Abstracts

Rationale: The prevalence of epilepsy is approximately 0.5%-1% in the general population. Stroke is known to cause epilepsy. However, epilepsy precedes stroke in some patients but the relationship is not clear. Seizures may be the initial presentation of subclinical cerebrovascular disease. Conversely, epilepsy and stroke may be independent but share similar risk factors. We hypothesize that individuals with epilepsy onset later in life with no history of stroke will have more stroke than those with lower extremity fracture. Methods: Using legally mandated healthcare data, we identified all non-federal South Carolina hospital and emergency department encounters from 2000 through 2011. Cases were those with a diagnosis of epilepsy. Controls were persons with lower extremity fracture (fracture of the tibia, fibula or ankle) representing an otherwise healthy population. The outcome of interest was stroke. All encounters were searched for a diagnosis of stroke more than 6 months after epilepsy or lower extremity fracture. Stroke from sub-arachnoid hemorrhage, which is known to be strongly associated with epilepsy, was excluded. Comorbid conditions suspected to confound the association with stroke, such as hypertension and myocardial infarction were also identified. Individuals under age 35 and those with a previous stroke diagnosis were excluded. Differences in the distribution of characteristics were assessed using descriptive statistics; proportions were compared using 95% confidence intervals. The time from the first epilepsy or lower extremity fracture encounter to the date of stroke, death or the end of the study was calculated and the association of the independent variables with stroke was assessed using Cox Proportional Hazard techniques. Results: 112,280 individuals without a prior diagnosis for stroke were seen in hospital or emergency department for epilepsy (70,258) or lower extremity fracture (42,022). Of those, 6,954 (9.9%) of those with epilepsy and 2,406 (5.7%) of those with lower extremity fracture suffered a subsequent stroke. The adjusted hazard of stroke for those with epilepsy was 1.60 (95% CI 1.53-1.68). The adjusted hazard of stroke was higher in African-Americans, those aged over 55 at first admission, those covered by Medicare or Medicaid, and in individuals with a history of hypertension, myocardial infarction, diabetes, hyperlipidemia, arteriosclerosis, and alcohol abuse. Conclusions: Those with epilepsy were 60% more likely to have subsequent stroke than controls even after adjustment for multiple factors including the presence of comorbid conditions. Possible explanations of this increased risk are adverse effects of medications, subclinical cerebrovascular conditions, or a sedentary lifestyle. Regardless of the underlying cause, clinicians should be aware of the increased stroke risk and consider aggressively treating known cerebrovascular risk factors in those with epilepsy.