Cause of death and predictors of mortality in a community-based cohort of people with epilepsy

From SUDEP Wiki
Jump to navigation Jump to search

Keezer MR, Bell GS, Neligan A (2016) Cause of death and predictors of mortality in a community-based cohort of people with epilepsy. Neurology. 2016 Feb 23;86(8):704-12.

Link to Article

Abstract: OBJECTIVE: The risk of premature mortality is increased in people with epilepsy. The reasons for this and how it may relate to epilepsy etiology remain unclear. METHODS: The National General Practice Study of Epilepsy is a prospective, community-based cohort that includes 558 people with recurrent unprovoked seizures of whom 34% died during almost 25 years of follow-up. We assessed the underlying and immediate causes of death and their relationship to epilepsy etiology. Psychiatric and somatic comorbidities of epilepsy as predictors of mortality were scrutinized using adjusted Cox proportional hazards models. RESULTS: The 3 most common underlying causes of death were noncerebral neoplasm, cardiovascular, and cerebrovascular disease, accounting for 59% (111/189) of deaths, while epilepsy-related causes (e.g., sudden unexplained death in epilepsy) accounted for 3% (6/189) of deaths. In 23% (43/189) of individuals, the underlying cause of death was directly related to the epilepsy etiology; this was significantly more likely if death occurred within 2 years of the index seizure (percent ratio 4.28 [95% confidence interval 2.63-6.97]). Specific comorbidities independently associated with increased risk of mortality were neoplasms (primary cerebral and noncerebral neoplasm), certain neurologic diseases, and substance abuse. CONCLUSIONS: Comorbid diseases are important causes of death, as well as predictors of premature mortality in epilepsy. There is an especially strong relationship between cause of death and epilepsy etiology in the first 2 years after the index seizure. Addressing these issues may help stem the tide of premature mortality in epilepsy.




Network Graph

Retrieving data for the network graph...