A social cascades model of addiction: Examining effects of psychological interventions on social and substance use outcomes via meta-analytic review.

Addiction has long been observed to manifest in social landscapes characterized by high levels of dysfunction, including interpersonal conflict, low social support, and a high density of substance-using peers. Two lines of work have emerged to understand these correlations-one examining how social environments might foster addiction and the other centered primarily around the detrimental effects of substance use on social life. Here, we test a novel cascades model positing bidirectional relationships between individuals' social environments and substance use, integrating theory from basic behavioral research with methods from intervention science. We explore both social and substance use outcomes derived from randomized intervention trials for participants with substance problems (k = 79 trials; N = 13,946 patients). Results indicate modest improvements in social outcomes from substance use intervention, d = 0.15 [0.08, 0.22]. Results further indicated that social and substance use outcomes were correlated at multiple levels, with intervention-related improvements in social outcomes temporally preceding improvements in substance use symptoms, b = 0.33, p = .007, and vice versa, b = 0.39, p < .001. Finally, targeted comparisons isolating effects of social versus substance use intervention components provided partial support for the cascades model, with significant cross-domain effects emerging selectively in models isolating social elements of randomized interventions, d = 0.12 [0.02, 0.22], and models further indicating superiority for therapies involving close social ties (e.g., family), d = 0.23 [0.05, 0.40]. In sum, results move toward an integrative view of social processes in relation to precursors and consequences of substance use, with potential implications for methodology, theory, and intervention. (PsycInfo Database Record (c) 2026 APA, all rights reserved).