Bidirectional associations between e-cigarette use and alcohol use across adolescence.

Abstract

Evidence on prospective bidirectional associations between e-cigarette and alcohol use among adolescents can inform prevention and policy but is largely absent from the literature. Data were drawn from a prospective cohort of students attending 10 Los Angeles high schools (N = 3396; baseline mean age = 14.1, SD = 0.4). Students completed surveys every 6-months from 2013 to 2017; 8 total waves. Analyses were restricted to (a) individuals who were never users of alcohol (N = 2394) or (b) individuals who were never users of e-cigarettes (N = 2704) at baseline. Repeated-measures, generalized linear mixed models were used to estimate the adjusted odds of past 6-month alcohol and e-cigarette initiation, in separate models. Among alcohol never-users at baseline, 15.7 % (N = 375) initiated alcohol use over the study period. Compared to never-users of e-cigarettes, those who reported use of e-cigarettes had 3.5 times the odds of subsequently initiating alcohol use in the following wave (OR = 3.54; 95 % CI: 2.81, 4.47). Stronger associations were observed for males (OR = 4.94; 95 % CI: 3.78, 6.45) than for females (OR = 3.21; 95 % CI: 2.33, 4.41; p_interaction = 0.04). Among e-cigarette never-users at baseline, 26.3 % (N = 709) initiated e-cigarette use over the study period. Compared to never-users of alcohol, those who reported use of alcohol had 3.2 times the odds of subsequently initiating e-cigarette use in the following wave (OR = 3.23; 95 % CI: 2.68, 3.89). This association did not differ by gender. E-cigarette and alcohol use can be markers to identify youth at risk for future alcohol and e-cigarette use, respectively. Research examining mechanisms underlying these associations is needed to infer causality.