Cannabis Use Disorder and All-Cause Mortality in Canada: A Population-Based Cohort Study Using Linked Survey and Vital Statistics Data

<title>Abstract</title> <p> <bold>Background:</bold> Evidence concerning the association between cannabis use disorder and mortality remains limited and has primarily arisen from treatment-based, hospital-based, or administrative cohorts. We examined the association between lifetime cannabis use disorder and all-cause mortality in a nationally representative Canadian household cohort. <bold>Methods:</bold> We conducted a population-based cohort study of respondents aged 15 years or older from the 2012 Canadian Community Health Survey–Mental Health linked to the Canadian Vital Statistics Death Database. Lifetime cannabis use disorder was defined as DSM-IV cannabis abuse or dependence assessed using the World Mental Health Composite International Diagnostic Interview. Survey-weighted Cox proportional hazards models estimated associations with all-cause mortality through December 31, 2017. The primary model adjusted for age and sex. <bold>Results:</bold> Respondents with lifetime cannabis use disorder were younger and more likely to be male than respondents without the disorder. Crude mortality was lower among respondents with cannabis use disorder than among those without it (1.61% versus 3.69%). The unadjusted hazard ratio was 0.63 (95% CI 0.35–1.15). Adjustment for age reversed the direction of the association (HR 1.58, 95% CI 0.87–2.85). In the primary age- and sex-adjusted model, lifetime cannabis use disorder was associated with an HR of 1.32 (95% CI 0.71–2.43). Adjusted predicted probabilities of death were 4.59% (95% CI 2.13–7.06) among respondents with cannabis use disorder and 3.52% (95% CI 3.23–3.81) among those without it. Cause-specific estimates could not be released because event counts did not meet Statistics Canada disclosure requirements. <bold>Conclusions:</bold> Lifetime cannabis use disorder was not associated with a statistically distinguishable increase in all-cause mortality during approximately 5.5 years of follow-up, although the adjusted estimate remained compatible with a potentially meaningful elevation in risk. Differences from hospital-based estimates may reflect variation in disorder severity, recency, healthcare engagement, and case ascertainment. </p>