Abstracts

Rationale: Late-onset unexplained epilepsy (LOUE) is associated with both cognitive impairment and functional changes . However, predictors of functional decline in this population remain underexplored . This study aims to investigate the utility of the Functional Activities Questionnaire (FAQ) in individuals with LOUE and its relationship with cognitive performance, and epilepsy related variables.

Methods: This cross-sectional study recruited participants with at least one unexplained seizure onset after age 55 and a nonlesional MRI at Brigham and Women’s Hospital. Inclusion criteria were seizure onset within 5 years and stable anti-seizure medication (ASM) use for at least 4 weeks. Exclusion criteria included non-English speakers, neurodegenerative diseases, Clinical Dementia Rating (CDR) global score ≥ 1.
Demographic and clinical variables extracted included: years of education, number of ASMs, drug refractoriness and epilepsy duration. The cognitive battery included the Mini Mental Status Exam (MMSE), the extended Preclinical Alzheimer’s Cognitive Composite (PACC5) and the FAQ. Z scores were generated using the Harvard Aging Brain Study participants’ performance as reference. Delayed verbal recall (DVR) scores were derived from the PACC5 components (Wechsler Logical Memory delayed recall and Free and Cued Selective Reminding Task). A cut-off score of ≥ 6 on the FAQ , was considered as indicating impairment on instrumental activities of daily living.

Results: 85 participants were enrolled in the study with a mean age at testing of 71.14 ± 7.26 years, 49.4 % Females, mean years of educational 15.71 ± 2.89 years, median number of ASMs 1 (range: 0-3), and time since seizure onset of 1.77 ± 1.59 years. Cognitive performance scores consisted of: MMSE Z-score (mean −0.66 ± 1.91), PACC5 Z-score (mean −0.70 ± 1.18), and the DVRZ Z-score (mean −1.05± 1.56). Functional impairment, measured by the FAQ, was prevalent in 12.05% of participants, with the mean FAQ score being 2.17± 4.91. The most impacted functional domains were memory-related tasks, with FAQ domains “remembering dates” and “travel” showing the highest number of participants with impairment (FAQ scores of 2 or 3). LASSO regression identified worse DVRZ performance, higher ASM number, and drug refractoriness as key predictors of functional impairment. Specifically, poorer DVRZ performance was associated with increased odds of impairment (OR=1.72, 95% CI: 1.15, 2.56), along with higher ASM number (OR = 3.41 [95% CI: 1.32, 8.81], p = .012) and being drug refractory (OR = 7.50 [95% CI: 1.60, 35.17], p = .010).

Conclusions: This study identifies key clinical factors that predict functional impairment in individuals with LOUE. Specifically, a higher number of anti-seizure medications (ASM) and drug refractory epilepsy (MedRef) and worse delayed verbal recall (DVRZ) performance were associated with increased odds of functional decline. These findings highlight the FAQ as a useful clinical tool to screen individuals with LOUE and emphasize the importance of minimizing the burden of antiseizure medications in this patient population. Additionally, individuals with drug-refractory LOUE are at risk of functional decline.

Funding: No funding received.