Desensitization to Violence Contributes to Violent Behavior in Adolescence

Exposure to violence, be it in the family, school or community, is a strong risk factor for both internalizing and externalizing problems in adolescence (Fowler et al., 2009; Wilson et al., 2009; Wolfe et al., 2003). However, several studies identified a pattern of “affective desensitization” where very high levels of violence exposure were associated with diminished, rather than increased, depressive symptoms (Gaylord-Harden et al., 2011; Ng-Mak et al., 2004; Mrug et al., 2008). By contrast, the relationship between violence exposure and externalizing problems was linear. Although it has been hypothesized that desensitization to violence would promote more violent behavior, this assertion has not been empirically tested. This study examines whether affective desensitization to violence in early adolescence contributes to perpetration of violence five years later.

The sample included 436 youth (52% male, 80% African American, 20% Caucasian) who participated in a three-wave prospective study of violence (mean ages 11.8, 13.2, and 18.0). At each wave, youth reported on witnessing violence and victimization at home, school/work, and community and delinquent behavior. At wave 2 they also reported on their symptoms of depression. Structural Equation Modeling (SEM) tested a model where violence exposure at waves 1 and 2 predicted delinquency and depression at wave 2, and squared total violence exposure also predicted depression (testing desensitization). In turn, wave 2 delinquency and depression predicted wave 3 violent delinquency, adjusting for violence exposure at wave 3. All paths also adjusted for gender, race, family SES, age, and previous levels of the outcome when available.

The results confirmed linear effects of early adolescent violence exposure on wave 2 delinquency (beta=.17, p<.001) and both linear and curvilinear effects on depressive symptoms (betas=.44 and -.17, both p<.01). Violent delinquency at wave 3 was predicted by higher delinquency (beta=.20, p<.001) and fewer depressive symptoms (beta=-.10, p<.05) at wave 2, as well as higher violence exposure at wave 3 (beta=.30, p<.001). The SEM model had excellent fit (CFI=.98, TLI=.93, RMSEA=.04).

The results showed that high levels of violence exposure in early adolescence were associated with more delinquency but fewer depressive symptoms. In turn, both delinquency and fewer depressive symptoms uniquely predicted more violent behavior five years later. These findings suggest that affective desensitization to violence represents one mechanism through which violence exposure promotes violent behavior. Addressing the emotional effects of violence through counseling may improve subsequent behavioral outcomes. Additional analyses will examine cognitive desensitization and gender differences.