Effect of cannabis liberalization on suicide and mental illness following recreational access: a state-level longitudinal analysis in the USA

Abstract

AbstractObjectiveTo standardize the implementation dates of various cannabis liberalization policies and determine whether previous research by Anderson et al. [D.M. Anderson, D.I. Rees, J.J. Sabia, American Journal of Public Health 104, 2369-2376] on medical marijuana access and population-level suicidality is robust to additional years of data and further cannabis liberalization in the form of recreational marijuana access.DesignA state-level longitudinal (panel) analysis. Suicide mortality rates from the National Center for Health Statistics and mental health morbidity rates from the National Survey on Drug Use and Health were employed with the procedures outlined by Anderson et al., using weighted ordinary least squares for three different specifications with various combinations of control variables as a sensitivity analysis to test for robustness.SettingAll 50 states and Washington, DC for the period 1999-2019.ParticipantsUSA population.InterventionsCannabis liberalization policies in the form of recreational or medical access.Primary and Secondary Outcome MeasuresState-level population mental health outcomes in the form of suicide mortality among various age groups for males and females defined by the International Classification of Diseases, Tenth Revision and rates of mental illness, serious mental illness, major depression, and suicidal ideation defined by the Substance Abuse and Mental Health Services Administration.ResultsRecreational marijuana access was associated with a 6.29% reduction (95% CI −11.82% to −0.42%) in suicide rates for males in the 40 to 49 age group. No other mental health outcomes were consistently affected by cannabis liberalization.ConclusionsAdverse mental health outcomes do not follow cannabis liberalization at the state level, confirming the findings of Anderson et al. In addition, there is evidence that recreational marijuana access reduces suicide rates for middle-aged males.Strengths and limitations of this studyCannabis liberalization policies, which vary greatly throughout the literature, are explicitly defined and corrected from previous studies.SAMHSA suppresses state-level geographical information for individual-level responses in the NSDUH and the analysis relied on population averages for a small number of age groups published in the NSDUH State Prevalence Estimates, which did not allow us to evaluate gender differences for mental health outcomes.The reliability of NSDUH and suicide data to estimate true population rates is highly debated.Population-level analyses of longitudinal data can be evaluated with multiple accepted methods from the medical literature and it is not clear whether weighted ordinary least squares is the most appropriate approach for this type of analysis.