Associations between cannabis policies and state-level specialty cannabis use disorder treatment in the United States, 2004-2019.

BACKGROUND: Cannabis use disorder (CUD) treatment prevalence decreased in the US between 2002 and 2019, yet structural mechanisms for this decrease are poorly understood. We tested associations between cannabis laws becoming effective and self-reported CUD treatment. METHODS: Restricted-use 2004-2019 National Surveys on Drug Use and Health included people ages 12+ classified as needing CUD treatment (i.e., past-year DSM-5-proxy CUD or last/current specialty treatment for cannabis). Time-varying indicators of medical cannabis laws (MCL) with/without cannabis dispensary provisions differentiated state-years before/after laws using effective dates. Multi-level logistic regressions with random state intercepts estimated individual- and state-adjusted CUD treatment odds by MCLs and model-based changes in specialty CUD treatment state-level prevalence. Secondary analyses tested associations between CUD treatment and MCL or recreational cannabis laws (RCL). RESULTS: Using a broad treatment need sample definition in 2004-2014, specialty CUD treatment prevalence decreased by 1.35 (95 % CI = -2.51, -0.18) points after MCL without dispensaries and by 2.15 points (95 % CI = -3.29, -1.00) after MCL with dispensaries provisions became effective, compared to before MCL. Among people with CUD in 2004-2014, specialty treatment decreased only in MCL states with dispensary provisions (aPD = -0.91, 95 % CI = -1.68, -0.13). MCL were not associated with CUD treatment use in 2015-2019. RCL were associated with lower CUD treatment among people classified as needing CUD treatment, but not among people with past-year CUD. CONCLUSIONS: Policy-related reductions in specialty CUD treatment were concentrated in states with cannabis dispensary provisions in 2004-2014, but not 2015-2019, and partly driven by reductions among people without past-year CUD. Other mechanisms (e.g., CUD symptom identification, criminal-legal referrals) could contribute to decreasing treatment trends.

Cannabis use disorder treatment use and perceived treatment need in the United States: Time trends and age differences between 2002 and 2019.

BACKGROUND: As trends in CUD treatment are understudied, we examined time trends in CUD treatment and perceived treatment need among people with CUD overall and by age group. METHODS: Data from the 2002-2019 National Survey on Drug Use and Health included 43,307 individuals who met past-year DSM-5-proxy CUD criteria. Last/current treatment for cannabis use (i.e., any or specialty CUD treatment) and perceived treatment need were regressed on survey year and age (12-17, 18-25, ≥26) using generalized linear spline models. Time-varying effect modification assessed the magnitude of age-treatment associations over time. RESULTS: Between 2002 and 2019, 6.1% of people with CUD used any CUD treatment, 2.8% used specialty treatment, and 2.2% perceived a treatment need. CUD treatment use decreased by 54.23% between 2002 and 2019 (9.11%-4.17%). Compared with adolescents, adults ages 18-25 were less likely to use specialty CUD treatment [aRR: 0.70: 95% CI: 0.52, 0.93] and ages ≥26 were more likely to perceive treatment need [aRR: 1.84: 95% CI: 1.19, 2.83]. Age-specific differences in the time-varying magnitude of associations were observed (e.g., in 2010 perceived treatment need was higher in ages ≥26 versus ages 12-17 [aOR: 2.34, 95% CI: 1.47, 3.71]). CONCLUSIONS: CUD treatment is decreasing and young adults have lower treatment use compared with adolescents. Attitudes towards cannabis use harms are shifting, potentially contributing to decreasing CUD treatment utilization and perceived treatment need. Future research should identify treatment barriers, especially among young adults with the lowest CUD treatment use.