As Imre Lakatos so eloquently stated in 1973, “Blind commitment to a theory is not an intellectual virtue: it is an intellectual crime.” Indeed, blind commitment to a bad theory in the public policy realm will have long lasting impacts, negatively affecting the pursuit of improved public health.

That is why it baffles me when otherwise smart researchers remain committed to bad theories. Indeed, the theory that e-cigarette use will lead to cigarette smoking is a bad theory.

Hypotheses based on the gateway effect are rarely supported, and a rudimentary examination of what would be required to show that certain substances are gateways indicate that it’s nearly impossible to do so. What we do know is that some people are going to do risky things – and we can’t always predict the trajectory of the risk behaviors in which they will choose to engage.

There are ways to minimize the initiation of risky behaviors, but prohibiting one product in hopes of decreasing the use of another should not be one of them. At best, doing so directs time and energy away from a better solution. At worst, people who would otherwise benefit from the reduced harm posed by the alleged gateway product – like e-cigarettes – are put at risk of relapsing to a more dangerous product – like their combustible counterparts.

Unfortunately, we are not deprived of misleading research that threatens to steer users away from reduced-harm products, like e-cigarettes, and toward readily available yet more dangerous tobacco products. A slew of studies have recently concluded that e-cigarette use is a gateway to combustible cigarette use in teens, and one study in particular concludes that kids who use e-cigarettes are seven times more likely to use combustible cigarettes when compared to kids who don’t use e-cigarettes.

On first glance, this shocking number would lead anyone to believe that the availability of e-cigarettes causes kids to graduate to combustibles. What’s missing from this conclusion – and what the data really suggest – is that kids experiment.

Over the course of the two-year study, the strength of the association between e-cigarettes and combustible cigarettes decreased, meaning that in the second phase of the study the odds that e-cigarette use would lead to cigarette use was half that than in the first phase. Also noteworthy is that cigarette use at each time point remained the highest predictor of future cigarette use throughout the study. While e-cigarette use is associated with cigarette use, this study design cannot demonstrate that e-cigarette use causes cigarette use.

What no study can capture – and this study is no exception – is the population of youth that would smoke combustible cigarettes regardless of e-cigarette availability. Therefore, no study can demonstrate that e-cigarette use is necessary or sufficient to future combustible cigarette use. Establishing the gateway theory would require proof of both statements.

The truth is, we cannot define e-cigarettes as a gateway to combustible cigarette use. In fact, population-level data provide evidence directly contradicting the gateway hypothesis: While e-cigarette experimentation tripled between 2013 and 2014, combustible use decreased by 27 percent between 2013 and 2015. In 2016, combustible use fell another 8 percent. If e-cigarettes were indeed a gateway, we would expect parallel increases in combustible smoking prevalence in the following years, not the other way around.

Furthermore, as it stands now, cigarette smoking among high school adolescents is at an all-time low of 13 percent, down from 17 percent in previous years. E-cigarette use is also down, falling to 11 percent from 16 percent in previous years. Youth tobacco use has never been this low.

It is incumbent on us, as public health advocates, to take the totality of data – from population studies, toxicology studies and physiological studies – and determine whether an intervention will result in the desired effect – improving public health. We would certainly fail at our goal if we were to ignore data showing that e-cigarette use increases smoking behavior in teens.

However, when the overwhelming majority of research does not support such a hypothesis, and the research that does draws weak conclusions, it would be egregiously irresponsible for us to craft public policies against the weight of evidence. We simply cannot let bad science in service of bad theories support public policies.

Featured Publications