select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to study suggesting a link between concurrent vaping in early teen smokers and persistent and heavier smoking in late teens

A study published in Tobacco Control looks at e-cigarette use among early adolescent cigarette smokers.


Prof Lion Shahab, Professor of Health Psychology and Co-Director of the UCL Tobacco and Alcohol Research Group, University College London, said:

“This paper provides a detailed analysis of the “disruption” vs. “entrenchment hypothesis”, i.e., the idea that co-use of e-cigarettes together with cigarettes may either divert youth away from continued use of cigarettes (disrupt) or strengthen cigarette use (entrench) later on. Using data from two longitudinal studies – one in the US and one in the UK – this paper reports that those youth who co-use e-cigarettes along cigarettes were more likely to be continuing and more frequent smokers, consistent with the “entrenchment hypothesis”.

“While the study provides some important insights into the concept of entrenchment, it needs to be contextualised: this study focused only on adolescents who were already smoking cigarettes. Unfortunately, it is unclear whether e-cigarette use started before or after initiation with cigarettes. It is likely that youth starting to smoke very early on (here only those who smoked by age 15 were included) and then try e-cigarettes are quite different from those who start using e-cigarettes and then try cigarettes. Not separating out these sequence effects makes it difficult to accurately assess disruption vs entrenchment as e-cigarettes may entrench smoking if used after initiation with cigarettes but disrupt smoking if used before initiation with cigarettes. There’s also another complication in interpretating results from this study. As this is an observational study, youth self-select into the categories of early cigarette smoker and early cigarette and e-cigarette co-user. The question is whether both groups are similar. One way to overcome potential bias is to control for important confounders that drive the behaviour of interest (cigarette smoking) going forward. While it is good to see that the authors attempt to control for some of these (including smoking in the home and smoking by friends), they did not adjust for two key determinants of smoking in youth and adults: dependence and mental health. We know that smokers who are more dependent or have mental health problems find it much harder to stop smoking and therefore are less likely to do so. It is conceivable that those youth already smoking who suffer from mental health problems may seek out more nicotine and those who are more nicotine dependent may also be more likely to start using e-cigarettes to overcome nicotine withdrawal in situations where smoking cigarettes is not possible but use of e-cigarettes is (because e-cigarette use is more easily concealed, e.g., in school or at home). As the analysis did not control for these confounders, it is difficult to delineate if it is e-cigarette use per se or if it is other factors leading to co-use of e-cigarettes with cigarettes, which are responsible for entrenching cigarette use later on.

“Either way, this useful study highlights the need to investigate this issue further and strengthen regulation to stop access to cigarettes and e-cigarettes in youth. We know that starting to smoke early makes it much more difficult to stop smoking later and results in more smoking-related harms. Increasing the age of sale of cigarettes and stricter reinforcement of cigarette and e-cigarette sale bans for kids is needed to ensure no-one becomes addicted to nicotine in the first place.”



Dr Jamie Hartmann-Boyce, Associate Professor in Evidence-Based Policy and Practice, University of Oxford, said:

“Whether or not e-cigarettes cause young people to start smoking is a critical public health issue, with important policy implications moving forward. This study adds to a large body of evidence showing that young people who use e-cigarettes are more likely to smoke, but like all these other studies it cannot establish whether one causes the other (because of the possibility of unknown/unmeasured confounders). Although well-conducted, this type of study cannot prove that vaping causes young people to start smoking or to smoke more than they would have in the absence of e-cigarettes – a limitation acknowledged by the authors. If vaping does cause young people to smoke, we would expect to see youth smoking rates increase as youth vaping rates rise – there is as of yet no clear evidence of this happening.”


Dr Sharon Cox, Principal Research Fellow, UCL Tobacco and Alcohol Research Group, University College London, said:

“The authors acknowledge their findings do not demonstrate that using e-cigarettes as well as smoking plays a causal role in entrenching smoking. However, this is at odds with their policy implications which state the availability of e-cigarettes should be restricted to prevent e-cigarette use entrenching smoking. Although it is reasonable, regardless of these data, to minimise the extent to which adolescents can access tobacco and e-cigarettes.

“The young people surveyed were very young when both smoking and vaping (before the age of 15). Evidence shows engaging in smoking in early teens is associated with other residual ‘risky’ behaviours, and while socioeconomic and education status is accounted for, engagement in a range of other behaviours is not. Therefore, we are not presented with a full picture of these young people’s lives and characteristics which could reasonably be expected to lead them to continued smoking, irrespective of any e-cigarette use.

