Bi-directional associations of epilepsy with dementia and Alzheimer's disease: a systematic review and meta-analysis of longitudinal studies

Objective To assess the bi-directional associations of epilepsy with dementia and Alzheimer's disease (AD). Methods We searched PubMed, Embase and the Cochrane Library for longitudinal studies assessing the associations of epilepsy with dementia and AD up to 4 August 2021. Two authors independently extracted study characteristics, exposures, outcomes and covariates. Summary hazard ratios (HRs) and 95% confidence intervals (CIs) were pooled using a random effects model. Results From 8,545 articles identified in the initial research, 27 publications describing 20 longitudinal studies were included in the final analyses. There were 10 studies on epilepsy predicting risk of dementia, 5 studies on epilepsy predicting risk of AD, 11 studies on dementia predicting risk of epilepsy, and 6 studies on AD predicting risk of epilepsy. Baseline epilepsy was associated with higher risk of dementia (pooled HR 2.00; 95% CI 1.73-2.33) and AD (pooled HR 1.81; 95% CI 1.19-2.75). The pooled HRs for epilepsy associated with baseline dementia and AD were 2.91 (95% CI) 2.11-4.01) and 3.11 (95% CI 2.47-3.90), respectively. These positive associations persisted in sensitivity and subgroup analyses. Conclusions Our findings suggested positive and bi-directional associations of epilepsy with dementia and AD. However, these associations should be carefully interpreted due to the presence of substantial heterogeneity, and they need to be verified in additional high-quality studies.