Typology of Adolescents Exposed to Non-medical Cannabis Marketing and Associations with Consumption Patterns.

Latent class analysis (LCA) identified subtypes of cannabis marketing exposure among adolescents and assessed whether the classes were associated with three cannabis use outcomes: past 28-day use, poly-cannabis use, and symptoms of cannabis use disorder (CUD). Survey data were from 471 adolescents (aged 15-19 years) who lived in four states with legal non-medical cannabis in 2018. Measures included social media engagement and cannabis outcomes. LCA with robust maximum likelihood estimation identified latent classes. Chi-squared tests assessed whether empirically derived classes differed across demographics, and logistic regression tested associations with cannabis use outcomes. Three classes were identified: digitally engaged (35.5%), digitally unengaged (36.5%), and traditional (28.0%). Both digitally engaged and unengaged classes were exposed to marketing on social media platforms, but youth in the engaged class interacted with posts and brands. Class membership differed by age (χ 2  = 14.89, p < 0.001) and school type, with the digitally engaged group older and not in school or in non-traditional schools (χ 2  = 16.22, p=0.01). As compared to the traditional class, youth in the digitally engaged class had 10.63 times the odds of past 28-day cannabis use (95% CI: 5.25, 21.51), 7.84 times the odds of poly-cannabis use (95% CI: 3.54, 17.33), and 13.85 times the odds of symptoms of CUD (95% CI: 3.96, 48.48). Youth in the digitally engaged class had higher odds of all cannabis use behaviors than the traditional class. These findings point to the possible use of algorithmic marketing to adolescents and suggest a need for monitoring and possible restrictions on digital cannabis marketing.