(Dis)continuity of violent behaviors during the transition to adulthood: The effects of stress and coping resources

Background and Purpose: Both victimization and violent behavior peak in adolescence and early adulthood (Nash & Kim, 2007), with those carrying violence propensities into young adulthood substantially more likely to experience grave outcomes such as arrest, incarceration, serious injury, and barriers to employment and educational attainment (Mun et al., 2008). Recent work is increasingly pointing to stress, emotional impairment, and protective resources (or lack thereof) in the childhood etiology of violence (Logan-Greene et al., 2011). This paper extends this work, testing the predictive utility of theory-guided, multi-level assessments of stress and coping in distinguishing individuals who manifest Desisting, Increasing, Persisting, or Always Low patterns of violent behaviors during the transition into adulthood. This research addresses a gap in lifecourse stress assessment, informing prevention efforts of malleable targets to reduce the prevalence of violence during this critical time.

Methods: This previously described NIMH-funded longitudinal study surveyed high school youth in two metropolitan areas, using sampling protocols based on school records and teacher reports (Nurius et al., 2008). An ethnically diverse sample selected on the basis of risk of school drop-out was followed for 3 timepoints from adolescence into young adulthood (6 years later) with satisfactory retention rates. Participants completed nurse-delivered surveys using established measurement scales with excellent psychometric properties spanning three domains: protective resources, emotional distress, and interpersonal stressors. Assessment in the transition to adulthood (roughly midway in the 6 year span) was used to predict patterns of violence in later adulthood in relationship to adolescent violence levels.

Results: Multinomial regressions successfully predicted membership in the four groups with strong fit statistics and explanation of variance (χ2=190.83, p<.001; Nagelkerke pseudo-R2=0.47). As hypothesized, all three domains contributed significantly to the model. Particularly strong predictive effects were seen with emotional distress and interpersonal stressors. Specifically, higher levels of anger predicted membership in the Increasing and Persisting groups, while higher levels of hopelessness predicted membership in the Increasing group relative to the Persisting group. Within the interpersonal stressors domain, lower levels of family disruption predicted membership in the Desisting and Persisting groups relative to the Always Low group, and higher levels of victimization had strong predictive effects for all group contrasts in the expected directions.

Conclusions and implications: This paper applies an innovative approach to developmental analysis of stress and coping resource contributions in predicting different trajectories of violence. Rather than a global “aging out” trend, findings demonstrate a robust predictive value of this multidomain that distinguishes separate pathways, better illuminating the roles of emotional distress and interpersonal stressors as contributors to violence at this developmental juncture. This is congruent with theories of violence positing mechanisms though which stressful exposures, such as child abuse, lead to emotional impairment and reactivity that foster violence. At present, emotional health and stressful events are often poorly addressed within juvenile and adult corrections contexts. Successful interventions for violent behaviors among youth must attend to these issues in judicial, clinical, and preventive settings (Farmer et al., 2007).