Jacques Le Houezec, Konstantinos Farsalinos, Jean-François Etter and Riccardo Polosa were on stage at Vapexpo 2016 to highlight current trends in e-cigarette research.

From left to right: Jacques Le Houezec (France), Konstantinos Farsalinos (Greece), Jean-François Etter (Switzerland), Riccardo Polosa (Italy).
From left to right: Jacques Le Houezec (France), Konstantinos Farsalinos (Greece), Jean-François Etter (Switzerland), Riccardo Polosa (Italy).

The Gateway Theory

If policy makers understood this, they would be less inclined to restrict access to vaping products.Jean-François Etter

Professor Jean-François Etter who is a specialist in political Sciences, addresses the gateway theory through and reviews what is needed to support this causality. Ultimately, he examines if science has the appropriate studies to approve it or to reject the gateway effect of vaping to smoking.

The origins

The Gateway Theory is saying that vaping among young people who don’t smoke will cause smoking in this population, later. It is a theory of causality and very hard to prove.Jean-François Etter

The Gateway theory originated in the 1970s, at a time when heroin consumption was increasing. It was a descriptive theory, most people thought that the use of marijuana led to the use of heroin, in a second step. This belief was a mix of scientific, political, popular and media theories of causality between marijuana and heroin use. It has still not been established, actually.

The essence of the gateway theory relies on 3 elements:
sequence of substance use (substance A is used first, B follows),
increased risk of subsequent use (when A is used, the risk of using A increases),
dose-response relationship (the more A is used, the more is the risk of using B).

In the domain of vaping, this theory has an enormous political influence in the development of the regulatory approaches. Appealing effects to the youth and gateway effects are some of the major concerns of politicians who have been asked to regulate the use of the e-cigarette, in Europe and worldwide.

“Electronic cigarettes can develop into a gateway to nicotine addiction and ultimately traditional tobacco consumption […]. For this reason, it is appropriate to adopt a restrictive approach”EU TPD

Vaping and smoking, a different description in scientific literature

Source: J.-F. Etter, Vapexpo 2016

In order to establish this theory, researchers first have to define the behaviors of vaping and smoking.

In most of the scientific literature the use of an e-cigarette is not satisfactorily described as a measure of vaping since a single episode is not enough to say that someone is vaping. The same applies for past 30-day use that is also a frequent proxy for vaping and not a good predictive variable, except for experimentation.

Finally, nicotine content is often not reported. This information is however relevant, even crucial, because some studies in the US show that some of the teenagers who experiment vaping are using non-nicotine liquids.

Regular smoking should also be the outcome of choice, not only few puffs, or the intention or the susceptibility to smoke. Finally, the validity of self-reporting is aslo questionable.

9 criteria to establish causality for the Gateway Theory

A British epidemiologist, Austin Bradford Hill, has had a lot of influence in the 1960s for establishing the link between smoking and cancer. He defined nine criteria to establish the causality link. A causes B if:


Jean-François Etter, Institute of Global Health (Geneva, Switzerland)
Jean-François Etter, Institute of Global Health (Geneva, Switzerland)

Strength of association

A few cases are not sufficient to support the causality, the minimal effect is also hard to establish since the prevalence of vaping among the youth is very low. The specialist is not confident that below 10% (relative risk), given the complexity of the study, any risk may be significant.

“Small effect + low prevalence of vaping in non-smokers: gateway affects are unlikely to represent a significant public health problem”Jean-François Etter


This is where the theory is the weakest, he believes. Can other things cause that these young people first vaped then smoked? These young people certainly have a propensity to use nicotine in any form, a common susceptibility of both vaping and smoking caused by social influences, family and friends, a personality (singer, actor), weaknesses, psychiatric disorders… A range of liabilites could explain that a young enters in both vaping and smoking.

The two behaviors are very close to each other: the scientists examine nicotine consumption but it is just the device that is changing. Multivariate statistical models are the way epidemiologist address this issue. Even when adjusted to the other variables evoked above, some of the variance of vaping still captures some of the variabilities of smoking to blur the final image just because vaping is a proxy for the susceptibility of nicotine use.

Smokers also do things differently than non-smokers, they take more risk in general, which propagates the susceptibility to much more variables (driving habits, for example).

The common variability theory (people vape and smoke because they have a common susceptibility) that is seen between vaping and smoking is supported by large amount of scientific evidence with other substances (marijuana, heroin, alcohol). It also accounts for the dose-response effect and is a powerful tool for building a policy and conducting research.


