Cannabis retail accessibility and adolescent use across a maturing market: Evidence from Canada.

Seabrook J.; Zhong S.; Cole A.; Philipneri A.; Anderson K.; Barakat C.; Martin G.; Leatherdale S.; Gilliland J. · Drug and alcohol dependence · 2026

**Background & Methods** This repeat cross-sectional study analyzed data from the COMPASS study (2018-2023) involving 419,887 secondary school students across four Canadian provinces (Alberta, British Columbia, Ontario, Quebec) to examine whether spatial accessibility to cannabis retailers was associated with adolescent cannabis use. Student survey responses were linked with geospatial measures of retail accessibility including distance to nearest retailer, retailer density, and inverse distance weighted density within 2000 m of schools, with multilevel logistic and ordinal regression models adjusting for individual and school-level covariates. **Key Findings** • Cannabis use among adolescents declined substantially over the study period, with past 12-month use decreasing from 22.2% (2018) to 10.0% (2023), despite continued expansion of retail outlets. • Associations between spatial accessibility measures (distance to retailers, retailer density) and cannabis use outcomes were generally small, inconsistent, and often non-significant across 264 sensitivity model specifications, with null effects predominating in mature retail markets. • While greater geographic accessibility was associated with higher use frequency among cannabis users in 2018 (early post-legalization), this relationship weakened and disappeared as retail availability expanded; perceived ease of access remained strongly associated with cannabis use but showed only modest correlation with actual retail accessibility. **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. **Evidence Quality** This large longitudinal study provides robust evidence from a substantial population with repeated assessments across multiple jurisdictions, strengthening generalizability within Canada. However, significant limitations include reliance on self-reported cannabis use data subject to recall and social desirability bias, inability to establish causality from observational data, and potential unmeasured confounding from individual characteristics (socioeconomic status, peer influence, family factors). The disconnect between perceived and actual retail accessibility suggests measurement challenges and highlights that proximity alone may not drive adolescent use in mature markets. Cross-sectional analysis at multiple time points strengthens temporal assessment but cannot definitively establish causation. The null findings in mature markets suggest policy implications may differ between early and established legalization contexts.

More recent cannabis research