Abstract
OBJECTIVE: Cannabis use is common among adolescents. Limited research has examined how specific cannabis use behaviors may differentially increase cannabis use disorder (CUD) risk. METHOD: Data were from two waves of a prospective cohort of Southern California adolescents who used cannabis in the past six months (N = 420; Fall 2022, Spring 2023). Multivariable logistic regression models were used to examine the association of each cannabis use behavior at baseline (number of modes of cannabis administration used, product type used first and most often, cannabinoid formulation used most often, past 30-day frequency, quantity) with probable CUD at a six-month follow-up (measured using the Cannabis Abuse Screening Test), adjusting for sociodemographic factors, other substance use, and probable CUD at baseline. RESULTS: Most adolescents (69.8%) used >1 mode of cannabis in the past six-months; concentrates (vaped or dabbed) were the product type used most often (37.5%). Greater odds of CUD were observed for each additional cannabis product used in the past 6 months (vs. only one; adjusted odds ratio [AOR] range = 2.83-4.13; ps < .05), and for frequent past 30-day cannabis use (10+ days/month vs. 1-2 days/month; AOR = 2.87, 95%CI = [1.31, 6.27]). No other cannabis behaviors or characteristics were associated with CUD in adjusted models. CONCLUSIONS: Frequent past-month cannabis use and using multiple modes of administration were predictors of probable CUD. In addition to monitoring the overall presence of adolescent cannabis use, these results demonstrate the importance of raising awareness about specific use behaviors associated with CUD risk among youth, particularly in regions with well-developed cannabis markets.