Importance Many individuals can use cannabis without harm, but individuals with psychosis are particularly vulnerable to negative effects from cannabis. No studies to date have investigated how recreational cannabis legalization (RCL) affects cannabis use in individuals with psychosis. Objective To assess impacts of RCL in cannabis use among individuals who have ever been diagnosed with a psychotic illness or episode in the US. Design, Setting, and Participants This study used data from 2014 to 2022 on state-level RCL with 5 years of follow-up from the Population Assessment of Tobacco and Health, a nationally representative longitudinal cohort study of 1856 adults aged 18 years and older in the US. The study sample consisted of individuals with a lifetime history of psychosis. Data were analyzed between January and June 2025. Main Outcomes and Measures Percentage point (pp) changes in 30-day use of cannabis after RCL compared with control states using difference-in-differences methods. Results The study sample consisted of 1856 individuals, with a mean age of 36.6 years (SD, 14.8 years) contributing 7465 responses, and was 50.2% White, 58.2% female, and had a 30-day cannabis-use rate of 31.8%. Individuals with psychosis in RCL states significantly increased their 30-day cannabis use by 9.53 pp (95% CI, 3.05-16.00 pp; P = .004), with sensitivity analyses showing significant estimates after retail outlets opened but not before, and no changes in higher frequency use. Conclusions and Relevance In this study, individuals with psychosis reported a large increase in current cannabis use following legalization and commercialization of cannabis in their state, and by larger amounts than previously reported estimates of the general population. Given how cannabis can negatively affect illness course and health service utilization in individuals with psychosis, these results should be considered by regulators designing policies around taxation, potency, advertising, and health warnings.
Hyatt et al. (Wed,) studied this question.