Bidirectional associations of e-cigarette, conventional cigarette and waterpipe experimentation among adolescents: A cross-lagged model

Purpose: With an increasingly diverse tobacco product market, it is imperative to understand the trajectories of product experimentation in order to design effective prevention programs. This study aims to explore the bidirectional associations of conventional cigarette, e-cigarette and waterpipe experimentation in a large adolescent sample. Methods: Longitudinal assessment of conventional cigarette, e-cigarette and waterpipe use initiation was conducted in a school-based cohort of 1369 9th graders (mean age = 14.88 SD = 0.48 at baseline) during fall 2014 and reassessed 6-months later using online self-reported questionnaires. Autoregressive cross-lagged analysis within structural equation modeling framework was performed to simultaneously estimate the initiation of these products over a six-month period, controlling for age, gender, and participation in an intervention program to reduce conventional cigarette initiation. Results: Tobacco product lifetime use was prevalent at baseline in the sample: conventional cigarettes (48.4%), e-cigarettes (35.8%), and waterpipe (20.8%). At six-month follow-up, trying conventional cigarettes predicted trying e-cigarette (adjusted odds ratio (AOR) = 3.78, CI95%: 2.66-5.37) and trying waterpipe (AOR = 2.82, CI95%: 2.00-3.97). Trying e-cigarette predicted trying conventional cigarette (AOR = 3.57, CI95%: 1.96-6.49) and trying waterpipe (AOR = 1.51, CI95%: 1.07-2.14). Although trying waterpipe predicted trying e-cigarette at follow-up (AOR = 2.10, CI95%: 1.30-3.40), its use did not predict trying conventional cigarette (AOR = 0.55, CI95%: 0.24-1.30). Conclusions: The high rates of poly-tobacco use and the bidirectionality of tobacco product experimentation demands comprehensive tobacco control and prevention programs that address the increasingly diverse tobacco product market targeting adolescents.