Fundamentally flawed

| May 1, 2018

The Chaffee vaping ‘gateway’ study should be retracted.

By Brad Rodu

Brad Rodu is a professor of medicine and holds an endowed chair in tobacco harm reduction research at the University of Louisville in Louisville, Kentucky, USA..

Despite a record decline in cigarette use among teens over the past five years, government officials and federally funded researchers have repeatedly claimed that e-cigarette use is creating a new epidemic of teen smoking. The allegations are based mainly on misrepresentations of federal survey findings and inaccurate medical journal claims that vaping is a gateway to smoking.

In some instances, published research is so fundamentally flawed it deserves retraction. A case in point is an article in Pediatrics by the University of California San Francisco’s (UCSF) Benjamin W. Chaffee, Shannon Lea Watkins and Stanton A. Glantz.

A 2013 survey of teens was funded by the U.S. Food and Drug Administration to find out which factors contributed to regular smoking one year later. Since experience with e-cigarettes is a relevant factor for smoking, teens were classified in 2013 as “never users,” “ever triers” or “current users” (i.e., at least one puff in the past 30 days).

Analyzing the survey data, the UCSF researchers concluded that teen e-cigarette users or triers in 2013 were much more likely to be regular smokers one year later—in short, e-cigarettes were a gateway to cigarettes.

This analysis is incorrect. Chaffee and his colleagues ignored the number of combustible cigarettes teens had smoked prior to the 2013 survey (their lifetime cigarette consumption, or LCC), which varied widely, from one puff but never a whole cigarette all the way to 99 cigarettes.


It turns out that e-cigarette triers and users in 2013 had smoked a lot more cigarettes than never e-cigarette users, as illustrated in the chart. Note that more never e-cigarette users had only puffed (i.e., were in the “blue”), while more e-cigarette users and triers had smoked a half pack or more (i.e., were in the “red”).

Using the survey data, my research team re-ran Chaffee’s calculations. When we added the respondents’ LCC, the positive results for e-cigarettes and the grounds for Chaffee’s claim essentially disappeared.

In short, Chaffee’s claim that using e-cigarettes is a gateway to smoking only stands if one ignores prior smoking of up to 99 cigarettes per teen.

The relationship of one factor, like e-cigarette use, to another, such as smoking, is called confounding—a basic epidemiologic principle that competent researchers understand and address in their analyses. Chaffee and his colleagues failed to account for the important confounding factor of LCC. When we documented this in an online journal comment, Chaffee et al. incorrectly described our analysis as a “statistical trick … akin to suggesting that a study of hypertension [high blood pressure] should adjust for confounding by systolic blood pressure.”

Their response further proved our case.

Consider the analogy of a study of the effect of dietary salt consumption on development of hypertension after one year of follow-up among participants who were not hypertensive at baseline. Knowing that those with blood pressure close to hypertension at baseline are far more likely to subsequently become hypertensive, compared to those with low blood pressure at baseline, it would be improper to blame salt for causing hypertension while ignoring baseline blood pressure readings.

Similarly, it is improper to ignore cigarette consumption at baseline and then to blame e-cigarette use for smoking one year later. The Chaffee article must be retracted.








Category: Also in TR, Editorial Archives

Comments are closed.