New England Journal of Medicine E-Cigarette Meta-Analysis is a Mega-Disaster

 

The scientific publishing community is in love with meta-analyses.  Metas apply statistical techniques to combine results from multiple studies with the aim of producing more precise estimates of specific effects.

A Medline search reveals almost 215,000 metas in the medical literature, with 75% of them published since 2015.  Of the 9,488 Medline studies involving e-cigarettes, 104 are metas.

I have written about the tsunami of deficient studies on e-cigarettes and vaping fueled by funding from the National Institutes of Health.  My colleague and I have spent considerable time over recent years documenting the flawed and downright fraudulent findings in many of these.  Unfortunately, we are now seeing an outbreak of metas that compound the distorted results of those prior studies, resulting in further obfuscation of the facts. 

One such meta-analysis is the work of retired professor Stanton Glantz and his colleagues, published in a spin-off journal, the New England Journal of Medicine – Evidence.  The article includes 181 references, 107 of which were used as sources for their risk calculations, and it runs to 18 pages, with a supplementary appendix of 117 pages. 

The main conclusion reads: “…the odds of disease between current e-cigarette and cigarette use were similar…There is a need to reassess the assumption that e-cigarette use provides substantial harm reduction across all cigarette-caused diseases.”

In short, “E-cigarettes = cigarettes = deadly = no harm reduction.”

Following publication, several issues were raised in letters posted on the journal website (here), including the small number of e-cigarette users and short duration of exposure, the lack of dose-response assessment, and two serious methodologic flaws involving assessing bias and certainty of evidence.

My colleague Nantaporn Plurphanswat and I had additional concerns.  We spent several months dissecting the meta, and we had help from Jordan Rodu, a statistics professor at the University of Virginia.  Our results have just been published (here).  In our abstract we identified “three principal deficits that were avoidable: (1) mixing unjustified and incomprehensible disease outcomes; (2) using survey datasets containing no temporal information about smoking/vaping initiation and disease diagnosis; (3) using longitudinal studies that didn’t account for changes in vaping and smoking during follow-up waves.”  Let’s take a closer look at these deficits.

We chose to focus on the Glantz meta results for e-cigarettes and cardiovascular disease (CVD), stroke and chronic obstructive pulmonary disease (COPD), because these are serious, and often fatal conditions.  When we looked at the CVD results, we were astounded to find that Glantz et al. included erectile dysfunction.  Maybe ED is associated with smoking, and it is medical problem, but there is no scientific rationale for including it in this study, but for the fact that it contributed the largest, most significant risk estimate to the CVD category. 

The meta authors made a similar error with COPD: they included a study of influenza, which isn’t remotely related. 

Medical diagnoses are not suggestions that can be categorized sloppily.  They are organized and structured by the World Health Organization in the International Classification of Diseases, and they should be employed legitimately.

We also documented that Glantz and colleagues based their results on numerous studies of cross-sectional datasets such as the National Health Interview Surveys (NHIS) and the Behavioral Risk Factor Surveillance System (BRFSS).  We have previously published research showing that these datasets contain no information about the age at initiation of e-cigarette and cigarette use, nor on the age that participants were diagnosed with diseases.  None of these studies should have been used in the Glantz meta. 

The Glantz team’s results were partially derived from longitudinal studies, mainly from the FDA’s Population Assessment of Tobacco and Health (PATH) survey, which is appropriate.  However, the only study of this type that showed a positive result for e-cigarettes and COPD was done by Xie et al. (here).  We re-analyzed that entire study, and we proved that the original results were almost entirely confounded by smoking. 

In summary, we demonstrated that the Glantz meta “failed to meet a basic criterion, described by Egger et al. as “Garbage in – garbage out?: The quality of component trials is of crucial importance: if the ‘raw material’ is flawed, the findings of reviews of this material may also be compromised.” 

Our study concluded, “the results of the [Glantz et al.] meta-analysis are invalid.”

     

Tobacco Controllers, Denying Harm Reduction Facts, Promote Orwellian Newspeak

 

Joanna Cohen is the Bloomberg Professor of Disease Prevention and the Director of the Institute for Global Tobacco Control at Johns Hopkins University’s Bloomberg School of Public Health.  Together with other editors of the prohibitionist journal Tobacco Control, she published a screed aiming to change the words people use to talk about tobacco, and particularly reduced risk products. 

Dr. Cohen, in a public email, further claimed that common, accurate terms such as “e-cigarette and heated tobacco products” serve “tobacco industry interests… We should not be doing the tobacco industry’s work for them.”

