Homicide and bipolar I disorder: A 22-year study

Jeong Hoon Yoon, Jeong Hyun Kim, Sang Sub Choi, Mi Kyung Lyu, Jee Hyun Kwon, Yun Ik Jang, Gun Tae Park

Research output: Contribution to journalArticlepeer-review

18 Scopus citations

Abstract

Background: Non-serious offenses in manic phase have been mainly studied in patients with bipolar disorder. However, some authors reported that depressive phase is related with the violent and homicidal manifestations of bipolar disorder. Aims: We investigated the characteristics of homicide by the polarity of mood episode in patients with bipolar I disorder. Methods: Among the offenders who were sentenced to undergo treatment at the National Institute of Forensic Psychiatry from October 1987 to January 2008, a total 219 offenders whose final diagnoses were bipolar I disorder based on DSM-III-R and DSM-IV were selected. Retrospective medical chart review was performed for characteristics of mood episodes. Descriptions of offenders were supplemented by review of the written records of the police or prosecutors. Results: The general rate of total offense was higher in the manic phase than in the depressive phase (86.8% vs. 13.2%). However, the rate of homicide was higher in the depressive phase than in the manic phase. The victims of homicide were more likely to be family members of the patients in depressive phase than in manic phases (96.2% vs. 63.9%, p= 0.001). However, parricide was committed only in manic phases. Altruistic motivation of homicide was significantly higher in depressive phase (34.6% vs. 0%, p< 0.001) whereas impulsivity was the most common one in manic phases. Conclusions: The risk of offenses, particularly homicide for family members, should not be overlooked in the depressive phases of bipolar I disorder.

Original languageEnglish
Pages (from-to)113-118
Number of pages6
JournalForensic Science International
Volume217
Issue number1-3
DOIs
StatePublished - 10 Apr 2012

Keywords

  • Bipolar depression
  • Homicide
  • Mania
  • Offense
  • Polarity

Cite this