Abstracts

Rationale: Epilepsy is associated with faster cognitive decline and higher mortality rates in people with dementia (PWD). Diagnosis and treatment of epilepsy likely has the potential to reduce morbidity and mortality in PWD. However, there is often a delay in diagnosing seizures in the context of dementia. Identifying predictors of epilepsy in PWD may aid in the timely diagnosis of epilepsy, but these remain understudied. The primary objective of our study is to identify predictors of epilepsy development in PWD.

Methods: This longitudinal, multicenter cohort is based on data of ~45000 participants recruited from 39 Alzheimer’s disease Centers in the US between September 2005 and December 2021. All participants with dementia and mild cognitive impairment (MCI) WITHOUT epilepsy at enrollment and at least 2 visits were included. Those with normal cognition, subjective cognitive impairment, active epilepsy, or remote seizures at enrollment were excluded. The primary outcome was epilepsy development during a follow-up visit. We studied the effect of the following variables on epilepsy development: demographics (age, education, sex, race, ethnicity), cardiovascular risk factors (hypertension, diabetes, hyperlipidemia), neurologic comorbidities (Parkinson’s disease[PD], Traumatic brain injury, stroke or transient ischemic attack[TIA]), cognitive status (age of onset of cognitive decline, severity of dementia, Alzheimer’s disease (AD) vs non-AD subtype of dementia), genetics (dominant AD mutation, APOE4 status), lifestyle factors(alcohol abuse and years of cigarette smoking) and depression. A multivariable survival analysis using the Cox regression model was used to determine the predictors of epilepsy in PWD.

Results: A total of 15,582 participants met the inclusion criteria (Fig1A, B). They had a mean age of 72.56 years with 50.2% (N=7826) female. Of these, 254 (1.6%) developed epilepsy during the follow-up

We conducted a multivariable survival analysis to identify predictors of epilepsy in PWD. Those with APOE4 heterozygous or homozygous allele (adjusted hazard ratio(aHR): 1.35, 95% confidence interval [CI]: 1.03, 1.78], p=0.0289), cognitive decline-onset before the age of 60 (aHR: 2.25, 95% CI: [1.44,3.55], p< 0.0004), AD as the subtype of dementia (aHR: 1.65, 95% CI: [1.15, 2.38], p< 0.0069), severe dementia (aHR: 2.13, 95% CI: [1.86, 2.44], p< 0.001), stroke or TIA (aHR: 1.98, 95% CI: [1.34, 2.92], p=0.0006), and presence of PD [aHR: 2.39, 95% CI: [1.02, 5.58], p=0.0437] were at a higher risk of developing epilepsy(Fig1C & 2) after adjusting for demographics, cardiovascular risk factors, neurologic comorbidities, genetics, cognitive status, and depression.


Conclusions: Our study identified the following predictors of epilepsy in PWD: APOE4 allele, dementia-onset before the age of 60, AD subtype of dementia, stroke/TIA, severe dementia, and PD. PWD with these risk factors may be considered for screening with EEG routinely for early identification and timely management of epilepsy.

Funding: American Epilepsy Society, Alzheimer's Association, NIH