Long-term mortality risk by cause of death in newly diagnosed patients with epilepsy in Finland: a nationwide register-based study.

To estimate long-term mortality by cause of death in a nationwide, register-based cohort of newly diagnosed patients with epilepsy (PWE). All noninstitutionalized Finnish PWE aged 10-74 years (n = 10,818) eligible for reimbursement for antiepileptic medication for the first time between 1990 and 1994 were identified in the database of Social Insurance Institution of Finland. Mortality was compared against a population-based reference cohort (n = 43,894). Hazard ratios (HR) and their 95 % confidence intervals (95 % CI) during a follow-up of 18 years were estimated using proportional hazards modeling. Potential years of life lost (PYLL) and excess fraction of causes of death attributable to epilepsy were estimated. PWE contributed 137,610 person-years of observation and there were 3,558 deaths. Mortality remained elevated up to 18 years post-diagnosis (HR 3.21, 95 % CI 3.07-3.35). Ischemic heart disease mortality in PWE was two-fold (HR 2.31, 95 % CI 2.09-2.54), and remained constantly elevated during entire follow-up in both men and women. Most premature mortality in terms of PYLL was attributable to brain cancer (17 %), other cancers (15 %), ischemic heart disease (11 %), as well as cerebrovascular diseases (10 %). The percentage of deaths in PWE statistically attributable to epilepsy was 3.9 % for accidents, 3.4 % for alcohol-related diseases, and 1.6 % for suicides. PWE had substantial excess mortality from non-communicable diseases, which did not disappear by 18 years. Diseases of the circulatory system and cancers, especially brain cancer, were the most important causes of death almost regardless of the mortality indicator.