Late relapse among ex-smokers using e-cigarettes following a quit attempt: a prospective one-year follow-up study

Background: Long-term e-cigarettes (EC) use by ex-smokers is increasing. This could reduce smoking relapse risk, by satisfying ex-smokers’ need to obtain nicotine from other sources, or it may promote relapse, either by maintaining nicotine dependence and/or through its behavioral similarity to smoking. This study aimed to assess the association of baseline exclusive EC use by smokers who quit in the past year with smoking status/relapse at 1-year follow-up compared with (i) a behavioral control group, smokers who quit in the past year and exclusively used nicotine replacement therapy (NRT) at baseline, and (ii), smokers who quit in the past year with no product (NP) use at baseline.

Methods: Data came from the Smoking Toolkit Study, a representative survey of adults in England, collected monthly from 04/2015 to 12/2022, with follow-up 1 year later. Eligible participants were ex-smokers who made a successful past-year quit attempt and were exclusive EC (N=125), NRT (40) or NP (220) users at baseline and who provided follow-up data. Main outcomes were follow-up smoking (current/ex-smoker) and relapse (relapsers/continuous abstainer) status. Multiple logistic regressions compared groups on outcomes, adjusted for baseline socio-demographic, smoking and quit-attempt related characteristics. Bayes Factors were calculated, testing against an estimated relapse risk for EC users reported in previous literature (odds ratio (OR)=2.4).

Results: Of past-year ex-smokers, with a median abstinence of 3-6 months at baseline, 39.2% (95% Confidence Interval (CI) 31.1-48.0) with exclusive EC use, 40.0% (95%CI 26.4-55.4) with exclusive NRT use and 31.4% (95%CI 25.6-37.8) with NP use were smokers at 1 year follow-up. Further, 52.0% (95%CI 43.3-60.6) of baseline ex-smokers with exclusive EC use, 56.4% (95%CI 41.0-70.7) with exclusive NRT use and 45.2% (95%CI 38.8-51.8) with NP use had a relapse since baseline. In adjusted analysis, relapse risk for EC use did not differ from NP (OR 1.3, 95%CI 0.8-2.3) or NRT users (OR 1.1, 95%CI 0.5-2.7), with data providing anecdotal evidence against an increased relapse risk for EC users (Bayes factors 0.5-0.7). Similar associations were found with smoking status at follow-up. Being a smoker or relapser at 1-year follow up was independently associated with younger age, greater urges to smoke and shorter quit length at baseline; being a relapser was additionally associated with lower educational attainment and having made more past-year quit attempts.

Conclusions: In England, exclusive use of EC by ex-smokers was not independently associated with a greater risk of smoking or relapse one year later compared with NRT or NP use. Data provide anecdotal evidence against increased relapse risks reported in the literature but cannot rule out potentially smaller associations.