A Medical Journal Retracts a 2022 Study That Linked Vaping to Cancer

The obvious problems with the article reflect a broader pattern that suggests a peer review bias against e-cigarettes.


The World Journal of Oncology recently retracted a February 2022 article claiming that nicotine vapers face about the same cancer risk as cigarette smokers. "After publication of this article," the editors explain, "concerns have been raised regarding the article's methodology, source data processing including statistical analysis, and reliability of conclusions." Because "the authors failed to provide justified explanations and evidence for the inquires [sic], subsequently this article has been retracted at the request of Editor-in-Chief."

Some of the concerns raised by this article are similar to the problems with other studies that have linked vaping to smoking-related diseases. Most conspicuously, this study failed to address the question of whether diagnoses were made before or after people started vaping, a minimum requirement for inferring causation. In 2020, the same problem led to the retraction of a Journal of the American Heart Association article that reported an association between vaping and heart attacks.

The World Journal of Oncology article—which was attributed to no fewer than 13 researchers at institutions such as the University of Missouri, Temple University Hospital, the Mayo Clinic, and the Icahn School of Medicine at Mount Sinai—has other obvious problems that should have been apparent before publication. It features enough inconsistencies, writing errors, non sequiturs, and failures of reasoning to make you wonder whether peer reviewers and editors actually read it, let alone carefully evaluated its strengths and weaknesses.

As critics have noted, the publication of such studies suggests that the peer review process is biased against vaping, favoring articles that highlight its potential hazards even when the science underlying them is weak. In an email, Brad Rodu, a University of Louisville professor of medicine who has been studying tobacco harm reduction for decades, says the "grossly flawed" study of vaping and cancer raises a troubling question: "How could it get through peer review?"

In the retracted study, University of Illinois internist Anusha Chidharla and her 12 co-authors analyzed data from the National Health and Nutrition Examination Survey. The sample included 154,856 respondents surveyed from 2015 through 2018, of whom 5 percent reported that they had ever used e-cigarettes, 31.4 percent said they were current smokers, and 63.6 percent said they did not smoke and had never used e-cigarettes. The survey also asked whether participants had ever been diagnosed with cancer.

Crucially, the study does not include information on when the e-cigarette users began vaping. But the authors note that "e-cigarettes [were] used as a strategy to quit smoking in most cancer respondents," which suggests that their diagnoses generally preceded their e-cigarette use. If so, that would be consistent with what Rodu and University of Louisville research economist Nantaporn Plurphanswat found when they analyzed data on other smoking-related diseases from the Population Assessment of Tobacco and Health Survey, which includes information on the timing of both diagnoses and e-cigarette use.

Chidharla et al. classified participants as "e-cigarette users" if they had ever vaped and were not current smokers. The researchers did not take into account whether the respondents in that group had a history of smoking, which is obviously problematic when you are trying to distinguish between correlation and causation.

"The authors reclassified former smokers as nonsmokers, thereby obscuring the effects of 'former,'" Rodu notes. "That also raised the cancers in their reference group, which was inappropriate."

Keeping those points in mind, what did the researchers find? They seemed confused about that.

According to the abstract, "the e-cigarette users [had] lower prevalence of cancer compared to traditional smoking (2.3% vs. 16.8%; P < 0.0001)." That is consistent with the numbers reported in Table 2. But according to the "Results" section of the article, "respondents with cancer [had] a lower prevalence of e-cigarette [use] compared to traditional smoking (2.3% vs. 16.8%; P < 0.0001)." The "Discussion" section reiterates that "cancer respondents had a lower prevalence of e-cigarette use than traditional smoking (2.3% vs. 16.8%)."

All those passages cite exactly the same numbers, but they are talking about two different things: prevalence of cancer among e-cigarette users and smokers (the study's "secondary aim") vs. prevalence of e-cigarette use and smoking among people diagnosed with cancer (the study's "primary aim"). "I have made a serious attempt," Rodu says, "but I cannot figure out how the authors switched from cancer prevalence among e-cig users to e-cig prevalence among participants with cancer." It seems neither the peer reviewers nor the journal's editors noticed that inconsistency prior to publication.

Assuming the version of the results presented in the abstract and Table 2 is the correct one, 2.3 percent of e-cigarette users reported cancer diagnoses, compared to 16.8 percent of current smokers and 9.5 percent of the nonsmokers. But when the researchers ran a regression analysis that included several demographic variables and "comorbidities" (including other diagnoses and use of other drugs), they calculated that e-cigarette users "had 2.2 times higher risk and traditional smokers had 1.96 times higher risk of having cancer compared to non-smokers."

In other words, the prevalence of cancer among the e-cigarette users was about one-quarter the prevalence among nonsmokers and one-seventh the prevalence among current smokers. But after the regression analysis, the risk for e-cigarette users was about the same as the risk for smokers—i.e., roughly twice the risk for nonsmokers.

Since the retraction mentions "concerns" about "source data processing including statistical analysis," we can surmise that the editors, after taking a closer look at the study, questioned that calculation. In any event, the lack of information about the timing of e-cigarette use makes it impossible to draw causal conclusions from whatever correlations these survey data do support. Or as Chidharla et al. put it, "causal or temporal association could not be established."

