Examining associations between monitoring-related parenting phenotypes and youth externalizing using a discordant sibling random-intercept cross-lagged panel model

Abstract

In this registered report, we introduce a new discordant sibling random-intercept cross-lagged panel model and apply it to investigate within-person/family and between-family bidirectional associations of parental monitoring-related phenotypes and youth externalizing behaviors. We used data from 1,956 twin/sibling pairs (Mage at first assessment = 10 years; 51% male, 59% White; 14% Black; 15% Hispanic; <1% Asian; 11% Other race/ethnicity) contributing up to six annual waves of data collection from the adolescent brain and cognitive development study, a nationally representative data collection effort including 11,875 children enrolled and assessed across 21 sites across the United States. DS latent curve models with structured residuals models were needed to account for change over time, converged, and fit the data well. We found few within-person/family effects consistent with hypotheses, concentrated from age 13–15 that were likely environmentally mediated. Potentially causal effects within-person/family effects included more youth-reported child disclosure at age 13 predicting lower parent-reported externalizing at age 14, and more youth externalizing problems at age 13 with less monitoring at age 14; other within-person/family associations were driven by familial confounding. Unexpected positive between-family lagged effects in the same developmental window reflect novel family-level processes that may be driven by evocative gene-environment correlations.