Early social media use linked to 17-fold increase in cannabis experimentation

Research on social media use has been largely cross-sectional, with limited longitudinal measurement of social media in relation to substance use. This study examined whether social media time trajectories were prospectively associated with substance use experimentation. The authors analyzed data from 7,166 adolescents (mean age, 10 years [SD=0.6]; 48.6% female) from the Adolescent Brain Cognitive Development Study, from baseline (2016-2018, ages 9-11) to year 5 (2021-2023, ages 13-16). Group-based trajectory modeling identified patterns of daily social media use from baseline to year 4. Alcohol, cannabis, and nicotine experimentation were assessed at year 5. Multiple logistic regression models estimated the associations between social media time trajectories and substance use experimentation, adjusting for baseline age, sex, race, ethnicity, household income, parental education, externalizing symptoms, internalizing symptoms, non-social-media screen time, study site, and respective substance use experimentation at baseline. Four social media time trajectories were identified: 1) no or very low use, 2) moderate, gradual increasing, 3) mid-onset, rapid increasing, and 4) early-onset, rapid increasing. Compared to the no or very low use trajectory, all increasing social media trajectories were associated with higher odds of alcohol, cannabis, and nicotine experimentation, with the early-onset, rapid increasing use trajectory having the largest magnitude of associations. For example, the early-onset, rapid increasing use trajectory was significantly associated with greater cannabis experimentation (adjusted odds ratio=16.86, 95% CI=10.28, 27.65). Increasing social media time trajectories were prospectively associated with a greater likelihood of substance use experimentation. These findings underscore the need for public health strategies that support healthier digital habits during early adolescence.