Sleep-related violence

Abstract

We hypothesized that sleep-related violent behavior associated with parasomnias occurs as the result of a diathesis and is precipitated by stressors and mediated by disturbed nonrapid eye movement (NREM) sleep physiology. Sixty-four consecutive adult patients (mean age 30 years) who were investigated for sleepwalking or sleep terrors were categorized according to clinical history into three groups: serious violence during sleep to other people or to property or self (n = 26); harmful, but not destructive behavior (n = 12); and nonviolent behavior (n = 26). Log linear analysis showed that a diathesis (childhood parasomnia and/or family history of parasomnia) and a stressor (psychologic distress, substance abuse and sleep schedule disorder) predicted the presence of sleepwalking or night terror. Serious violent acts were more likely to occur with males (p < 0.004) who showed sleep schedule disorder (p < 0.03). Both harmful and serious violent sleep behavior occurred with drug abuse (p < 0.009). In comparison to all other groups, those who were violent to other people were males who experienced more stressors (p < 0.02), drank excessive caffeinated beverages, abused drugs (p < 0.03) and showed less stage 4 sleep (p < 0.02) and less alpha (7.5-11 Hz) electroencephalogram NREM sleep (p < 0.02) on polysomnography. Being male and having < 2% stage 4 sleep provided 89% sensitivity, 80% specificity and 81% diagnostic accuracy for individuals who were violent to others. The forensic implications of these findings are discussed.