New Study Purports to Show that E-Cigs are a Gateway to Smoking, But Provides No Evidence to Support that Conclusion

A new study published yesterday in the journal Tobacco Control purports to show that e-cigarettes are a gateway to smoking. The study followed 1,136 nonsmoking youth in grades 9 to 11 in Hawaii for one year to determine which youth initiated smoking during this follow-up period. Baseline surveys were conducted in 2013, and follow-up surveys were conducted in 2014. Rates of smoking initiation were compared between youths who had used e-cigarettes at baseline and those who had not used e-cigarettes. The study found that youths who had used e-cigarettes were significantly more likely to initiate smoking and that this association was stronger for youth who had a lower propensity to start smoking. Based on that finding, it concluded that e-cigarettes are a gateway to smoking.

Propensity to smoking was assessed by measuring rebelliousness, parental support, and intentions to smoke.

The study concludes as follows: "This study provides evidence that ecigarettes are recruiting lower risk adolescents to smoking, which has public health implications."

The Rest of the Story

The rest of the story is that this study does not actually provide evidence that e-cigarettes are recruiting youths to smoking. What it shows is that experimentation with e-cigarettes is a much more sensitive marker of propensity to smoking than simply assessing rebelliousness or parental support or intentions to smoke.

The major problem with the study is that it does not examine whether or not youth who experiment with e-cigarettes become addicted to vaping and then progress to smoking, which would be indicative of a gateway effect. Instead, it compares youth who have simply tried e-cigarettes with youth who never even had a puff of an e-cigarette. The study does not document that any of the youth who experimented with e-cigarettes actually became regular vapers or that they were addicted to vaping before progressing to smoking.

In fact, the paper hides a critical piece of information: although 29% of the high school students in the sample had experimented with e-cigarettes, only 2% were using them more than once a week. Therefore, at least 98% of the e-cigarette experimenters were clearly not addicted to vaping and could not be considered regular vapers.

The key problem is that the study did not determine the trajectory of e-cigarette use prior to the initiation of smoking. It is entirely possible that many of the youth who had experimented with e-cigarettes actually failed to take up vaping and that because of this failure, they started smoking. Thus, the study results are entirely consistent with the conclusion that vaping is protective against smoking initiation.

There is an alternative explanation for why youth with a lower propensity to smoke who experimented with e-cigarettes were more likely to start smoking. Namely, what this shows is that e-cigarette experimentation is a much more sensitive marker of propensity to smoke than the measure used in the study, which relied on measures of rebelliousness, parental support, and future intentions to smoke.

In fact, if you look carefully at Figure 1, you'll see that among youth who used e-cigarettes at least 5 times at baseline, propensity to smoke failed to predict smoking initiation risk. Having experimented with e-cigarettes was a much better marker of propensity to smoke than the propensity measure actually used in the study.

Let me provide an analogy to help explain this. Suppose that you were interested in studying whether using marijuana is a gateway to smoking. You do a study showing that nonsmoking youth who experiment with marijuana a few times are more likely to have initiated smoking one year later. Does this provide evidence that marijuana is a gateway to smoking?

I would argue that it doesn't. There is another plausible explanation for the study results, which is that marijuana experimentation does not cause youth to smoke, but merely is a marker for a risk-taking personality trait that predisposes to smoking.

And if we were to find that the relationship between marijuana experimentation and smoking initiation were stronger among youth who had a lower propensity to smoke, would that demonstrate that the observed relationship is because marijuana use causes kids to smoke, rather than that it is a marker for an increased baseline propensity to smoke?  No, it would not. It could be that experimentation is just a much more sensitive marker of propensity to smoke because it shows whether or not a youth has actually taken up a drug-related risk-taking behavior.

If a youth has taken up a drug-related risk-taking behavior, how could it possibly be that this youth is not more likely to also take up another drug-related risk-taking behavior?

In other words, it would be absolutely shocking if we found that youth who try e-cigarettes are not more likely to initiate smoking. I would have predicted the results found in this new study before the study was even conducted, and I would have bet my professional reputation on the fact that the study would indeed find an increased risk of smoking among youth who had experimented with e-cigarettes and that this risk would be greater among youth with a lower propensity to start smoking as measured by less definitive indicators such as parental control and general level of rebelliousness, or even stated intention to smoke.

To be clear, the rest of the story is that this new study provides no evidence that e-cigarettes are a gateway to smoking. Instead, it confirms that actual drug-related risk-taking behavior is a much better predictor of other drug-related risk-taking behavior than simply asking a kid if he thinks he will try another drug in the future or asking a kid how rebellious he is or how much his parents support him.

