Two-part models identifying predictors of cigarette, e-cigarette, and cannabis use and change in use over time among young adults in the US

Document Type

Journal Article

Publication Date



The American journal on addictions




BACKGROUND AND OBJECTIVES: Limited longitudinal research has examined differential interpersonal and intrapersonal correlates of young adult use and use frequency of cigarettes, e-cigarettes, and cannabis. This study aimed to address these limitations. METHODS: We analyzed five waves of longitudinal data (2018-2020) among 3006 US young adults (M = 24.55, 44% male, 32% sexual minority, ~30% racial/ethnic minority). Two-part latent growth models examined likelihood of past-month cigarette, e-cigarette, and cannabis use (binary part) and days used (continuous part) and identified predictors (depressive symptoms, personality traits, adverse childhood experiences [ACEs], parental use) of baseline use and changes over time. RESULTS: Regarding baseline past-month use (27% cigarettes, 38% e-cigarettes, 39% cannabis), depressive symptoms, ACEs, and parental substance use predicted use outcomes (i.e., likelihood, frequency) for each product; extraversion predicted cigarette and e-cigarette use outcomes; openness predicted e-cigarette and cannabis use outcomes; conscientiousness negatively predicted cigarette and cannabis use outcomes; and agreeableness negatively predicted cannabis use frequency. Regarding longitudinal changes, conscientiousness predicted accelerated increase of cigarette use frequency at later timepoints; depressive symptoms predicted increases in likelihood of e-cigarette use but the association weakened over time; and parental cannabis use predicted decreased cannabis use frequency but the association weakened over time. DISCUSSION AND CONCLUSIONS: Young adult substance use interventions should target high-risk subgroups and focus on distinct factors impacting use, including chronic, escalating, and decreasing use. SCIENTIFIC SIGNIFICANCE: This study advances the literature regarding distinct predictors of different substance use outcomes and provides unique data to inform interventions targeting young adult cigarette, e-cigarette, and cannabis use.


Prevention and Community Health