Suicide and Attempted Suicide in Bipolar Disorder: A Systematic Review of Risk Factors

Keith Hawton, D.Sc., F.R.C.Psych.; Lesley Sutton, M.Sc.; Camilla Haw, M.R.C.Psych.; Julia Sinclair, M.R.C.Psych.; and Louise Harriss, M.Sc.

Objective: To determine the main risk factors for suicide and nonfatal suicidal behavior in patients with bipolar disorder through a systematic review of the international literature.

Data Sources: Studies were identified through electronic searches of MEDLINE (1966-December 2003), EMBASE (1980-December 2003), PsycINFO (1872-November 2003), and Biological Abstracts (1985-December 2003) using index and free-text search terms for bipolar disorder, bipolar depression, manic depression, mania, and affective disorders; combined with terms for self-harm, self-injury, suicide, attempted suicide, automutilation, self-mutilation, self-poisoning, and self-cutting; and combined with terms for risk, case control, cohort, comparative, longitudinal, and follow-up studies. No language restrictions were applied to the search.

Study Selection: Included studies were cohort, case-control, and cross-sectional investigations of patients with bipolar disorder in which suicide (13 studies) or attempted suicide (23 studies) was reported as an outcome. The selected studies also used diagnostic tools including the DSM, International Classification of Diseases, and Research Diagnostic Criteria.

Data Synthesis: Meta-analysis of factors reported in more than 1 study identified the main risk factors for suicide as a previous suicide attempt and hopelessness. The main risk factors for nonfatal suicidal behavior included family history of suicide, early onset of bipolar disorder, extent of depressive symptoms, increasing severity of affective episodes, the presence of mixed affective states, rapid cycling, comorbid Axis I disorders, and abuse of alcohol or drugs.

Conclusions: Prevention of suicidal behavior in patients with bipolar disorder should include attention to these risk factors in assessment and treatment, including when deciding whether to initiate treatment aimed specifically at reducing suicide risk.

(J Clin Psychiatry 2005;66:693-704)

Received Nov. 19, 2004; accepted Feb. 1, 2005. From the Centre for Suicide Research, University of Oxford Department of Psychiatry, Oxford, United Kingdom.

This study was supported by grant number SEO 016 from the National Health Service Executive South East Region Research Committee.

Dr. Hawton is supported by Oxfordshire Mental Healthcare Trust, Dr. Haw by St Andrew's Hospital, Northampton, and Dr. Sinclair by the Medical Research Council. Mss. Sutton and Harriss report no significant individual commercial relationships relevant to the study.

The authors thank Heather Wilder, M.I.S., for assisting with the literature search; G. Brown, Ph.D., E. Hansen Høyer, Ph.D., E. King, Ph.D., M. Oquendo, M.D., and R. Tavcar, M.D., for supplying us with additional data from their studies; and J. Geddes, M.D., E. Hansen Høyer, Ph.D., and K. Jamison, Ph.D., for checking the list of identified studies.

Corresponding author and reprints: Keith Hawton, D.Sc., F.R.C.Psych., Centre for Suicide Research, University of Oxford, Department of Psychiatry, Warneford Hospital, Oxford, OX3 7JX, UK (e-mail: keith.hawton@psychiatry.ox.ac.uk).