Historically, empathy has been thought to motivate prosocial behaviour and inhibit aggressive behaviour. Contrary to current assumptions and theoretical support, a recent meta-analysis revealed a small effect of empathy on aggression among adults (Vachon, Lynam, & Johnson, 2013). The current study sought to determine whether broadening the focus from empathy to include other socially relevant affective characteristics (such as in CU traits) was advantageous in predicting aggressive behaviour. As little is known about the strength of this association among youth, the current study meta-analytically examined 192 unique effect sizes drawn from published and unpublished studies reporting on samples of children and adolescents. Analyses were conducted across general, cognitive, and emotional empathy, as well as callous-unemotional traits, and general, direct, indirect, proactive, and reactive aggression. Significant variability was noted across effect sizes. Consistent with a recent meta-analysis involving adults (Vachon et al., 2013), small to moderate associations were identified between aggression and traditional measures of empathy (i.e., general, emotional, cognitive); these effects ranged from r = −0.06 to −0.26. Among broader measures of emotional style (i.e., callous-unemotional traits), moderate to large effects were found; ranging from r = 0.30 to 0.37. Results suggested that broader affective measures may be more strongly associated with aggression than empathy alone. The results raise questions about the nature of empathy assessment and indicate the utility of targeting multiple emotion-related factors during treatment to effectively reduce aggressive behaviour. In particular, the results underscore of the importance of considering the limited prosocial emotions specifier (perhaps trans-diagnostically given the varied nature of the sample) when considering implications for prognosis and treatment targets.
The prevalence of disorders characterized by aggression (e.g., conduct disorder, oppositional defiant disorder) is alarmingly high among children and adolescents (approximately 10%; Angold & Costello, 2001; Nock, Kazdin, Hiripi, & Kessler, 2007). Youth aggression has demonstrated continuity into adulthood and has been associated with several negative outcomes later in life (e.g., antisocial behaviour in adulthood, work/school problems, substance use, physical and mental health concerns; e.g., Frick, Ray, Thornton, & Kahn, 2014; Frick & Viding, 2009; Karantanos, 2012; Moffitt, 2018). Given this problematic trajectory, early detection and effective interventions are imperative to preventing disruptive patterns of behaviour in adulthood. While deficits in empathy have been incorporated into the criteria used to diagnose disruptive behaviour disorders (American Psychiatric Association, 2013) and have been identified as a potential treatment target (Frick & Kemp, 2021), empirical support for an association between empathy and aggression is mixed. Disagreement surrounding the relevance of these constructs may be fueled by variations in the conceptualization of empathy (broader versus more narrowly defined) and type of aggression assessed (Vachon et al., 2013; Frick & Kemp, 2021). As such, the true magnitude of association remains unclear. To help address this gap in the literature, the current study sought to clarify the association between subtypes of empathy and aggression, and to examine whether broader affective measures (i.e., callous unemotional [CU] trait measures) predict aggression above and beyond traditional empathy measures.
The current meta-analysis provides evidence of an association between empathy and aggression among children and adolescents. With 188 unique studies, the current meta-analysis provides the most comprehensive review of the association between empathy, CU traits, and the forms and functions of aggression among youth to date. Consistent with adult samples (Vachon et al., 2013), only a small to moderate association was found between traditional measures of empathy (i.e., general, emotional, cognitive empathy) and aggression in children and adolescents. By considering multiple types of empathy and aggression, and expanding beyond the traditional definition of empathy to include indices of emotional style indexed via CU traits (e.g., lack of guilt/remorse, shallow affect), we were able to, for the first time, provide evidence that broader measures of emotional style appear to be better predictors of aggressive behaviour than empathy alone; where moderate to larger correlations were noted between CU trait measures and aggression. These findings suggest consideration of a constellation of affective traits, rather than empathy deficits exclusively, is important to the prediction of aggression among children and adolescents.
Although contrary to contemporary theory, the ostensibly small to moderate effect between traditional measures of empathy and aggression in itself does not dismiss the potential relevance of considering empathy in relation to risk for aggression. Instead, the current study highlights the utility of targeting multiple affective features in treatment to effectively reduce aggressive behaviour. In the DSM-5 (APA 2013), the Limited Prosocial Emotions specifier was incorporated into the conduct disorder diagnosis. This work underscores the importance of considering these traits in the diagnosis and treatment of not only conduct disorder, but also trans-diagnostically given the varied nature of the sample. As effect sizes estimated in the current meta-analysis are reflective of a single time-point, it is also important to acknowledge that continuous dynamical interactions may substantially intensify the effects when compounded over time. Before dismissing the importance of empathy to aggression, one must consider the function of empathy both within a larger constellation of affective traits, and a broader range of socially relevant behaviours.