Exploring Oxytocin and Callous-Unemotional Traits as Mediators of the Relationship between Maltreatment and Aggression

Abstract

Recent advances in our understanding of the consequences of childhood maltreatment have offered new insights into the biological and psychological mechanisms that underlie the increased risk for aggression among abused and neglected children; however, the majority of this research has examined reactive (i.e., impulsive) aggression. The processes by which maltreatment increases the risk of proactive aggression are less understood. The present study tested a serial mediation model to explore the potential intermediary roles of oxytocin (OXT) and callous-unemotional (CU) traits in the maltreatment-proactive aggression relationship. Two at-risk samples of young adults (54 probationers/parolees and 47 undergraduate students with histories of adverse childhood experiences) provided saliva samples and completed self-report measures of maltreatment histories, CU traits, and aggression. Separate serial mediation tests of the indirect effects of OXT and CU traits were conducted for each form of maltreatment. As hypothesized, low salivary OXT predicted elevated CU traits and proactive aggression, but not reactive aggression. Analyses supported the primary hypothesis that low OXT and elevated CU traits sequentially mediated the relationship between total maltreatment and proactive aggression. Although emotional abuse and neglect were expected to exert the strongest effects on OXT, serial mediation was only supported for the model of emotional abuse. Specifically, severity of emotional abuse predicted lower levels of salivary OXT, which in turn predicted elevated CU traits, and ultimately more proactive aggression. This study illustrates the role of reduced peripheral levels of OXT in CU traits and proactive aggression that develop in the context of childhood maltreatment. Findings support neurodevelopmental and biosocial theories of psychopathy, which postulate that socio-emotional adversity in early childhood hinders the development of biological systems that are responsible for prosocial emotions. Although the hypothesized serial mediation model was statistically significant, the cross-sectional design precludes our ability to establish causal order. Furthermore, the indirect pathway from maltreatment to proactive aggression through OXT and CU traits only explained a small portion of the total variance in proactive aggression, and should be interpreted as one of several mechanisms that lead to proactive aggression.