Cancer Research UK funded research does not support the concern that e-cigarettes are a ‘strong gateway’ into smoking

Cancer Research UK

The following response has been issued to address the misinterpretation by the media of a Cancer Research UK funded paper looking into a potential ‘gateway effect’ between e-cigarette vaping and cigarette smoking in young people.   

Carl Alexander from Cancer Research UK said: “While this study shows young people who experiment with e-cigarettes are likely to try smoking and vice versa,  the researchers didn’t look at whether the youngsters then became regular users or whether they might have tried smoking anyway. Research like this is important to help us understand the potential impact of e-cigarettes on young people.

“It’s illegal to sell e-cigarettes to under 18s in this country, and regular use among those who’ve never smoked tobacco is very low. Tobacco is the biggest preventable cause of death in the world. The evidence so far shows that e-cigarettes are much less harmful than smoking and people are successfully using them to give up tobacco.”

How the research was misinterpreted

Cancer Research UK is concerned that research on e-cigarette use among young people carried out by King’s College London, and funded by Cancer Research UK, is being misinterpreted as demonstrating “strong evidence of a so-called ‘gateway effect’” from e-cigarettes to smoking.

The study did find an association between both e-cigarette use and smoking, and between smoking and e-cigarette use. However, finding this association does not mean one behaviour caused the other. This study found that it  is equally likely that trying an e-cigarette “causes” trying smoking, as trying smoking “causes” trying an e-cigarette.

The proportion of young people regularly using e-cigarettes is very small. In this study, the proportion of young people using these products at least monthly was low (5% > monthly smoking, 2% > monthly e-cigarette use), consistent with other findings in Great Britain. This means it was only possible to explore associations between trying each product, and not regular use. We can’t say for sure whether the young people in this study who tried these products would become regular smokers or whether they would have gone on to smoke regardless of e-cigarettes.

Furthermore, the study found that it was much more common for young people to have tried smoking than e-cigarettes: only 21 had tried an e-cigarette but not smoked, compared with 118 who had tried smoking but not e-cigarettes. 

Youth smoking rates in the UK continue to decline, and regular use of e-cigarettes is rare and is almost entirely confined to those who have smoked before.

The evidence does not therefore support the concern that e-cigarettes are a strong gateway into smoking. In the UK sale of e-cigarettes to under-18s is prohibited, and advertising for e-cigarettes on TV, radio, the internet and the press is prohibited. All of these measures are designed to protect young people in particular.

Notes on methodology

The study used causal mediation analysis to investigate whether there was a causal relationship between e-cigarette experimentation and smoking, and vice versa. These analyses confirmed an association, but also found that the strength of the apparently causal association between e-cigarette use and smoking was similar to that between smoking and e-cigarette use. So it is equally likely trying an e-cigarette “causes” trying smoking, as trying smoking “causes” trying an e-cigarette.

Additionally, it is likely that there are underlying factors or covariates that are driving both smoking and e-cigarette use – people who try e-cigarettes are likely to try similar things, like smoking. Or the types of people who try one are similar to the types of people who try the other. This could be driving an association between e-cigarette use and smoking, and vice versa.

Although this study did test some of these covariates, it didn’t look at all the possible factors that might be at play, and overall this effect is difficult to study completely. This means that we can’t completely rule out the effect of covariates creating an apparently causal association between e-cigarette use and smoking.

A further explanation of the findings can be found on NHS Choices