TY - JOUR T1 - Suicide after a stroke: a population study JF - Journal of Epidemiology and Community Health JO - J Epidemiol Community Health SP - 863 LP - 866 DO - 10.1136/jech.55.12.863 VL - 55 IS - 12 AU - T W Teasdale AU - A W Engberg Y1 - 2001/12/01 UR - http://jech.bmj.com/content/55/12/863.abstract N2 - STUDY OBJECTIVE To establish whether there are increased rates of suicide after a stroke and the degree to which any increase is related to gender, age at stroke, diagnosis, duration of hospitalisation, and time since stroke. DESIGN Cross linkage of national registers for hospitalisations and causes of death. SETTING The population of Denmark, 1979–1993. PATIENTS A study cohort was defined comprising all 114 098 stroke patients discharged alive from hospital during the period 1979–1993. These patients were then screened in a register of causes of death over the same time period, and 359 cases of suicide were identified. MAIN RESULTS Annual incidence rates, both observed and expected, together with standardised mortality ratios (SMR) were computed based on annual population and suicide statistics, stratified by age and gender. The overall annual incidence rate of suicide in the cohort was 83 per 100 000 per year compared with an expected figure of 45 (difference = 38, 95% CI = 27, 49). Correspondingly, SMR were increased for stroke patients. Across all age bands the SMR for mens was 1.88 (95%CI 1.66, 2.13) and for women 1.78 (1.48, 2.14). SMR were greatest (2.85; 2.17, 3.76) for patients under 50 years of age group and were least for patients 80 years or older (1.3; 0.95, 1.79). There was no clear relation to stroke diagnosis. Suicides were negatively related to duration of hospitalisation, being lowest for those hospitalised for more than three months (0.88; 0.65, 1.19) and highest for those hospitalised for less than two weeks (2.32; 1.92, 2.80). Survival analysis suggested that the risk for suicide is greatest up to about five years after a stroke. CONCLUSIONS Stroke patients are at an approximately doubled risk for suicide. This risk is greater among younger patients and among patients hospitalised for a relatively shorter time. The risk appears to decline with time after a stroke, being greatest within the first five years. ER -