Incidence, methods and circumstances of suicide in epilepsy: A population-based study in Sweden

Objective

Persons with epilepsy have a higher risk of suicide compared to the general population, but limited data makes it unclear how extensive this risk is and who is most vulnerable. Our study aimed to explore the incidence of suicide among persons with epilepsy in Sweden, and compare to the general population. To facilitate prevention, we also wanted to examine methods and circumstances of suicide.

Methods

We identified all individuals with a diagnosis of epilepsy (ICD G40) in the Swedish Patient Register between 1998 and 2005 who were alive in 2006 (n = 60,952). Among them, 190 cases of suicide were recorded in the National Cause of Death Register during follow-up 2006 to 2011. We reviewed their medical records, death certificates, and autopsy records to validate the cause of death and epilepsy diagnosis, as well as collect information on suicide method and circumstances. After validation we calculated the incidence rates by age and sex and standardized mortality ratios (SMRs) with 95 % confidence intervals (CI).

Results

Overall suicide incidence was 40.0/100,000 person-years (95 % CI 33.0–47.9). Incidence was highest in age 45 to 64 years (61.3, 95 % CI 46.4–79.1) and appeared higher in men than in women with epilepsy. Compared to the general population, individuals with epilepsy had twice the risk of suicide (SMR 2.03 CI 1.67–2.45) and the excess risk appeared more pronounced in women (SMR 2.70 CI 1.92–3.68) than in men (SMR 1.80 CI 1.40–2.26). Intoxication (50 %) was the most common method, followed by hanging, cutting weapons and guns (25 % combined).

Significance

Our results confirm that suicide is overrepresented in individuals with epilepsy, especially in middle age. Incidence was higher in men but the SMR was higher among women, suggesting that the effect of epilepsy on suicide risk is greater for women than for men. Identifying subgroups that are particularly vulnerable is important for suicide prevention.