Predictors of trait aggression in bipolar disorder.

OBJECTIVES

Although aggressive behavior has been associated with bipolar disorder (BD), it has also been linked with developmental factors and disorders frequently found to be comorbid with BD, making it unclear whether or not it represents an underlying biological disturbance intrinsic to bipolar illness. We therefore sought to identify predictors of trait aggression in a sample of adults with BD.

METHODS

Subjects were 100 bipolar I (n = 73) or II (n = 27) patients consecutively evaluated in the Bipolar Disorders Research Program of the New York Presbyterian Hospital-Payne Whitney Clinic. Diagnoses were established using the Structured Clinical Interview for the DSM-IV (SCID-I) and Cluster B sections of the SCID-II. Mood severity was rated by the Hamilton Depression Rating Scale (HDRS) and Young Mania Rating Scale (YMRS). Histories of childhood maltreatment were assessed via the Childhood Trauma Questionnaire (CTQ), while trait aggression was measured by the Brown-Goodwin Aggression Scale (BGA).

RESULTS

In univariate analyses, significant relationships were observed between total BGA scores and CTQ total (r = 0.326, p = 0.001), childhood emotional abuse (r = 0.417, p < 0.001), childhood physical abuse (r = 0.231, p = 0.024), childhood emotional neglect (r = 0.293, p = 0.004), post-traumatic stress disorder (t = -2.843, p = 0.005), substance abuse/dependence (t = -2.914, p = 0.004), antisocial personality disorder (t = -2.722, p = 0.008) and borderline personality disorder (t = -5.680, p < 0.001) as well as current HDRS (r = 0.397, p < 0.001) and YMRS scores (r = 0.371, p < 0.001). Stepwise multiple regression revealed that trait aggression was significantly associated with: (i) diagnoses of comorbid borderline personality disorder (p < 0.001); (ii) depressive symptoms (p = 0.001); and (iii) manic symptoms (p < 0.001).

CONCLUSIONS

Comorbid borderline personality disorder and current manic and depressive symptoms each significantly predicted trait aggression in BD, while controlling for confounding factors. The findings have implications for nosologic distinctions between bipolar and borderline personality disorders, and the developmental pathogenesis of comorbid personality disorders as predisposing to aggression in patients with BD.