Wednesday, August 16, 2017

Government-Funded Researchers Continue to Misrepresent Smokeless Tobacco Cancer Risks



The American Journal of Epidemiology in October 2016 published a report by Annah Wyss of the National Institute of Environment Health Science and 20 government-funded coauthors.  It revealed that American men had ZERO mouth cancers associated with dipping or chewing tobacco (Odds Ratio, OR = 0.9), while women, who mainly use powdered dry snuff, had a 9-fold elevated risk (here). 

After trying for several months to correspond with the authors, I wrote to the journal editor.  My letter was accepted, and the authors were invited to reply.  The journal published both items on August 4 (here).

Here is my letter.  Because the references are important, I provide links to the sources.


“In the article… Wyss et al. omit key information that would likely yield critical insights about their most important result.

“Wyss et al. reported that ever use of snuff among never cigarette smokers was associated with head and neck cancer (OR = 1.71, [95% Confidence Interval] CI: 1.08, 2.70), based on 44 exposed cases and 62 exposed controls.  The OR among women ever snuff users (8.89, CI: 3.59, 22.0, based on 20 exposed cases and 12 controls) was an order of magnitude higher than that in men (0.86, CI: 0.49, 1.51, based on 24 exposed cases and 50 controls). 

“The striking difference between women and men reflects completely different snuff exposures.  It is widely known that in the southern U.S. women primarily use powdered dry snuff, whereas men throughout the U.S. use moist snuff.  Powdered dry snuff use is associated with excess oral cancer risk in four previous studies (Reference 2), all of which were cited by Wyss (3, 4, 5, 6).  In contrast, moist snuff is associated with minimal to no risk in eight previous studies (2).

“Wyss and colleagues are knowledgeable about the use of powdered dry snuff by women and its cancer risk; one was first author of a 1981 study reporting that “[t]he relative risk [for oral and pharyngeal cancer] associated with snuff dipping among white [women] nonsmokers [in North Carolina] was 4.2 (95 per cent confidence limits, 2.6 to 6.7).” (4)  Subjects in that study had exclusively used powdered dry snuff.  

“Wyss et al. used pooled data from 11 US case-control studies located throughout the U.S.  They have information to confirm that the 20 female [powdered dry] snuff users with cancer were from study sites in the Southern U.S., whereas the 24 male [moist] snuff users were from more diverse locations.  This information is not available from the 12 publications cited by Wyss et al. as data sources.  Six publications had insufficient information about exposed and unexposed cases and controls (i.e. no specificity with respect to smokeless tobacco type, gender, and/or smoking), three reported no results related to any smokeless tobacco product and three did not mention smokeless tobacco at all.

“Wyss et al. should provide in tabular form the number of exposed and unexposed cases and controls separately for never smoking men and women from each of the 11 source studies.  This information is routinely provided in publications of pooled analyses, and it is available in the existing data and results files.”



In their response, Dr. Wyss and her coauthors provided additional results, per my request.  While they confirmed no risks among men who dip, their conclusion about women was grossly inaccurate: “Snuff use was strongly associated with elevated [head and neck cancer] risk in both Southern women…and non-Southern women.”

In Southern women, the risk was quite high (OR = 11.25) and statistically significant, but with a wide 95% confidence interval (2.14 – 59.07), owing to the fact that the OR was based on only 26 cases.  In non-Southern women, the OR was even higher, at 15.91, but this was based on a mere 6 cases.  While statistically significant from a technical perspective (CI = 1.20 – 211.43), the CI is so large and unstable that the estimate is virtually meaningless.

Dr. Deborah Winn of the National Cancer Institute knows that powdered dry snuff is the only smokeless product with mouth cancer risk (evidence here, here, here and here); she published a study in 1981 that falsely implicated dip and chew.  Even though Dr. Winn co-authored the new report, the authors’ only comment on powdered dry snuff was, “we cannot comment on whether snuff users – female or male, Southern or non-Southern – in our analysis used dry or moist snuff.”

To be clear, numerous studies since 1981 have demonstrated that the huge difference in mouth cancer risk between women and men is attributable to powdered dry and moist snuff.  Yet federally funded researchers continue to misrepresent these established facts in order to sustain the fiction that all smokeless products are hazardous.

 

Wednesday, August 9, 2017

E-Cigarettes: The Case for Regulatory Restraint



Debate over FDA regulation of e-cigarettes often degenerates into bitter argument and ad hominem attack.  In contrast, a point/counterpoint presentation in the Journal of Policy Analysis and Management is calm, thoughtful and enlightening.

The case for stringent FDA regulation (citations here and here) is presented by Kurt Ribisl, a professor at the University of North Carolina, along with graduate students Andrew Seidenberg and Elizabeth Orlan (hereafter RSO).  Cornell University Professor Donald Kenkel makes the case for light regulation (here and here).

Describing the problem – “Smoking is estimated to result in 480,000 deaths per year in the United States” – Kenkel underscores a death toll that has not changed in 20 years, despite increasingly harsh anti-smoking policies.  Tobacco prohibitionists frequently cite this horrid data, but they rarely support vastly safer alternative products like e-cigarettes, referred to by Kenkel as “perhaps the most revolutionary product innovation in the market…”

Kenkel puts the risk differences in striking contrast.  He notes that if Britain’s health authorities are correct that vaping is no more than 5% as hazardous as smoking, then a world in which all smokers vape would suffer only 24,000 deaths attributable to e-cigarettes, compared to 480,000 actual deaths due to smoking.  This projection is similar to that in my 1994 publication (here), where I estimated 6,000 deaths among an imagined 46 million smokeless tobacco users – a mere 2% of the 419,000 actual deaths from smoking.

RSO acknowledge that vaping is less risky than smoking, but they reject the 5% estimate, arguing that it “lacks empirical basis, as it came from a multicriteria decision analysis model developed from one meeting of 12 people. This type of expert opinion is at the bottom rung of the evidence hierarchy…”  That is true, but the North Carolinians offer no counter-estimate.  Kenkel suggests that even if the e-cigarette risk was as high as 10% as that of smoking – which is highly implausible, the result “would still be a revolutionary public health gain.”

RSO “supports regulating e-cigarettes less stringently than combustible tobacco products,” however that position is not fully reflected in their policy recommendations.  As Kenkel notes, they still call for “restrictions on flavors, advertising and marketing restrictions, taxes, bans on indoor vaping, and mandatory product warnings.” 

Using “the conceptual framework of economics,” Kenkel brings a refreshing perspective to e-cigarette use: “the key question is whether vaping and smoking are substitutes or complements… if they are complements, a policy that encourages vaping will increase smoking… If they are substitutes, a policy that encourages vaping will cause consumers to substitute vaping products for combustible cigarettes and thus reduce smoking.”

Kenkel finds that “Accumulating evidence suggests that vaping is a substitute for smoking,” and he concludes: “Policies to encourage vaping could facilitate such a revolutionary change in the U.S. market for tobacco products.”

Government and insurance incentives are worth considering, says Kenkel: “the benefits of reducing smoking could also justify a more activist policy approach. Taxing cigarettes and subsidizing a substitute good are alternative approaches to correct market and individual failures in the cigarette market. Many private and public health insurance plans provide generous coverage for one class of substitute goods—nicotine replacement products approved as smoking cessation aids. The same rationale might support subsidies for vaping products.”

Observing that vaping is “an existential threat to the market for combustible cigarettes, unless their spread is restricted by government regulation,” Kenkel seems to be in sync with new FDA Commissioner Scott Gottlieb, who has signaled a lighter approach toward e-cigarette regulation.