Cause-specific mortality in adults with unprovoked seizures. A population-based incidence cohort study

Abstract

PURPOSE: To determine the cause-specific mortality relative to that expected in a population-based incidence cohort of people with unprovoked seizures. METHODS: The cohort comprises 224 inhabitants of Iceland first diagnosed as suffering from unprovoked seizures during a 5-year period from 1960 to 1964. The expected number of deaths was calculated by multiplying person-years of observation within 5-year age categories for each year from diagnosis through 1995 by cause-specific and sex-specific national death rates for those aged 20 years and above. The standardized mortality ratio (SMR) and 95% confidence intervals (95% CI) were calculated. RESULTS: All-cause mortality was increased among men (SMR 2.25, 95% CI 1.56-3.14) but not women (SMR 0.79, 95% CI 0.38-1.46). Among men, there were 8 deaths from accidents, poisoning and violence observed versus 2.82 expected (SMR 2.84, 95% CI 1.22-5.59) and 4 deaths from suicide versus 0.69 expected (SMR 5.80, 95% CI 1.56-14.84). All-cause mortality for men was still elevated after restriction of analysis to those with seizures of unknown etiology (SMR 1.73, 95% CI 1.05-2.67) with the excess deaths attributable to suicide (SMR 5.26, 95% CI 1.06-15.38). Both males and females with remote symptomatic unprovoked seizures had an increase in all-cause mortality due to excess mortality from all cancers, cerebrovascular disease and accidents. CONCLUSION: When compared with the age-, time-period- and gender-specific mortality in the general population, there is excess mortality in men but not women. The increased mortality for men is partly attributable to excess mortality from accidents and suicides.