Biegun, Jeff2026-09-022026-09-022026-08-252026-08-25http://hdl.handle.net/1993/40055Following the 2018 legalization of recreational cannabis, Canada remains at the forefront of the developing post-legalization environment for cannabis and the consumption of cannabis-related products. However, there remains a need to understand patterns of cannabis consumption to inform harm-reduction policy. This dissertation applies the biopsychosocial framework incorporating understandings of the social determinants of health to assess relationships between hypothesized factors associated with risky cannabis consumption. This dissertation consists of two studies, each investigating consumption patterns evident among Manitoban and Canadian cannabis consumers in relation to factors including cannabis use frequency, cannabis use disorder, cannabis use motivations, cannabis use purposes and cannabis use methods, as well as sociodemographic covariates including age, income, sex or gender, and education. Study one involves an analysis of the Manitoba Liquor, Gambling and Cannabis Survey II (MLGCSII), where two OLS regression models were produced examining predictors of cannabis use frequency, and cannabis use disorder symptoms (CUD-S). Results suggest the coping motivation was the strongest predictor of cannabis use frequency (β=0.36, p<.001) and CUD-S (β=0.41, p<.001). A three-step mixed-indicators latent class analysis (LCA) examined the existence of latent classes on indicators including social, coping, and enhancement motivations, cannabis use purpose (recreational, medical, ‘both’) and cannabis use frequency. Three latent classes were identified which differed in terms of motivational intensity across motivation factors, cannabis use purposes, and cannabis use frequency, labelled low-, moderately- and highly-motivated consumers. CUD-S scores were a significant predictor of class membership, with higher CUD-S scores associated with greater odds of being in the highly-motivated group than the other two groups. Study two utilized the Canadian Community Health Survey (CCHS2020) to analyze predictors of cannabis use frequency and disorder at the national level, finding that cannabis use frequency and SDS scores were the strongest multivariate predictors of one another. This study extended this analysis to groups of individuals based upon method of cannabis consumption and cannabis use-purpose, utilizing multinomial logistic regression to examine odds ratios associated with hypothesized covariates of cannabis use disorder. SDS did not predict cannabis use-method group membership but did predict cannabis use-purpose group membership, with higher SDS scores predicting decreased odds of being a medical consumer versus recreational or both (medical and recreational). The results of these studies contribute to filling gaps in the research on cannabis consumption patterns, and the findings are discussed in light of policy implications and avenues for future research.engCannabis consumptionCannabis legalizationHarm ReductionSubstance useCannabis use motivations