Predicting the Rate of Decline in Early Alzheimer Disease: the Role of Neurocognitive Performance Features
Alzheimer disease (AD) is a neurodegenerative disorder that characteristically begins with episodic memory impairment followed by other cognitive deficits over time; however, the course of illness varies, with significant variability in terms of the rate of cognitive decline across affected individuals. Several studies have examined demographic, clinical, biological, and neurocognitive performance markers to predict rate of AD progression, but findings are mixed. The current study utilized neurocognitive performance features along with disease-specific and health features to determine the best prediction model for the rate of future cognitive decline in subjects with mild AD. Ninety-six subjects with mild AD at baseline were administered a comprehensive battery of neurocognitive tests and clinical measures. Based on Clinical Dementia Ratings (CDR) of functional and cognitive decline within two years, subjects were determined to be Faster (n = 45) or Slower Progressors (n = 51). Stepwise logistic regressions using neurocognitive performance features, disease-specific, health, and demographic variables were performed in a hierarchical fashion to determine optimal predictors of rate of progression. Several individual neurocognitive measures distinguished Faster from Slower Progressors at baseline, including Trail Making Test - A, Digit Symbol, California Verbal Learning Test (CVLT) Total Learned, CVLT Primacy Recall, and the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) Neuropsychological Battery Total Score. No disease-specific, health, or demographic variables predicted rate of progression; however, history of cardiac illness showed a trend. In a stepwise logistic regression of neurocognitive performance features alone, a combination model of three measures (Trail Making Test - A, Semantic Fluency, and CERAD Total) distinguished Faster from Slower Progressors with 76% accuracy. In an omnibus model including neurocognitive, disease-specific, health, and demographic variables, only Trail Making Test - A distinguished groups (68% correct classification). Several neurocognitive performance features may play a role in predicting rate of decline in mild AD. Notably, three relatively brief and commonly used measures were found to predict differences in rate progression with good accuracy. Results from the current research provide important advances in understanding the role of neurocognitive measures in predicting rate of decline in AD.