero, because the correlation is statistically insignificant. The result covers a negative (inverse) correlation between lung cancer and passive smoking in North America and Europe of 0.96 (0.86 to 1.07), while there is a small increased risk in Asia of 1.18 (1.07-1.30).
Because of this obscure link between passive smoking and lung cancer the statisticians discount any notion that secondhand smoke (ETS) could be a cause of lung cancer:
“One cannot reliably conclude that any true ETS effect on lung cancer risk exists. Our results suggest caution in drawing inferences from weak epidemiological associations where known biases exist.”
To put in clear layman’s language: When no clear link between passive smoking and lung cancer exists, it cannot be claimed that passive smoking causes lung cancer.
No risk of passive smoking in US & Europe
but figures on Asian women push up risk
The 102 studies have been performed in North America, Europe and Asia during the period 1981-2015. They are all epidemiological studies – that is, statistical population studies that have examined lung cancer-rates among non-smokers who in questionnaires have stated that they have been exposed to secondhand smoke at home, work or elsewhere. Their lung cancer rates have been compared to the corresponding rates of those who reported that they were not exposed to passive smoking in these places.
This is the standard method used to find potential disease associations for most risk factors.
The new analysis is an update of the statiticians’ 15 year old series of articles on passive smoking, in which they also disputed the official findings on lung cancer and passive smoking. Since then, many new studies have been published, but they have only watered down the weak statistical link between passive smoking and lung cancer further, according to the analysis.
The British point out many strange features in the 102 studies, when taken as they are – ie before correction: Only a few of the studies (app. 16%), report a statistically significant increased relative risk – this is most frequently in the smallest and simplest studies of low quality. The rest of the studies (nearly 85%) show no statistically significant increased risk. And more than 25% of the total number of studies show a negative (inverse) correlation between passive smoking and lung cancer – some of them statistically significant.
Another remarkable circumstance: It is only in the studies from Asia that a significant association between passive smoking and lung cancer can be found. In the rest of the world the relationship does not exist.
The unadjusted association between lung cancer and passive smoking between spouses in the 102 studies are insignificant in North America: 1.07 (0.94 to 1.23) and insignificant in Europe: 1.17 (0.99 to 1.39) while there is a statistically significant increased risk in studies from Asia: 1.33 (1.20 to 1.46), which pulls the entire ‘world-wide’ estimate up to an overall significant risk increase of 1.22 ( 1.14 to 1.31). Link to graphic
The British statisticians explain the difference between the Western and Asian studies by cultural differences: Asian women are more likely to deny that they are smokers in surveys than Western women.
No risk from
in European workplaces
This trend among Asian women affects the risk in the Asian studies in an upward direction because smokers usually marry other smokers – so when smoking Asian women are pretending to be non-smokers, then their much larger (smoking) risk of lung cancer is attributed to the non-smoking passive-smokers in the studies. This is called ‘smoker misclassification’ – it is found everywhere, but it is far more common in Asian studies, according to the analysis.
Graphics of the link between lung cancer and passive smoking at workplaces in North America, Europe and Asia: Almost all the results are distributed on both sides of the 1.0 line, indicating no risk. (Click to enlarge)
The picture does not seem clearer with the unadjusted association between lung cancer and passive smoking in workplaces. This relationship is of course interesting because it is the one that underlies the smoking bans.
Again, risk estimates are much higher in Asian studies than in studies from the rest of the world. In the studies of passive smoking in the European workplace the relationship is non-existent: None of the 11 published studies from Europe show a statistically significant association between lung cancer and passive smoking in the workplace. And 27% of the studies from Europe show a negative correlation (3 out of 11). See graphic.
Furthermore, the British examined the association between lung cancer and passive smoking in childhood and adolescence, in which there also can be seen a small increased risk in Asia. However the majority of studies on passive smoking in childhood from USA & Europe is negative, and the overall result of the Western studies is zero. In the United States: 1.06 and in Europe: 1.02, both before correction. Seegraphic.
Children from smoking homes
have less lung cancer risk
than children from smoke-free homes
An interesting – and statistically significant – negative result occurs when all studies of the relationship between lung cancer and passive smoking from parents in childhood and adolescence is combined: 0.78 (from 0.64 to 0.94). This means that children who grow up in smoking homes are significantly 22% less likely to get lung cancer than children who grow up in smoke-free homes. See table.
The link between passive smoking and lung cancer is in other words totally unclear. If there exists any lung cancer risk from passive smoking, the risk is so close to zero that it has no meaning and can be ignored.
The British point out that this is what you can expect if you extrapolate the risk of passive smoking down from the risks of active smoking: From measurements among non-smokers in smoking environments, it is known that the amount of smoke passive-smokers inhale is more than 1000 times lower than the amount which the smoker inhales directly from the cigarettes. Passive smoking in the home and on the job on average is equivalent to smoking 7-10 cigarettes a year. [ 6 ]
If extrapolation is done with this method the increased risk of passive smoking would end up at 1.008 – i.e. an increased risk of lung cancer less than one per thousand – compared to the 1.22 higher increased risk shown in the uncorrected statistical studies.
The analysis concludes, that the difference is due to the faults and biases in some of the studies. With a true understatement the Brits express it this way: The evidence that passive smoking (ETS) causes lung cancer is not ‘convincingly demonstrated’:
Most, if not all, of the ETS/lung cancer association can be explained by confounding adjustment and misclassification correction. Any causal relationship is not convincingly demonstrated.
Are the smoking bans
built on a simple lie?
The authors diplomatically refrain from writing about the smoking bans, which are based on the claim that passive smoking can cause lung cancer and heart disease. As it is clear from the published evidence, passive smoking has not been shown to cause more lung cancer than no passive smoking has done.
The lung cancer relationship has been the authorities’ strongest case against passive smoking. The link between passive smoking and heart disease is even more uncertain, which can be ascertained in the British statisticians’ unpublished paper ( 7 ) and associated meta-analysis ( 8 ) of the relationship between passive smoking and heart disease from 2015.
Here it appears that only one out of 22 studies (app 5%) has shown a significantly increased risk of heart disease from passive smoking in workplaces. The other 95% of the studies have shown no statistically significant increased risk. Half of the studies are negative ( Table 5 ), and the overall result before corrections are insignificant: 1.08 (0.99 to 1.19) ( page 2 ).
It may seem strange that the authorities have succeeded in banning passive smoking in many private companies all over the West, for there is no evidence of disease risk from passive smoking in public space when all of the published literature is considered. Nevertheless, many authoritative health reports have been particularly vocal on this – that smoking bans had to be implemented and maintained because passive smoking caused fatal diseases.
That is why the former medical chief of France’s biggest medical institute in an interview in 2010 stated that the smoking bans were built on a lie. ( 9 )