E-cigarette use and subsequent tobacco smoking initiation: an umbrella review with Bayesian model meta-analysis

Electronic nicotine delivery systems (ENDS) have seen a global surge in their usage. However, their role in smoking initiation, particularly among non-smokers, remains a critical concern. This umbrella review evaluated the risk of tobacco-smoking initiation following ENDS use, focusing on non-smokers, by synthesising evidence using Bayesian meta-analysis.

Systematic review and Bayesian meta-analysis.

We searched databases including PubMed, EMBASE and Cochrane Library up to 1 May 2025. We included systematic reviews assessing tobacco-smoking initiation after ENDS use among non-smokers. Quality was evaluated using the AMSTAR-2 (A Measurement Tool to Assess Systematic Reviews - version 2). A Bayesian hierarchical meta-analysis was conducted to estimate pooled ORs with 95% credibility intervals (CrIs), using non-informative priors to model heterogeneity and uncertainty. This study was registered with PROSPERO (CRD42023464207).

From 5055 records, 13 systematic reviews were included. Of them, five were rated as 'high quality', one as 'moderate', four as 'low' and 'three' as critically low, as per AMSTAR-2 quality assessment. The meta-analysis revealed a pooled median OR of 3.1 (95% CrI 2.56 to 3.72) for smoking initiation among non-smoking ENDS users, with moderate heterogeneity (τ mean: 1.23). Among adolescents, the OR was 2.7 (95% CrI 2.27 to 3.21), with moderate heterogeneity (τ mean: 0.63). The risk of relapse in former tobacco smokers who used ENDS was doubled (RR: 2.03, 95% CI 1.39 to 2.96). Individuals who had never smoked but engaged with ENDS had more than twice the odds of developing an intention to smoke, with an OR of 2.21 (95% CI 1.86 to 2.61). The overall certainty of the evidence was rated very low due to the risk of bias in the studies and evidence of publication bias across the studies.

ENDS use is significantly associated with increased risk of tobacco smoking initiation among non-smokers, particularly adolescents, supporting the gateway hypothesis. These findings indicate the need for stringent regulatory and public health strategies to protect people from nicotine addiction.

CRD42023464207.