Abstracts

Rationale: There is an emerging, bidirectional association between dementia, and particularly Alzheimer’s Dementia (AD) and epilepsy. Data suggest that amyloid is pro-epileptogenic, and certain mutations found in early onset AD (presenilin 1/2 and APP) have been linked to seizures. Conversely, case series of patients with dementia without history of clinical epilepsy show evidence of subclinical hippocampal seizures, raising the possibility of hippocampal hyperexcitability contributing to the pathogenesis of AD. The prevalence of both epilepsy and dementia increases with increasing age and their incidence is highest among those over the age of 65. The elderly population is estimated to reach 100 million in the US alone by 2060, therefore, understanding the link between these two conditions is of significant public health importance. The goal of this epidemiological study was to define the association between dementia and epilepsy in a large, population-based cohort, the Framingham Heart Study (FHS). Methods: The study population included participants of the original (n=5,209) and offspring (n=5,124) FHS cohorts. Surveillance for dementia at FHS is ongoing and based on consensus review that considers neuropsychological testing, neurologist exams, family interviews and brain autopsy data. We performed a broad screen for participants with potential epilepsy. We used FHS collected data (self-reported history of seizures, history of cerebral tumor, brain injury, 345.xx and 380.4x ICD-9 codes and antiepileptic drug (AED) use), as well as CMS data claims (ICD-9 codes, AED use) for those participants enrolled in CMS. We identified 533 participants and their medical records were reviewed in detail. Consensus review by two epileptologists based on definitions of epilepsy proposed by the International League Against Epilepsy yielded 182 epilepsy cases. A nested case-control design was used for testing the association in each direction; participants free of epilepsy at the time of dementia diagnosis with 3 cohort, age-, and sex-matched controls were followed for incident epilepsy. Conversely, dementia-free participants at the time of epilepsy diagnosis with 3 cohort-,age-, and sex-matched controls were followed to at least age 65 for incident dementia. Cox proportional hazard regression was used for statistical analysis. Results: Among 860 dementia cases (mean age 82, 65% women) matched to 2,580 controls, there were 68 incident epilepsy cases and the entry age and sex adjusted cox regression yielded HR=1.85 (1.12-3.04), p=0.016. In the group of 78 epilepsy cases (mean age 64, 60% women) matched to 234 controls, there were 76 incident dementia cases. Cox model adjusted for age at entry and sex yielded HR=2.06 (1.27-3.32), p=0.003. Conclusions: Our study offers evidence from a large, population-based, cohort that there is a bidirectional association between epilepsy and dementia; namely, there is a twofold risk of developing either condition in the presence of the other.The FHS cohort can be instrumental in defining the epidemiology of these two conditions and we aim to leverage the rich FHS database to elucidate potential biological pathways linking dementia to epilepsy. Funding: FHS NS017950