A Person-Centered Approach to Identifying Polyvictimization Trajectories and Their Associations with Cannabis Use

Background: Childhood maltreatment is a well-established risk factor for revictimization across development, including community violence exposure (CVE) and intimate partner violence (IPV). While victimization across the life course heightens risk for cannabis use, it remains unclear whether and how specific trajectories of polyvictimization differentially shape that risk. Standard approaches examine child maltreatment and later victimization separately, while others rely on cumulative measures that obscure variation in timing and exposure type—which may be differentially associated with cannabis use severity.

Violence victimization prevalence is highest in emerging adulthood (ages 18-25), a period characterized by major life transitions. Individuals experiencing victimization across both childhood and emerging adulthood may be at elevated risk for greater cannabis use severity, with implications for risk of substance use disorder during a critical developmental period. Identifying heterogeneity in victimization trajectories may improve early identification of young adults at greatest risk for severe cannabis use. This study identifies latent classes of violence victimization trajectories spanning childhood through emerging adulthood and examines their associations with cannabis use outcomes in adulthood.

Methods: The present study uses Waves 1-4 of the National Longitudinal Study of Adolescent to Adult Health (Add Health; n≈11,863), a nationally representative cohort of US adolescents in grades 7-12 during the 1994-95 school year (Wave 4; ages 24-32 years). Violence victimization included child maltreatment (i.e., physical, emotional, and sexual abuse) before age 18, and CVE (witnessing and direct victimization) and IPV (physical and sexual) in emerging adulthood (Wave 3; ages 18-26 years). Cannabis outcomes included current use (past 30 days), frequent use (weekly or more), and symptoms of cannabis use disorder. Latent class analysis with robust maximum likelihood estimation identified victimization trajectory classes, with the Bayesian Information Criterion guiding selection of the number of classes. We sought solutions with average posterior predicted class probabilities >0.80 and entropy >0.60. The Bolck-Croon-Hagenaars method examined associations between latent class membership and each cannabis outcome after accounting for classification uncertainty.

Results: On average, participants reported 1.24 total incidents of child maltreatment. Approximately 26% of participants reported physical IPV victimization, 13.4% reported CVE victimization, 12% reported sexual IPV victimization, and 8.7% reported witnessing CVE. We hypothesize at least four victimization trajectory classes will emerge: chronically high victimization, childhood-only, emerging adulthood-onset, and chronically low. We further hypothesize that childhood-onset victimization will confer elevated odds of current cannabis use, frequent cannabis use, and symptoms of cannabis use disorder in adulthood, above and beyond classes characterized by later-onset victimization alone.

Conclusions and Implications: Polyvictimization research has largely focused on cumulative burden, but some trajectory types, particularly childhood-onset victimization, may differentially predict cannabis use above and beyond overall exposure severity. This reframes cannabis use among survivors as a potentially patterned response to developmentally timed and type-specific harm, rather than an individual failing. Findings also underscore the importance of centering survivors as informed stakeholders for cannabis policy and trauma-informed program design and have further implications for optimizing screening practices for high-risk cannabis use in social work and related professions.