Examining Predictors of Callous Unemotional Traits Trajectories Across Adolescence Among High-Risk Males

Abstract

Among high-risk youth, those with high levels of callous unemotional (CU) traits show more severe and chronic forms of antisocial behavior. Although ecological models have linked factors across multiple domains of risk to broader antisocial behavior development, fewer studies have adopted this approach in relation to understanding the unique development of CU traits. Further, a paucity of evidence exists from studies that have examined predictors of trajectories of CU traits. In the current study using data from the Pathways to Desistance data set, we examined prospective risk factors for CU traits trajectories modeled from ages 14 to 24. The sample included male adolescents who had interacted with the justice system (N = 1,170). CU traits were assessed using the Youth Psychopathic Traits Inventory. Risk factors were assessed at baseline via youth self-report across multiple domains of risk (individual, parenting, and broader contextual risk). Our results demonstrated higher risk factor scores across individual characteristics (higher anxiety and more substance use), parenting (higher harshness, and lower monitoring and knowledge), and broader contextual risk (more violence exposure) for youth with a “high” and stable CU traits trajectory. Adolescents with stable “high” CU traits likely need interventions capable of addressing and changing multiple aspects of their ecology across individual-, parent-, family-, and community-level targets.

Notes

¹ As an exploratory post hoc test, we examined physiological anxiety within the high CU traits group to determine if the group mean belied primary and secondary groups. We created subgroups within the high CU traits group classifying youth who scored in the lowest 25% as low anxious (n = 48) and youth in the top 25% as high anxious, (n = 45). Providing further support for the existence of a secondary CU traits group, the high anxious, high CU traits group had experienced more neighborhood disorder (p < .01), and marginally greater parental harshness and lower knowledge than the low anxious, high CU traits group (p < .10). Results were similar if we used a median split on physiological anxiety scores to create subgroups. Nevertheless, the post hoc nature of our analyses highlight the need for future studies that examine anxiety subgroups within high CU traits youth, especially using sophisticated analytic approaches, such as model-based cluster analysis (e.g., Kimonis et al., 2012, p. 1094).