“The entrenchment hypothesis is similar to the ‘gateway hypothesis’. The conclusions suggest e-cigarettes have played a causal role in entrenching the youth in their smoking. Nicotine is uniquely addictive and difficult to quit when smoked in cigarettes, therefore we would predict those who are smoking and vaping may still be doing so at follow-up. At a population level, smoking rates among young people are falling in both the UK and the US. Therefore, at an individual level, another interpretation of this study could be that both smoking and vaping simply maintains nicotine use. 

“Finally, tobacco smoking is by far the more dangerous product of the two, and by implicating vaping as the cause for and maintenance of smoking, we risk losing sight of how addictive and harmful smoking is. Cigarettes have long been addicting young people, from this paper we see that has not changed.”


Prof Peter Hajek, Director of the Tobacco Dependence Research Unit, Queen Mary University of London, said:

“Seeing an association between having an experience with both cigarettes and e-cigarettes at 15 and smoking more frequently at 17, the first question is obviously whether the two groups differed in frequency of smoking already at 15. Surprisingly, the paper did not check this.  It claims that e-cigarettes caused more frequent smoking at 17, although the link may have been present at 15 already.

“More importantly though, an association between two similar behaviours is much less likely to indicate a causation than to simply reflect a common disposition behind both behaviours. In this case, that youth with a stronger reaction to or interest in nicotine products will try both products; and also use them more often.

“There is a simple check to see which interpretation is more likely. If vaping led to smoking, we would see an increase in smoking rates among young people since vaping came along. In reality, the decline in smoking among young people has accelerated.”


Prof Caitlin Notley, Professor of Addiction Sciences, Norwich Medical School, University of East Anglia (UEA), said:

“This study took two cohorts of young people, in the US and the UK, who smoke tobacco and asked them about ever use of an e-cigarette before the age of 15. The important point here is that the first behaviour measured is tobacco smoking, and also importantly, the measure of e-cigarette use is ever use – this includes a range of behaviours from a one off instance of trying something, through to regular daily use. It is unsurprising that young people who smoke tobacco are also likely to have tried an e-cigarette – this is a phenomena known as ‘common liability’, where the propensity to try certain consumption behaviours correlates with a propensity to also try other consumption behaviours (this could be smoking tobacco, using an e-cigarette, underage alcohol drinking or experimenting with illicit substances, for example). The authors themselves point out, in the limitations section of the paper, that ‘the observational design and the lack of measurement of early life confounders in the PATH do not permit full causal inference’, and that ‘the analyses do not account for the sequencing of tobacco and e-cigarette use’. Given this, it is inappropriate to conclude in the abstract that there ‘is evidence e-cigarette use among early adolescent smokers in the UK and USA leads (emphasis added) to higher odds of any smoking and more frequent tobacco cigarette use later in adolescence’. As association does not equate to causation, a more appropriate conclusion would be that early reported tobacco smoking is associated with trying an e-cigarette, which is also associated with later continued tobacco smoking. Or, simply, that young people who are attracted towards tobacco smoking at an early age are more likely to continue to report tobacco smoking at a later age, and may also experiment with other products.

“This paper does not, in my view, support an ‘entrenchment theory’ as reported, since e-cigarettes are not tobacco products, although both contain nicotine. The data presented support a common liability explanation of youth experimental behaviour.”


Prof John Britton, Emeritus Professor of Epidemiology, University of Nottingham, said:

“This study concludes that an association between vaping and continued smoking in adolescents who started smoking very young is consistent with entrenchment of smoking by vaping. An alternative hypothesis, that vaping by adolescent smokers is confounded with more severe nicotine addiction, does not appear to be considered. Most adolescent vaping is transient; those who persist with it are likely to be the most addicted smokers and hence those who are least likely to quit.”



E-cigarette use among early adolescent tobacco cigarette smokers: testing the disruption and entrenchment hypotheses in two longitudinal cohorts’ by Brian C Kelly et al. was published in Tobacco Control at 23:30 UK time Tuesday 18 April 2023.

DOI: 10.1136/tc-2022-057717



Declared interests

Prof Caitlin Notley: “I have no conflicts of interest to declare.”

Prof John Britton: “None.”

Prof Peter Hajek: “No conflict of interest.”

Dr Sharon Cox: “I have no competing interests.”

Dr Jamie Hartmann-Boyce: “I don’t have any conflicts of interest to declare, but for transparency am currently funded by Cancer Research UK to conduct research on the same topic.”

Prof Lion Shahab: “No specific COI.”

in this section

filter RoundUps by year

search by tag