Dr. Cohen knows something about wording and misperception.  Here are the results and conclusions of a 2022 study for which she served as senior author:

“About 61.2% of smokers believe nicotine causes cancer or don’t know… High perceived threat of tobacco may be overgeneralized to nicotine…The current study supports the need for corrective messaging to address the misperception that nicotine causes cancer. Identifying that nicotine misperceptions are associated with higher harm perceptions about tobacco smoking suggests that there may be unintended consequences of high perceived harm of tobacco smoking that need to be addressed. As nicotine misperceptions are significantly more prevalent among those already at higher risk of tobacco smoking caused diseases, care should be taken to ensure equity in message dissemination.” (Cohen text, Brad Rodu corrections for accuracy)

The point of the above corrections is to underscore that Tobacco Control editors should not be advocating changes in terminology when they are guilty themselves of conflating the tobacco plant with lighting it on fire and inhaling smoke.  This is a topic I have blogged about before (here and here).

Dr. Cohen ought to direct her efforts toward the worst example of improper terminology – “e-cigarette or vaping product use-associated lung injury (EVALI).”  First coined by the CDC, this misleading formulation has been adopted by tobacco harm reduction opponents in order to obfuscate differences between vaping marijuana and vaping nicotine.  Rather than correcting their error, prohibitionists have doubled down, as seen in a recent EVALI endorsement by the American Thoracic Society (ATS).

Clive Bates explains in the following 10 paragraphs the damage caused by adoption of the EVALI fallacy:

“Truly appalling to the point of being cynical and sinister...The most notably absent feature of the [American Thoracic Society] workshop is any real recantation of the massive wave of misinformation about EVALI generated by the anti-vaping tobacco control community from 2019 to the present day, despite the fact that nicotine vaping was not (and could not have been) implicated in the 2019-20 outbreak of lung injuries in the United States (see analysis here).   Instead, there is something far worse: an attempt to redefine EVALI so that the misinformation was right all along.  

“‘EVALI: This term will be used to refer to all e-cigarette–related lung injury. This term will be used as an umbrella, as the EVALI epidemic has brought attention to e-cigarette–related health effects and is used broadly to document lung injury/disease attributable to e-cigarettes. It should be noted that the CDC does not limit EVALI diagnosis to those exposed to particular active ingredients, and use of all e-cigarettes were considered under the diagnostic criteria.’

“One wonders if this was a premise or finding of the 2021 ATS workshop or whether it has been added to the write-up retrospectively as an ex-poste justification for the flawed framing of the issue at the workshop. As we've seen on Stanton Glantz's blog and with Laura Crotty Alexander's statements, this definition is a kind of escape from accountability for the misleading attribution of EVALI (the 2019 US lung injury outbreak) to nicotine vaping.  There is nothing on the (now archived) CDC website that suggests CDC intended EVALI to refer to anything but the US lung injury outbreak and the agents and mechanisms that caused it.  

“This formulation allows a single case of an adverse respiratory reaction to nicotine e-cigarettes anywhere in the world to be classed as ‘EVALI’, and for academics/activists to say, ‘nicotine vaping causes EVALI.’ The effect of that, however, is to load the negative perceptions associated with the US outbreak of lung injuries (a large outbreak of 2,800 hospitalisations and 68 deaths) into risk perceptions about nicotine vaping, which did not cause these. It is deeply unethical and misleading to do this. Scientists should be trying to clear up misunderstandings and confusions (many of which they have created or amplified), not adding to them through unwarranted conflations. 

“They could try to justify this on the basis that ‘we are scientists, and we can use whatever terminology we like, as long as we are clear about it.’ However, it would still be unethical because the predictable consequence (whether unintentional or deliberate) will be a conflation of radically different risks and, therefore, the promotion of misunderstanding. Just like saying, ‘all tobacco products are harmful,’ but worse. Scientists behaving ethically would actively take care not to do this and to dismantle the confusions that they had previously worked so hard to promulgate.  There is perfectly good language available to describe effects other than the EVALI caused by THC-VEA for the extremely rare (and usually contested) cases where nicotine vaping may be implicated in adverse pulmonary reactions, for example, in people with allergies or pre-existing conditions.

“The inclusion of this is designed to increase the confusion and inappropriate conflation: 

‘Although most affected individuals reported use of cannabinoid e-cigarettes, approximately 20% reported using only nicotine e-cigarettes [citation removed]. It is unknown whether these patients were unintentionally exposed to VEA through cross-contamination of e-liquids or sharing of e-cigarettes or whether additional ingredients, such as medium-chain triglycerides (MCT), can lead to EVALI  [citation removed].’

“We need to be absolutely clear here. This 20% of cases were not and cannot have been caused by nicotine vaping. That is just inconsistent with the epidemiological evidence (the outbreak was confined in time and place and ended without any changes made to nicotine vaping products. the causal agent VEA cannot be added to nicotine e-liquids and would serve no purpose if it could be).  The reason people deny using THC is that illicit drug use presents problems with law enforcement, parole officers, employers, colleagues, schools and parents. How obvious do things need to be before they are understood by tobacco control academics?  It is also possible that some people were sold fake THC vapes with no active ingredients, just the cutting agent. So the statement above is pure merchant-of-doubt. 

