Research has suggested that childhood-onset conduct problems (CPs) are more strongly related to individual predispositions, whereas adolescent-onset CP is more strongly associated with social factors, such as peer delinquency. Neighborhood disadvantage (ND) increases the risk for associating with deviant peers. Thus, peer delinquency could mediate the relationship between ND and adolescent-onset CP. This mediational hypothesis has not been tested previously. We tested this hypothesis in 1,127 justice-involved adolescent males using self-reported delinquency and official arrest records over 3 years after the youth’s first arrest as outcomes. Predictors were self-reported and census-derived indicators of ND and self-reported peer delinquency. Age of onset moderated the associations between self-reported ND and arrests and between self-report of peer delinquency and arrests. In both cases, the association was stronger for those with adolescent-onset CP. Peer delinquency mediated all relationships between ND and CP. Our results also showed some unexpected differences in associations depending on whether self-reported ND or census-derived indicators were used as predictors. Specifically, census-derived ND was negatively related to self-reported offending, which could be due to the use of an arrested sample and the need for youth in more advantaged neighborhoods to show a more severe pattern of antisocial behavior to be arrested.

Given the equivocal literature on the relationship between internalizing symptoms and early adolescent alcohol use (AU) and AU disorder (AUD), the present study took a developmental perspective to understand how internalizing and externalizing symptoms may operate together in the etiology of AU and AUD. We pit the delayed onset and rapid escalation hypothesis (Hussong et al., 2011) against a synthesis of the dual failure model and the stable co-occurring hypothesis (Capaldi, 1992; Colder et al., 2013, 2018) to test competing developmental pathways to adolescent AU and AUD involving problem behavior, peer delinquency, and early initiation of AU. A latent transactional and mediational framework was used to test pathways to AUD spanning developmental periods before AU initiation (Mage = 11) to early and high risk for AUD (Mage = 14–15 and Mage = 17–18). The results supported three pathways to AUD. The first started with “pure” externalizing symptoms in early childhood and involved multiple mediators, including the subsequent development of co-occurring symptoms and peer delinquency. The second pathway involved stable co-occurring symptoms. Interestingly, chronically elevated pure internalizing symptoms did not figure prominently in pathways to AUD. Selection and socialization effects between early AU and peer delinquency constituted a third pathway.