Abstracts

Rationale: Given the risk of cognitive decline and dementia among older adults with epilepsy, the International League Against Epilepsy (ILAE) Taskforce on Epilepsy in the Elderly recommended routine cognitive-behavioral screening for all older adults. However, there is a paucity of research aimed at validating the ability of brief cognitive screening tools to identify cognitive impairment in older adults with epilepsy. This study evaluated the diagnostic performance of the Montreal Cognitive Assessment (MoCA), a commonly used screen in international research to detect cognitive impairment in older patients (age ≥55) with epilepsy residing in the U.S., using the International Classification of Cognitive Disorders in Epilepsy (IC-CoDE) as the gold standard.


Methods: Fifty older adults with focal epilepsy completed the MoCA and neuropsychological measures of memory, language, executive function, and processing speed/attention. The IC-CoDE taxonomy divided participants into IC-CoDE Impaired and Intact groups. Sensitivity and specificity across several MoCA cutoffs were examined. Spearman correlations examined relationships between the MoCA Total score and clinical and demographic variables and MoCA domain scores and individual neuropsychological tests. The relationship between MoCA domain scores and neuropsychological tests was further examined with a principal component analysis (PCA).


Results: IC-CoDE Impaired patients demonstrated significantly lower scores on the MoCA Total, Visuospatial/Executive, Naming, Language, Delayed Recall, and Orientation domain scores (Cohen’s d range: 0.336-2.77; Figure 1). The recommended MoCA cutoff score < 26 had an overall accuracy of 72%, 88.2% sensitivity, and 63.6% specificity (Figure 2). A MoCA cutoff score < 24 yielded optimal sensitivity (70.6%) and specificity (78.8%), with overall accuracy of 76%. Higher MoCA Total Scores were associated with greater years of education (p=.016) and fewer antiseizure medications (p=.049). The MoCA Memory domain was associated with several standardized measures of memory, MoCA Language domain with category fluency, and MoCA Abstraction domain with letter fluency. A PCA identified five factors accounting for 69.42% of the variance, with individual MoCA domain scores demonstrating high factor loadings with standardized neuropsychological measures.


Conclusions: This study provides initial validation of the MoCA as a useful screening tool for older adults with epilepsy that can be used to guide clinical decision-making in epilepsy and identify patients who may benefit from comprehensive neuropsychological testing. Further, we demonstrate that a lower cutoff (i.e., < 24) better captures cognitive impairment in older adults with epilepsy than the generally recommended cut-off and provides evidence for construct overlap between MoCA domains and standard neuropsychological tests. These findings can inform validity studies in regions of the world where neuropsychology is undeveloped, and there is a need for sensitive and widely available screening measures.


Funding: NIH/NINDS R01 NS120976.