Lonelier Young Adults Used Cannabis More Often and Had Higher Rates of Cannabis Use Disorder

Using marginal structural modeling to control for multiple confounders, loneliness at ages 21 and 23 predicted greater cannabis use frequency at subsequent waves (OR 1.13 per SD increase, 95% CI 1.06-1.20). Loneliness also predicted higher cannabis use disorder risk (prevalence ratio 1.20, 95% CI 1.12-1.29) and hazardous alcohol use (PR 1.10, 95% CI 1.03-1.17). The cannabis associations held after accounting for depression, prior substance use, and other psychosocial factors.

Data from 4,407 young adults in the Community Youth Development Study. Loneliness measured at ages 21 and 23; substance outcomes at ages 23 and 26. Marginal structural modeling accounted for time-fixed and time-varying confounders including depressive symptoms, prior substance use, and psychosocial factors. Cannabis use disorder diagnosed using validated criteria.

In an era of rising loneliness and social isolation, this study establishes that loneliness independently predicts problematic cannabis use, separate from depression. This suggests that addressing loneliness and social connection could be a novel approach to preventing cannabis use disorders in young adults.

Loneliness is increasingly recognized as a public health crisis. This study adds substance use to the growing list of loneliness-related health consequences. The finding that loneliness predicts cannabis problems independently of depression suggests loneliness may operate through distinct pathways, potentially social reinforcement or emotion regulation.

Two-year lag between loneliness and outcome assessments may miss shorter-term dynamics. Self-reported loneliness measure (4 items). Marginal structural models address but cannot eliminate all confounding. Sample from community study may not represent all young adults.