Abstract
PURPOSE: Cannabis use is associated with increased risk for suicidality. However, the directionality of this association and the role of shared familial influences-such as genetics and environment-remains uncertain. Further, despite higher rates of cannabis use and suicidality endorsement in clinical samples, research in these populations is limited. This study examines cross-sectional and prospective associations between cannabis use and suicidality in a sample characterized by earlier cannabis initiation and heavier use than typically reported, while accounting for shared familial influences. METHOD: Adolescent sibling groups were recruited from Denver and San Diego (N = 1,261); at least one sibling was recruited from a substance use treatment program, alternative school, or juvenile probation. Participants completed clinical interviews assessing substance use and suicidality at three waves (2001--2019). Cannabis use frequency was examined as a predictor of suicidality using multilevel models accounting for shared familial influences and within-family clustering. Covariates included alcohol, tobacco, other substance use, age, and sex. Reverse associations and exploratory models assessing tobacco as a predictor were also examined. RESULTS: Cannabis use was not associated with suicidality (all p's > 0.05). Exploratory analyses suggested a possible association between tobacco and suicidality (e.g., Wave 1 within-family effect: OR = 1.037, p = 0.016), though these associations were largely reduced to non-significance after accounting for other substance use. CONCLUSIONS: Findings in this high-risk clinical sample are inconsistent with literature linking cannabis use to suicidality in community samples. Results underscore the need for further research on the association between general and polysubstance use risk and suicidality.