Emotion Dysregulation and Violence in Borderline Personality Disorder

Background and Purpose: Borderline personality disorder (BPD) is a complex and disabling disorder that presents considerable challenges for the mental health treatment system. One of the most problematic features of BPD is dysregulation of emotion, particularly negative affect. We have recently shown that persons with BPD commonly engage in significant interpersonal violence and, given that emotion dysregulation is associated with self harm, we hypothesize that it may be one of the core processes producing the increased risk for interpersonal violence among this population. This is important because such violence can result in repeated hospitalization, incarceration, injury or death, exacting considerable costs in human suffering. Determining the role of emotion dysregulation in incidents of violence among individuals with BPD has important implications for the development of strategies to successfully treat such individuals in the context of community based services.

Methods: As part of the MacArthur Violence Risk Assessment study, 220 individuals with BPD and 581 individuals with other severe mental illnesses were followed in the community for 30 weeks after discharge from psychiatric hospitals in 3 major metropolitan areas. Detailed measures of violence, emotion dysregulation, and antisocial characteristics were collected at study baseline and every 10 weeks post-discharge. Violence data were collected from self-reports and triangulated with collateral informants and arrest records. A series of latent growth curve analyses were constructed to examine individual differences in rates of change in emotion dysregulation and antisocial characteristics throughout the one-year study, and the degree to which such changes mediated the documented risk-enhancing effect of BPD on subsequent interpersonal violence.

Results: Latent growth curve analyses indicated that BPD was associated with greater initial levels of emotion dysregulation at study entry and less favorable trajectories of change in the construct over time, as evidenced by comparatively little post-hospitalization improvement in emotion dysregulation. Mediator analyses indicated that the greater initial levels and lack of reduction in emotion dysregulation over time experienced by those with BPD significantly and completely mediated the relationship between BPD and subsequent violence, with initial level (B = .04, p < .05) and growth parameters (B = .05, p < .05) in emotion dysregulation both making a unique contribution to this effect. These results remained even after adjusting for antisocial characteristics, and estimates of model fit indicated that this growth model provided an adequate representation of the observed data, ÷2(29, N = 801) = 66.24, p < .01, CFI = .94, RMSEA = .04.

Conclusions and Implications: Our findings show that emotion dysregulation is a significant longitudinal mediator of violent behavior among individuals with BPD, and may serve as the primary mechanism that enhances risk for violence among this population. Existing evidence-based treatments, such as dialectical behavior therapy, that teach emotion regulation skills have been successful in reducing self harming behavior in individuals with BPD. Future efforts toward developing treatments to reduce interpersonal violence in this population might begin with examining possible modifications of these existing treatment protocols that would aim to reduce frequency of violence by teaching skills in managing dysregulated emotions.