“Personally, I suspect at least some of the authors are doing this deliberately. Firstly to cover their tracks over the massive misinformation sprayed over the public in 2019-20, but more insidiously because they want these products to be seen as harmful to deter use of them.  Further, they need nicotine vaping to be harmful, or they lose the rationale for control (the purpose of tobacco control - it's in the name) and, therefore, their reason to exist.  

“There are two ways to look at this: a disgraceful cynical, defensive fear play or so naive as to be negligent. Sorry, no nice way to put it.”

George Orwell, in his seminal book “1984”, coined the term “newspeak”, defined today by Merriam-Webster as propagandistic language marked by euphemism, circumlocution, and the inversion of customary meanings. George, meet Dr. Cohen and her fellow-travelers.

 

Stanton Glantz’s Tainted Science: The Rest of the Story

 

Journalist Marc Gunther recently authored articles for Undark (here) and Medium (here) about “The tainted science of Stanton Glantz.” 

It is evident that Gunther interviewed or corresponded with numerous Glantz research critics. Gunther’s sources include the author of this blog, who provided extensive documentation of Glantz’s questionable work. However, since Gunther fails to tell the whole story, I will provide the rest of it here.

Fake Heart Attacks

The most important items in Gunther’s articles are his descriptions of Glantz’s flawed heart attack study that appeared in the Journal of the American Heart Association – a work that was retracted eight months after its publication in June 2019. Gunther described that study and its repercussions in his Undark article as follows:

“…when describing the second study, published in 2019 in the Journal of the American Heart Association, Glantz said it provided ‘more evidence that e-cigs cause heart attacks.’… Critics pounced on what they called glaring flaws in the analyses. Some of the e-cigarette users had previously smoked, for example, muddying the correlation. Brad Rodu, a University of Louisville professor who has numerous and longstanding connections to the tobacco industry, dug into the raw data and found that at least 11 of the 38 heart-attack victims cited in the Journal of the American Heart Association study had suffered their heart attacks before they started vaping — some as many as 10 years before. Glantz was made aware of the temporality problem before publication because it was raised by a peer reviewer, the journal’s editor subsequently realized.

 

“Sixteen tobacco researchers wrote to the journal editor asking for a retraction, and the Journal of the American Heart Association ultimately did just that — something it has done only a handful of times in its history. Its editor, though, was careful to state in a letter to Glantz that ‘the retraction notice is intentionally absent of any language suggesting scientific misconduct.’

 

“Andrew Gelman, a professor of statistics at Columbia University who followed the controversy on his blog, was unimpressed with Glantz’s response to the retraction, calling it ‘anti-scientific.’ He wrote: ‘If someone points out an error in your work, you should correct the error and thank the person. Not attack and try to salvage your position with procedural arguments.’”  

Gunther is correct: I had discovered that Glantz deliberately deceived the journal’s editors and readers by counting heart attacks before people had vaped. My comments on this appeared in a USA Today article by Jayne O’Donnell on July 17, 2019. O’Donnell wrote, “However, when Rodu obtained the federal data, he found the majority of the 38 patients in the study who had heart attacks had them before they started vaping — by an average of 10 years earlier. In his letter to the editors [dated July 11, 2019], Rodu called Glantz's findings ‘false and invalid…Their analysis was an indefensible breach of any reasonable standard for research on association or causation…We urge you to take appropriate action on this article, including retraction.’” (emphasis added) 

As noted, I publicly called for a retraction on July 17, 2019. But Gunther credited the retraction to “sixteen tobacco researchers,” linking to a letter they wrote to the journal editor on January 20, 2020. In the USA Today article I also called for a federal investigation, and I followed up a year later with a full analysis of the fraudulent findings, which was published in the journal Addiction (discussed here). 

Gunther correctly cited influential statistician Andrew Gelman. I had corresponded with Gelman in 2018 about Glantz’s woefully defective Pediatrics study, which also elicited a retraction demand from me (here, here, here and here). Gelman’s account of our correspondence is here (see Episode 1). Three months before Pediatrics retracted the study, Gelman ran the analysis of it that I had recommended to researchers world-wide in November 2019. 

In his Medium article, Gunther identifies a group of “respected veterans of the anti-smoking movement (Steven Schroeder, Ken Warner, David Abrams, Raymond Niaura and [Michael] Siegel),” and he provided links to their university profiles. Later in the article, he attributes the JAHA retraction to “other scholars”, followed by a comment from me, who he characterizes again with “whose work has been supported by the tobacco industry.” Gunther makes no mention of my 27 years of research into tobacco harm reduction, and provides no link to my university webpage profiles (here and here). 

