Impact of Recreational and Medical Marijuana legalization on cannabis use disorders, serious mental illness, and mortality outcomes among Medicaid enrolled youth - ABSTRACT Adolescent cannabis use is a major public health crisis. Marijuana laws (ML) carries significant health impli- cations for American youth. Decriminalization, medicalization, and legalization of cannabis by a majority of US states over the past 25 years has dramatically shifted societal perceptions and adult use patterns. How mariju- ana policy changes have affected population-wide health of US youth and the downstream public health impli- cations of ML remain topics of significant debate. Cannabis is the most commonly used illicit drug by US ado- lescents and the main drug for which US teens obtain substance use (SU) treatment. Adolescent cannabis use is associated with negative long-term consequences on mental health outcomes, risk-taking behaviors (eg, motor vehicle accidents: MVA), and academic/job achievement. Currently, 36 states and the District of Colum- bia (DC) have enacted medical ML (MML); of these, 18 states and DC have enacted recreational ML (RML). While studies have assessed the association between MML and marijuana use in youth, little is known about the impact of ML on mental health outcomes for American youth. No studies to date have evaluated the impact of MML and RML enactment on youth mental health or mortality outcomes. This represents a major knowledge gap in research that, if answered, will provide critical information to guide ML policymaking. The overarching goal of this NIDA RFA-DA-22-037 application is to characterize the effects of state-level MML and RML, including degree of ML restrictiveness and the effect of varying provisions, on changes in can- nabis use disorder (CUD), serious mental illness (SMI), non-cannabis substance use disorder (NCSUD), sui- cide-related outcomes (SROs), all-cause mortality (ACM), and treatment utilization for SU and MH services, including prescription medication, in American youth. Our central hypothesis is that ML will increase CUD, SMI, NCSUD, SROs, and ACM, with less restrictive ML being associated with worse health outcomes. To achieve study objectives, a difference in difference (DID) quasi-experimental design will be implemented. The main in- tervention of interest is ML, specifically, enactment of MML and RML. We will use national Medicaid data from all 50 states and DC for the period 1/1/2008 to 12/31/2020 merged with the National Death Index data. The study population will include a 20% random sample of all US adolescents (12-17 years) and young adults (18- 25 years) (N=55 million). Medicaid-enrolled adolescents and young adults have higher prevalence rates of mental health and substance use disorders and poorer physical and mental health outcomes compared to age matched peers; thus, they are a large and vulnerable population in whom MML and RML effects are unknown. Multilevel models will be fit to obtain estimates of before vs. after changes among adolescents and young adults in states enacting MML and RML compared to changes in other states, controlling for Individual and state-level covariates.
