Multi-directional pathways of tobacco and marijuana use, including comorbid use, among a population-based cohort of adolescents in Texas: A six-wave cross-lagged model.

Purpose: Past 30-day tobacco and marijuana use commonly occur among adolescents. It is unclear whether use of one product precedes the other, especially given the new climate surrounding marijuana legalization and the increasing popularity of e-cigarettes.

Methods: Six-panel cross-lagged regression models, with six months between each panel/Wave (2014-17), were used to model stability paths, bi-directional paths, and comorbid paths (i.e., correlations) between past 30-day use of marijuana and tobacco products. Data were derived from three cohorts of adolescents (n = 3907; weighted N = 461,069) in 6th, 8th, and 10th grades at baseline.

Results: Few bidirectional relationships between past 30-day tobacco and marijuana use were observed in early adolescence (6th grade). During the middle adolescence developmental period (8th grade), past 30-day marijuana use was prospectively associated with greater risk of past 30-day tobacco use. In late adolescence (10th grade), increased odds of past 30-day marijuana use given prior past 30-day e-cigarette use, and vice versa, were observed. For all cohorts, stability paths were common, especially for past 30-day marijuana use. Comorbid use was common in middle adolescence (8th grade) but small in magnitude.

Conclusions: This is the first study to longitudinally situate comorbid, past 30-day use of tobacco and marijuana and simultaneously examine bi-directional past 30-day use of these products for adolescents. Marijuana use more often and more strongly predicted subsequent tobacco use than the reverse, especially during middle adolescence (13-15 years old). Marijuana use should be considered when creating interventions that address adolescent e-cigarette use in the U.S.