(PS3-49) A Socio-ecological Model of Substance Use in Truant Adolescents: Examining Family, Peer, and Individual Influences

Abstract Text: Truant adolescents engage in more substance use than non-truant adolescents. Research has also shown a bidirectional relationship between truancy and substance use, such that as truancy behaviors increase, substance use also tends to rise. Socio-ecological models highlight the various layers of influence on adolescent substance use. In particular, family, peer, and individual-level factors have been identified as predictors of substance use among adolescents. This study aimed to characterize the relative contributions of these socio-ecological factors on adolescent alcohol and marijuana use in a truant population. The current sample includes truant adolescents (N = 110) aged 12-19 (Mage = 15.52, 64.5% male, 52.7% White). Adolescents completed measures of parental monitoring (Parental Monitoring Questionnaire; Stattin & Kerr, 2000), parent-child communication (Targeted Parent-Child Communication about Drugs; Miller-Day, 2004), parental conflict (Conflict Behavior Questionnaire; Robin & Foster, 1989), peer deviance (Peer Deviancy Scale; Institute of Behavioral Science, 1987) psychosocial problems (Pediatric-Symptom Checklist; Gardner et al., 1999), drug-related problems (Add Health, 2008), and alcohol-related problems (Add Health, 2008). Adolescents also completed questions about peer alcohol and marijuana use and Timeline Follow-Backs (TLFBs) to assess past 90-day marijuana and alcohol use.

A hierarchical linear regression was conducted to test the relative contributions of parenting, peer, and psychological factors on past 90-day cannabis use. Hierarchical logistic regressions were run for any past 90-day alcohol use, drug-related problems, and alcohol-related problems. For all models, demographics (age, race, ethnicity) were included in Step 1. Family factors (monitoring, communication, and conflict) were added in Step 2, followed by peer factors (marijuana use, alcohol use, peer deviance) in Step 3. Finally, individual factors (total, internalizing, and externalizing symptoms) were added in Step 4. For cannabis use, significant predictors included age (β = 6.42, p = .012), peer deviance (β = 2.18, p = .040), and peer marijuana use (β = 7.26, p = .021). For alcohol use, significant predictors were age (OR = 1.84, 95% CI [1.14, 2.96], p = .012), ethnicity (OR = 5.28, 95% CI [1.45, 19.20], p = .012), and parental monitoring (OR =.96, 95% CI [.93, 1.00], p = .041). In terms of alcohol-related problems, significant predictors included parent-child communication (OR =.91, 95% CI [.84, .99], p = .032) and peer alcohol use (OR = 2.05, 95% CI [1.12, 3.78], p = .020). There were no significant predictors of drug-related problems. Results suggest that peer and family factors have distinct relationships to alcohol and marijuana use among truant adolescents. Future research should further explore the role that these factors play in the initiation and ongoing use of substances in this population.