OBJECTIVE The objective of this study was to estimate the association between vaping and subsequent initiation of smoking among Australian adolescents and explore the impact of design and analytical methods in previous studies.

METHODS We conducted a retrospective cohort analysis of cross-sectional data from 5114 Australian adolescents aged 14-17 recalling information on smoking and vaping initiation from age 12 to 17. The outcome was smoking initiation, analysed with negative-binomial regression to estimate incidence rate ratios (IRRs) for vape status (ever-vaped vs never-vaped) as a time-varying exposure. We also re-analysed using the methods of previous studies not accounting for the time-varying nature of e-cigarette exposure.

RESULTS Participants (n=5114) were retrospectively followed for 20478 person-years. After adjusting for socio-demographic variables and proxy measures of common liabilities for vaping and smoking, the rate of smoking initiation for those who ever-vaped was nearly 5 times that of those who never-vaped (IRR=4.9; 95% confidence interval: [3.9, 6.0], p<0.001), with IRRs considerably higher at younger ages. Not accounting for the time-varying nature of e-cigarette exposure in re-analysis attenuated the estimated IRR by 44%.

CONCLUSIONS Controlled analyses indicate that vaping markedly increases the risk of subsequent smoking initiation among Australian adolescents from age 12 to 17, with those aged 12, 13, and 14 bearing an alarmingly disproportionate burden of the elevated risk. Additionally, the relative risk of future smoking due to vaping may have been underestimated in other studies due to methodological differences.

IMPLICATIONS FOR PUBLIC HEALTH Our findings highlight the need for public health interventions and strict e-cigarette access laws.