Authors :
Presenting Author: Jessica Kania, BS – University of Wisconsin-Madison
Anny Reyes, PhD – Cleveland Clinic
Victoria Williams, PhD – University of Wisconsin School of Medicine and Public Health
Dace Almane, MS – University of Wisconsin-Madison
Ifrah Zawar, MD MS-CR – University of Virginia Heatlh
Natalia Mendez, BS – University of California at San Diego
Kayela Arrotta, PhD – Cleveland Clinic
Rani Sarkis, MD, MSC – Massachusetts General Brigham
Lisa Ferguson, MA – Cleveland Clinic
Vineet Punia, MD – Cleveland Clinic, Cleveland, OH, USA
Robyn Busch, PhD – Cleveland Clinic
Jana Jones, PhD – University of Wisconsin–Madison
Bruce Hermann, PhD – University of Wisconsin-Madison
Carrie McDonald, PhD – University of California, San Diego
Rationale:
Rationale: The increasing incidence of late-onset epilepsies and aging in individuals with chronic or remitted epilepsy raises critical questions about long-term care, particularly in light of emerging data showing neurodegenerative biomarker-positivity in these populations. Efficient cognitive assessments will be essential for tracking cognitive changes and evaluating pharmacologic interventions, should biomarker-guided care become standard. The Alzheimer’s Disease Assessment Scale-Cognition (ADAS-Cog) is a widely used tool in AD trials, but its utility in people with epilepsy (PWE), who are often excluded from such studies, remains unclear. This study evaluates the sensitivity, specificity, and clinical relevance of the ADAS-Cog in older PWE.
Methods:
Methods: From the Brain, Aging, and Cognition in Epilepsy (BrACE) study, 83 older adults with focal epilepsy (mean age=66.4 years, 57% female) underwent neuropsychological assessment including administration of the ADAS-Cog 13. The International Classification of Cognitive Disorders in Epilepsy (IC-CoDE) assigned participants to intact vs. abnormal cognition groups based on deficits in at least one cognitive domain. The ADAS-Cog 13 cutoffs for mild cognitive impairment (MCI) were assessed using sensitivity and specificity analyses. Spearman correlations and independent t-tests examined the relationship of ADAS-Cog 13 total scores to socio-demographic and clinical epilepsy variables. Age and education matched controls (n=83) were obtained from the ADNI cohort. Independent t-tests assessed ADAS-Cog performance between IC-CoDE memory domain subgroups and between BrACE participants and ADNI controls. Results:
Results: Of the 83 BrACE participants, 32 (38.6%) were classified as IC-CoDE impaired and 19 (22.9%) had at least memory domain impairment. IC-CoDE impaired PWE had significantly higher (worse) ADAS-Cog 13 total scores than intact cognition PWE (t=-3.172, p< 0.001). An ADAS-Cog cutoff score of ≥15 displayed the highest sensitivity (68.8%) and specificity (66.7%) with an overall accuracy of 67.5%. Males had significantly higher (worse) ADAS-Cog total scores than females (t=-1.986, p=0.025). Years of education, age, and duration of epilepsy were not significantly correlated with ADAS-Cog total scores. IC-CoDE-memory impaired PWE had significantly worse ADAS-Cog 13 total (p< 0.001) and ADAS-Cog delayed recall scores (p< 0.001). ADAS-Cog 13 total scores were significantly worse in BrACE participants than ADNI controls (t=-7.228, p< 0.001) and across most ADAS-Cog test items except commands, constructional praxis, ideational praxis, orientation, and language.