Prohibitionist tirades against smokeless tobacco invariably focus on a presumed association with cancer. For example, these “fact sheets” from the American Cancer Society and the Campaign for Tobacco Free Kids are terribly misleading for all but very sophisticated readers. They boldly declare that smokeless tobacco causes cancer, but they fail to provide ANY evidence to support the claim. There’s a good reason for this deliberate misrepresentation: Scant evidence exists to link smokeless tobacco use with any cancer.
The Cancer Society and other anti-tobacco extremists cherry-pick various studies, taking a few isolated epidemiologic findings to justify their claims and ignoring the rest of the evidence that shows little or no risks from smokeless tobacco. This isn’t a scientific approach, and it should play no role in tobacco policy development. Unfortunately, this is exactly what has driven tobacco policy in the U.S.
Last week, United Kingdom epidemiologists Peter Lee and Jan Hamling published a comprehensive analysis of smokeless tobacco use and cancer in BMC Medicine, the flagship medical journal of the Biomed Central series, covering all aspects of medical science and clinical practice. Articles in this journal “need to be of outstanding quality, broad interest and special importance;” the Lee-Hamling study meets these criteria.
This study looked at ALL the epidemiologic evidence linking smokeless tobacco use and cancer. It compiled statistics from 89 studies, and it used a straightforward technique to separate the risk related to smokeless tobacco use from the risk related to smoking and alcohol consumption. That is especially important, since smokeless tobacco users may also have a history of smoking and heavy drinking, both of which are established risk factors for cancers of the oral cavity, throat and esophagus.
For each cancer, Lee and Hamling produced a summary of the relative risk (RR) of cancer among smokeless tobacco users, compared with non-users of tobacco. This is followed by a confidence interval (CI), which is the range within which the RR lies with 95% confidence. An RR of 1.0 indicates that the risk among smokeless users is the same as that among non-users, and any CI that spans 1.0 indicates that the RR is not statistically significant. An RR of 1.2 indicates an increase of 20%, while an RR of 2.0 indicates a doubling of the risk. It is also important to note that small RRs (those under 2) should not be seen as definitive evidence that ST caused those diseases. In fact, the National Cancer Institute advises: “Relative risks or odds ratios less than 2 are viewed with caution,” because they “are sometimes difficult to interpret.”
Lee and Hamling found 41 studies that reported risks for oral cancer. For all studies, the RR was 1.79 (CI = 1.36-2.36), indicating a modest elevation in risk. However, in the 19 studies that accounted for smoking, the RR was 1.36 (CI = 1.04-1.77); in the 10 studies that accounted for both smoking and alcohol, the RR was 1.07 (CI = 0.84-1.37). Thus, there is virtually no evidence that smokeless tobacco is an independent cause of oral cancer.
Lee and Hamling also found that, for studies published since 1990, the RR for smokeless use was 1.28 (CI = 0.94-1.76). This means that no significant oral cancer risk has been detected in users of modern smokeless tobacco products, which should provide reassurance for contemporary users.
Lee and Hamling examined the evidence for many other cancers. Here is a summary of the risks among smokeless users (accounting for smoking), which can also be seen in Table 30 of their manuscript:
Esophagus: RR = 1.13 (CI = 0.95-1.36)
Stomach: RR = 1.03 (CI = 0.88-1.20)
Pancreas: RR = 1.07 (CI = 0.71-1.60)
All Digestive Tract: RR = 0.86 (CI = 0.59-1.25)
Larynx: RR = 1.34 (CI = 0.61-2.95)
Nasal: RR = 1.14 (CI = 0.73-1.77)
Lung: RR = 0.99 (CI = 0.71-1.37)
Prostate: RR = 1.29 (CI = 1.07-1.55)
Bladder: RR = 0.95 (CI = 0.71-1.29)
Kidney: RR = 1.09 (CI = 0.62-2.94)
All Cancers: RR = 0.98 (CI = 0.84-1.15)
The bottom line is that there is very little evidence that smokeless tobacco use is associated with any cancer. The only statistically significant finding in that list is for prostate cancer (RR = 1.29, CI = 1.07-1.55). But that RR is based on only four epidemiologic studies. As Lee and Hamling point out, the data for this finding “are inadequate for a clear conclusion…Prostate cancer is not considered smoking related, and more information on its relationship with smokeless tobacco is needed before any clear conclusion can be drawn.”
What If All Smokers Had Used Smokeless Instead?
Lee and Hamling also calculated how smokeless tobacco use might have changed cancer deaths among American men. In 2005, 142,205 men in the U.S. died from the 7 cancers associated with smoking. If no American men had ever smoked, there would have been only 37,468 cancer deaths, so 104,737 were directly attributable to smoking. Using the RRs above, Lee and Hamling calculated the number of cancer deaths that would have occurred if all smokers had instead used smokeless tobacco. The number attributable to smokeless tobacco would have been 1,102, which is only 1.1% of the deaths currently attributable to smoking.
Lee and Hamling then calculated another extraordinary statistic, a sort of worst-case scenario in which every man in the U.S. used smokeless tobacco. I assume that the investigators did this in order to counter anti-tobacco extremists who claim that releasing accurate information about the risks of smokeless tobacco would cause everyone to use it. In that case, according to Lee and Hamling, there would be 2,081 deaths attributed to smokeless use – a mere 2% of the deaths currently attributable to smoking.
It is difficult to exaggerate the importance of this study. The clear implication is that the deaths each year of 103,635 male smokers in the U.S. could be avoided if public health leaders shared the truth about safer smokeless tobacco, and if tobacco manufacturers were allowed and encouraged to urge smokers to switch.
As profound as the Lee/Hamling research is, it is astounding how little public and professional attention has been paid to it. Their study has to date been the subject of only a single media blog article in the Los Angeles Times on July 29.
The scourge of smoking-related deaths, preceded in most cases by years of debilitating and costly illness, should be viewed as an indictment of our public health and regulatory systems. Given the wealth of published research on the role of smokeless tobacco in tobacco harm reduction, aggressive smoker education on the subject should be a national and global public policy priority.