Aggression is more strongly associated with suicidality, independent of depression, than emotional instability and impulsivity in people with epilepsy

Abstract

Purpose: We evaluated the relative contributions of emotional instability, impulsivity, and aggression to the presence of suicide risk in people with epilepsy after adjusting for depressive symptoms. Methods: This was a cross-sectional study that used the short form of the Affective Lability Scale (ALS-18), the Barratt Impulsiveness Scale Version 11 (BIS-11), the Brief Aggression Questionnaire (BAQ), and the Patient Health Questionnaire-9 (PHQ-9). Suicidality was assessed using the Mini International Neuropsychiatric Interview (MINI), and the presence of suicide risk was defined as a MINI suicidality score ≥1. Stepwise logistic regression and mediation analyses using a two-stage regression method were performed. Results: Of the 171 subjects (63.2% men) included, suicide risk was present in 38 subjects (22.2%). The stepwise logistic regression analysis identified four variables that were independently associated with suicide risk: higher PHQ-9 score, higher BAQ score, longer duration of epilepsy, and unemployment. The univariate analysis showed that ALS-18 and BIS-11 scores were significantly associated with suicide risk; however, they were backward eliminated from the logistic model according to the criterion of p > 0.1. The mediating effects of ALS-18 and BIS-11 scores on suicide risk via PHQ-9 scores (but not BAQ scores) were significant, with the proportion mediated 61.5% and 54.0% of the total effect, respectively. Conclusions: Alongside depressive symptoms, aggression may be a more useful concept than emotional instability and impulsivity for identifying suicidal risk in people with epilepsy.