Temporality says that the cause must precede the effect. But antecedence is not sufficient to demonstrate a gateway effect, because causality is necessary to make temporality relevant to the theory. Antecedence is very hard to establish, based on self-reports.

Many studies of gateway effect show only antecedence and increased relative risk: this is not sufficient, because it does not eliminate:
– confounding by other factors (even after statistical adjustment),
– doubts about the plausibility of the theory.


Due nicotine use, which is the fact of both vaping and smoking, the gateway theory is circular. In most cases with kids, smoking precedes vaping and this constitutes by itself a strong argument against the gateway theory. In the 1970s, it was told that people had to “graduate” in drugs consumption with marijuana and easy-access drugs before trying drugs other more scary and harder to obtain drugs like heroin. It is definitely not the case with vaping.

Why would people who chose to vape rather than smoke suddenly changed their mind to smoke anyway? Because of their addiction to nicotine?

Most e-cigarette models are not very addictive, they do not deliver nicotine quickly enough to become as addictive as cigarettes. And if nicotine delivery was not enough, why turning to smoking as they could turn to e-cigarette models that are more performant at delivering nicotine? We don’t have testimonies in the literature of vapers who are not smokers and who switch to go on to smoking. We rather have the reverse, people who vape because they don’t like to smoke, because it’s risky, smelly, dangerous, expensive, because of social pressures to quit smoking (smoking bans everywhere, family and friends).


Do we have consistency with other lines of evidence? In the USA, the UK, France where vaping is frequent among kids, prevalence of smoking is low and decreasing fast. Prevalence data cannot be used as a proof but is it reassuring that vaping prevalence is not increasing at the same time as smoking prevalence and it makes the scheme coherent.

There is also no proof that vaping caused nicotine addiction in non-smokers.


Do nicotine medication and smokeless tobacco act similarly? Nicotine medications are not very addictive because the pharmaceutic industry did not want them to be addictive. The non-tobacco users who get addicted to nicotine medications are very rare. And there is no case described in the literature of a non-tobacco user, getting addicted to nicotine gums, who turn to smoking.

Smokeless tobacco is a different product since it delivers a larger amount of nicotine and is addictive. But is smokeless tobacco a gateway to smoking? In the countries where smokeless tobacco products are used, smoking prevalence is low and their consumption by young people is not paralleled by consumption of cigarettes. Apparently, even if more addictive than NRTs there is not gateway effect for snus in Norway or Sweden.

What researchers need are models that not only describe the transition from A to B but a comprehensive models that explain movements in all directions between A and B like the one proposed by Dave B. Abrams.

Even if there was a gateway effect, because e-cigarettes probably help people quit smoking, the global effect toward public health is still going to be positive. A gateway effect would need to be very strong to counteract the cessation effect.Jean-François Etter

Conclusion: The evidence of a gateway effect is very hard to obtain

Source J.-F. Etter, Vapexpo 2016
Source J.-F. Etter, Vapexpo 2016

In conclusion, evidence of a gateway effect is very hard to obtain. The common liability model is plausible and a better base for policy making and research.

Nevertheless, in spite of its weaknesses and the lack of data to support the gateway theory, it has an enormous political influence to discourage people from vaping and increase smoking with a strong negative effect.

If policy makers understood this, they would be less inclined to restrict access to vaping products.

Comment by Konstantinos Farsalinos


In order to support the gateway theory, we need to demonstrate that people who never used any tobacco cigarette got addicted to nicotine through vaping and then, switch to smoking because of their addiction. So many other parameters could interfere which make this theory very weak.

If there were no restriction on e-cigarette sales, it would probably be easier for a young who wants to inhale something to start vaping but it is not because he started vaping that he will start smoking. Most of the studies defending the gateway theory, so far, do address vaping as “ever use of a vaporizer” which is very different from daily use of it. The question should be “Are you a daily user of a nicotine delivering e-cigarette?“.

Comment by Riccardo Polosa


The Gateway theory is completely destroyed by looking at the smoking rates. In countries where vaping has been adopted by people, the smoking prevalence is decreasing. If vaping was the cause of smoking, there should be the other way. The gateway is out of smoking, not to smoking.

  • Kosta Kurteev

    Does it not have something to do with the fact that tobacco products are heavily taxed in many countries, so the real danger is, that when more and more people switch to vaping, much of the revenue will be lost? Isn’t it why there is such a need to come up with nonsensical and evidence-lacking reasons to demonise e-cigarettes? This may actually be more of a case in countries which lack a universal healthcare, such as the USA.