Family processes underlying adolescent substance use and conduct problems: disentangling correlation and causation

PROJECT SUMMARY/ABSTRACT During adolescence, substance abuse and conduct problems are co-occurring and mutually reinforcing problems with significant acute and long-term burden to adolescents, their families, and society. Family-based treatments and prevention programs can reduce their incidence, but these interventions are limited in the size, duration, and scope of effects and can be cumbersome to implement. One factor limiting the refinement of these existing interventions is the difficulty of conducting rigorous science about the specific causal effects of the family factors that are targeted as mechanisms of change. Factors such as parental monitoring, parental warmth, and family conflict covary with each other as well as many other variables (e.g., genes, socioeconomic status, living environment, parent psychopathology), so it is difficult to isolate their unique impacts on adolescent outcomes from the larger causal network of risk factors. This proposal pursues a causally grounded, mechanistic approach to disentangle the impacts of family processes and characterize how they vary by the severity, chronicity, and developmental timing of exposure at the population level. We draw data from the Adolescent Brain and Cognitive DevelopmentSM (ABCD) Study—a diverse, longitudinal, nationwide, population-based cohort of 11,880 youth ages 9/10 years old followed prospectively until ages 19-20, with simultaneous measurement of many intercorrelated aspects of development at biannual assessments throughout. We will attempt to isolate the effects of parental monitoring, parental warmth, and family conflict from other influences using three distinct, and complementary approaches: (A) a between-subjects approach using marginal structural models; (B) a within-subjects approach using the fixed effects panel estimator; (C) a twin/sibling discordancy approach using 336+ pairs of monozygotic twins, 430+ pairs of dizygotic twins, and 1,805 non-twin siblings. We pursue four specific aims. Aim 1: Determine the causal impacts of three family processes—parental monitoring, parental warmth, and family conflict—on proximal adolescent substance abuse and conduct problems between ages 11 to 18 years. Aim 2: Determine whether the effects of the family processes vary by youth sex, racial/ethnic identification, or predisposition to substance abuse and conduct problems. Aim 3: Determine whether the effects of three family processes vary by severity, chronicity, or developmental timing of exposure. Altogether, this research will yield a more scientifically accurate developmental psychopathology of adolescent substance abuse and delinquency and thereby inform the design, refinement, and personalization of family-based interventions to reduce the incidence of these problems at the population level.