Abstract
OBJECTIVE: Little research has examined psychosocial influences differentially associated with driving under the influence of alcohol and marijuana (DUIA/DUIM) or of being a passenger in a vehicle driven by someone under the influence of alcohol or marijuana (PVA/PVM). METHOD: Using 2019 survey data among 2,375 young adults (M age = 24.66, 57.7% female, 51.2% in states with legal marijuana retail) in a longitudinal study examining vaping retail, we conducted latent class analysis (LCA) to identify classes of DUI-related behaviors and multinomial logistic regression to identify correlates of classes (e.g., individual, interpersonal, policy related). RESULTS: A total of 22.6% reported past-month DUIA, 24.1% DUIM, 29.7% PVA, and 27.7% PVM. Of those reporting DUIA, 41.4% reported DUIM, 71.8% PVA, and 40.7% PVM. Of those reporting DUIM, 47.7% reported DUIA, 49.6% PVA, and 83.8% PVM. LCA indicated four classes: alcohol and marijuana risk takers (5.9%), all of whom reported DUIA and DUIM, 81.2% PVA, and 89.2% PVM; marijuana risk takers (14.2%), 45.1% DUIM, 100% PVM, 7.8% PVA, and 0% DUIA; alcohol risk takers (24.1%), 40.1% DUIA, 98.6% PVA, 2.6% DUIM, and 33.8% PVM; and low risk takers (55.8%), 8.7% DUIA, 4.5% DUIM, and 0% PVA/PVM. Relative to the other classes, alcohol and marijuana risk takers were more likely male, heterosexual, and White, and alcohol and marijuana risk takers reported greater depressive symptoms. Influences of parent/peer alcohol versus marijuana use were specific to the respective DUI risk behaviors. No differences emerged for legal marijuana retail. CONCLUSIONS: Findings regarding those at particular risk, social influences, and situations under which DUI-related behaviors occur should inform future interventions and research.