Perhaps Gunther was influenced by the established (anti-) tobacco research and policy community, some of whom switched to support safer products only after e-cigarettes became popular. Gunther writes that they tolerated Glantz’s defective studies.

“‘Stan has always been an advocate and ideologue willing to twist the science,’ says David Abrams, a New York University professor and veteran tobacco researcher. He says that some scientists ignored flaws in his work when Glantz focused on combustible tobacco because they, too, strongly opposed smoking. ‘Frankly, none of us cared if he was a little bit sloppy with his research because the ends justified the means,’ Abrams says.” 

I never tolerated Glantz’s flaws and sloppy research, and I published numerous critical comments in journal forums and in my blog starting in 2004. Gunther mentioned two examples but did not acknowledge the active role I played in correcting Glantz’s record.

Fake Teen Gateway

Gunther’s biggest error is his description of Glantz’s “2018 study in Pediatrics, [which] also claimed that e-cigarette usage encourages more young people to smoke…a gateway effect, but the alleged link between vaping and smoking disappeared when other teen behaviors, such as using marijuana, were taken into account.”

I replicated that study’s analysis and proved that Glantz’s team had fabricated false results, which I described in detail in a letter to the editors that included a call for retraction (here). That led to an extended exchange with Glantz and Pediatrics editors, with the latter going to great lengths to defend the flawed research (here, here and here).   

The “Helena Miracle”

Gunther noted the “Helena miracle”, in which Glantz credited public smoking bans for a decrease in hospital admissions for heart attacks in the small Montana town. Gunther wrote, “The small sample size in Helena — four cases per month during the ban, compared to seven beforehand — should have raised red flags; random fluctuations could have explained the drop in hospital admissions.” As I had told Gunther, bright red flags were raised by me in comments to the BMJ in 2004 (here) and 2006 (here). My analysis showed that the number of hospital admissions in Helena and other miracle localities was so small that “the relevant question is whether [the Helena] report involves anything more than random variation.” (here)

I attempted to reproduce Glantz’s tiny city reports by conducting an analysis using state-wide data. My study, published in 2011 (discussed here), was subsequently referenced by other researchers (here). I found that rates of death due to heart attacks in states with smoking bans were no different than those in states with no ban, and that heart attacks had been declining everywhere for years. The last point is something Glantz never took into account; in fact, he freeloaded his second-hand smoke heart-attack results on the decades-long downward trend in heart attack rates.

Fake “Softening” of the Smoking Population

In 2015, Glantz proclaimed in Tobacco Control that there is no public health basis for advising smokers that smokeless tobacco and e-cigarettes are safer cigarette alternatives, because the smoking population in the U.S. was “softening”, i.e., becoming more likely to quit (abstract here). Once again, I found that his analysis was seriously flawed, as he failed to consider a number of other significant factors. My research group recreated his analysis and took into account the missing factors; we found that Glantz’s “softening” disappeared.  We published our study in the journal Addiction (discussed here). 

Other Glantz Publication Problems

Glantz Redefined Youth Smoking, 2014  https://rodutobaccotruth.blogspot.com/2014/04/ucsf-redefines-youth-smoking-journals.html

Glantz Falsely Linked E-Cigarettes to Smoking, 2014 https://rodutobaccotruth.blogspot.com/2014/03/ucsf-study-falsely-links-e-cigarettes.html

Glantz Misrepresented IQOS Studies to the FDA, 2017 https://rodutobaccotruth.blogspot.com/2017/11/smoke-but-no-fire-iqos-opponent.html

Glantz Used Complicated Models to Try to Change NYTS Data, 2017 https://rodutobaccotruth.blogspot.com/2017/02/complicated-models-cant-alter-data.html

https://rodutobaccotruth.blogspot.com/2017/02/complicated-models-cant-alter-data-part.html

Glantz’s Failed Attempt to Link Teen Smoking to Cinematic Smoking, 2017 https://rodutobaccotruth.blogspot.com/2017/08/teen-smoking-unconnected-to-cinematic_30.html

Another E-Cigarette Gateway Claim by Glantz Based on Tiny Numbers, 2018 https://rodutobaccotruth.blogspot.com/2018/04/another-uc-san-francisco-e-cigarette.html

Trace Toxins in Teens Improperly Blamed on E-Cigarettes, 2018 https://rodutobaccotruth.blogspot.com/2018/03/beyond-headlines-trace-toxins-present.html

Federal Funds Misspent on Glantz’s Research, 2020 https://rodutobaccotruth.blogspot.com/2020/02/federal-funds-misspent-on-anti-vaping.html

 

Since 2006, the National Institutes of Health has funneled $51 million to Glantz at the University of California San Francisco. In addition to the flawed heart attack work, which may be viewed as research misconduct using federal funds, he published 300 other articles. I list above only the worst offenders and note my attempts to correct them.