Despite that concession, the authors' conclusion assumes a risk they have not proven. "Our study found e-cigarette users had an early age of cancer onset as well as higher odds of having cancer compared to non-smokers," they write. "Prospective studies should be planned to mitigate the risk."

In light of the study's fundamental weaknesses, what value would it have had even if the "source data processing including statistical analysis" had been sound? "Despite the limitations," the authors say, "to our knowledge, this is the first large population-based study to find [a] potential association between e-cigarette use and cancer in humans." So there's that.

Stanton Glantz, the American Legacy Foundation Distinguished Professor of Tobacco Control at the University of California, San Francisco, thought that was good enough. In a blog post, he hailed the "first epidemiological evidence linking e-cigs to cancer in people." Glantz, who co-authored the retracted Journal of the American Heart Association article alleging a connection between vaping and heart attacks, said Chidharla et al.'s study provided "direct evidence that people who use e-cigarettes are at increased risk of some cancers" (emphasis his). He added that "e-cigarette use was associated with higher risks of some cancers than smoking cigarettes, including cervical cancer, leukemia, skin cancer (non-melanoma), skin (other) and thyroid cancers" (again, emphasis his).

For those of us who are less eager to undermine the case for vaping as a harm-reducing alternative to smoking, what would this study have meant if it had not been retracted? Chidharla et al. were not sure.

If you ignore the glaring methodological weaknesses, the study suggests that vapers and smokers face similar cancer risks. Yet the authors note that "vaporized nicotine emissions from e-cigarettes contain carcinogens generally in lower concentrations with cancer potencies < 1% that of tobacco smoke." In other words, the carcinogenic potency of e-cigarette aerosol is more than 99 percent lower than the carcinogenic potency of cigarette smoke. The researchers add that "mean lifetime cancer risks decline from traditional smoking to e-cigarettes."

Chidharla et al. say the "exponential increase in the use of e-cigarettes due to their widespread promotion as safer alternatives to traditional smoking" is a "dangerous threat" and a "public health risk." But they also describe e-cigarettes as a promising harm reduction tool.

"Although smoking in any form is never safe, e-cigarettes can be recommended by clinicians as an alternative to traditional smoking in populations with a history of cancer who would otherwise continue to smoke or those who want to start smoking at all cost," the authors write. "This could dramatically decrease the risk of serious disease in nicotine users and other high-risk groups."

That is exactly the point of harm reduction. A dramatic decrease in health risks among people who otherwise would be smoking is unambiguously an improvement. So why do Chidharla et al. seem ambivalent at best about products that help people achieve that outcome?

"Due to higher prevalence of certain types of cancers in e-cigarette [users] and
unknown consequences of e-cigarette use, more guidelines are needed regarding the use of e-cigarettes and their association with cancer," the authors write. "E-cigarette[s] should not be considered as a safe alternative to dual or traditional smoking without stronger clinical evidence on [their] safety." But to reduce health risks, e-cigarettes do not have to be "a safe alternative"; they need only be a safer alternative, which Chidharla et al. concede they are.

The authors sometimes seem desperate to obscure that point. "The recent outbreak of e-cigarette vaping-associated lung injury (EVALI) in the USA suggests caution," they write. "EVALI is primarily attributable to vitamin E acetate in cannabis oils distributed through illicit channels [emphasis added]." What does that have to do with the risks posed by the legally distributed nicotine products the article is ostensibly discussing?

Chidharla et al. worry that e-cigarettes are "gaining popularity" among "never-smokers and adolescents." They add that "the potential for negative health effects from exposure to nicotine or other chemicals in e-cigarettes among non-smokers is concerning."

Adolescent vaping actually has been falling in recent years. And according to a survey the researchers cite, never-smokers account for less than 9 percent of Americans who report that they have tried e-cigarettes. A more recent survey found that less than 3 percent of Minnesota never-smokers reported "current" e-cigarette use, meaning they used e-cigarettes "every day or some days."

Chidharla et al. also worry that vaping products "are commonly marketed as a safe alternative" when "the long-term effect of e-cigarettes is not known yet." But as they concede, the evidence indicates that vaping, even if it may pose some long-term risks, is far less hazardous than smoking.

Surveys suggest that Americans generally do not understand that, thanks largely to deliberate obfuscation by anti-smoking activists and public health officials. When it comes to public perceptions, the problem is not that people mistakenly think vaping is completely risk-free; the problem is that less than 3 percent of Americans recognize that e-cigarettes are "much less harmful than combustible cigarettes."

Brian King, director of the Food and Drug Administration's Center for Tobacco Products, acknowledges the gap between what the evidence shows and what Americans commonly think. "I'm fully aware of the misperceptions that are out there and aren't consistent with the known science," he told the Associated Press in September. "We do know that e-cigarettes, as a general class, have markedly less risk than a combustible cigarette product."

Since those "misperceptions" discourage smokers from switching to vaping, they are a significant obstacle to reducing smoking-related disease and death. Ill-conceived, poorly reasoned studies like this one compound that problem.