Longitudinal Effects of Cannabis Use on Suicidal Behaviour in Early Adulthood in Washington State, United States and Victoria, Australia.

Fitzpatrick A.; Heerde J.; Rowland B.; Batmaz E.; Toumbourou J. · Prevention science : the official journal of the Society for Prevention Research · 2026

**Background & Methods** This longitudinal study compared matched cohorts from Washington State, USA (recreational cannabis legalized 2012) and Victoria, Australia (recreational cannabis illegal) to investigate cannabis use effects on suicidal behaviour in early adulthood. Data from 1,470 participants (55.4% female) in the International Youth Development Study were collected via self-report surveys at ages 25 (2014) and 29 (2018), assessing cannabis use, substance use, proximity to sales outlets, and suicidal behaviour, with analysis performed using multivariate negative binomial regression and propensity score methods. **Key Findings** • Washington State participants reported significantly higher past-year cannabis use at age 25 compared to Victoria, with increased use continuing through age 29, while Victorian participants showed declining cannabis use over the same period. Washington State also demonstrated lower self-harm rates at age 29 despite higher cannabis prevalence. • Higher cannabis use at age 25 and increased cannabis consumption at age 29 independently predicted increased suicidal behaviour in both regions. Suicidal behaviour at baseline (2014) was a significant predictor of suicidal behaviour at follow-up (2018), with stronger predictive risk in Washington State. • Propensity score analysis confirmed cannabis use had a significant effect on suicidal behaviour. Legalisation-related indicators (Washington State location and proximity to cannabis stores) significantly predicted increased cannabis use, yet Washington State maintained lower overall suicidal behaviour rates than Victoria despite higher cannabis prevalence. **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. **Evidence Quality** Study limitations include reliance on self-report data, which may introduce recall bias and social desirability bias. The observational design cannot establish causation despite regression adjustments. The paradoxical finding of higher cannabis use but lower suicidal behaviour in Washington State suggests unmeasured confounding variables and indicates that cannabis use represents only one determinant of suicidal behaviour. Cross-national differences in healthcare systems, mental health services, and cultural factors were not controlled. This represents moderate-quality evidence due to longitudinal design and large sample size, though causality remains unestablished.

More recent cannabis research