What would it take to actually provide evidence that marijuana use is a gateway to smoking?

Two things.

First, one would have to examine the trajectory of marijuana use. If youth who experiment with a few joints then go on to become regular marijuana users, and then over time progress to "harder" drugs like cigarettes, that would provide some evidence of a gateway effect of marijuana. But the present study presents no evidence that e-cigarette experimenters go on to become regular vapers, and then over time progress to cigarettes.

Second, one would have to provide at least some evidence of a mechanism by which the use of marijuana creates changes that then causally lead to smoking.

In the case of marijuana, there is some evidence for such an effect. As explained by the National Institute on Drug Abuse: "Early exposure to cannabinoids in adolescent rodents decreases the reactivity of brain dopamine reward centers later in adulthood.⁴⁶ To the extent that these findings generalize to humans, this could help explain the increased vulnerability for addiction to other substances of abuse later in life that most epidemiological studies have reported for people who begin marijuana use early in life.⁴⁷ It is also consistent with animal experiments showing THC's ability to "prime" the brain for enhanced responses to other drugs.⁴⁸ For example, rats previously administered THC show heightened behavioral response not only when further exposed to THC but also when exposed to other drugs such as morphine—a phenomenon called cross-sensitization.⁴⁹"

In the case of e-cigarette experimentation, there is no evidence of a mechanism by which use of a highly flavored and sweet-tasting product would lead youth to be dissatisfied and craving for a harsher and less tolerable alternative. In fact, conceptually, there is evidence that regular vaping is likely to be protective against smoking initiation. The existing evidence suggests that most youth vapers are doing so because they enjoy the flavorings, rather than because they are deriving pleasure from any kind of nicotine hit or reward. The Monitoring the Future survey revealed that the majority of youth who use e-cigarettes report that they vape flavored e-liquids that do not contain nicotine. It appears, then, that nicotine is a minimal if not non-existent reason for youth e-cigarette experimentation. Thus, the hypothesis that youth use e-cigarettes, become addicted to the nicotine, are not satisfied enough with vaping and feel the need for something more hard-core, and therefore progress to cigarette smoking is dubious at best, and at worst, is a little on the ridiculous side.

At very least, it would necessary to demonstrate that a significant number of youth who experiment with e-cigarettes are in fact addicted to nicotine. So far, I'm not aware of a single study which has demonstrated that a single youth vaper is addicted to nicotine. We're not even sure that a substantial proportion of youth who vape are using nicotine-containing products in the first place. Nor is there any evidence that the delivery of nicotine from the products that youth are using is sufficient to create and sustain nicotine addiction. In fact, studies of nicotine delivery indicate that with most e-cigarettes, it is not conducive to creating addiction because there is no significant spike in blood nicotine levels such as to create a "hit" or a psychological reward that would lend itself to a propensity for addictive behavior.

There is a second problem with the study which deserves mention, although even without this second problem the study conclusion would be invalid.

This second problem is that smoking initiation was measured by any experimentation with cigarettes, even if just a puff. So many of the youth in the sample may have puffed on a single e-cigarette at baseline and then puffed on a single cigarette some time over the next year and that would be considered as providing evidence that e-cigarettes are a "gateway" to smoking. I provide this example only to show how flimsy the study conclusions are. In reality, I would actually expect to see higher rates of actual smoking initiation and sustained, regular smoking among youth who at baseline had experimented with e-cigarettes, and that would not provide evidence of a gateway effect. In fact, such a finding would be entirely consistent with the hypothesis that it is the failure of youth to become regular vapers that predisposes them to try smoking. It is still entirely possible that youth who become regular vapers are somewhat immune from smoking because it becomes very difficult to switch from a sweet, flavored product to a harsh, real cigarette like a Marlboro.

Although I disagree with the study conclusions, I do want to note that I agree with the overall study recommendations, which include providing education to youth about the potential health risks of both vaping and smoking, taking efforts to keep e-cigarettes out of the hands of youth, and restricting marketing of vaping products to youth. These actions should be taken regardless of whether e-cigarettes are a gateway to smoking or not.

Nonetheless, I'm afraid this study is going to do significant damage because anti-smoking groups are going to use this to convince the public and policy makers that e-cigarettes are a gateway to smoking, and this is going to lead to the continued promulgation of irrational policies that promote cigarette smoking by discouraging vaping. The media are already reporting this study as having proved that e-cigarettes are a gateway to smoking. It's almost impossible to get the cat back in the bag. Even if people now know "the rest of the story."