Differential cued recall memory impairment in mild cognitive impairment due to Alzheimer's disease versus Parkinson's disease

Both Alzheimer's (AD) and Parkinson's disease (PD) are often associated with memory dysfunction, but their pathophysiological underpinnings differ. The current research aimed to differentiate specific profiles of memory impairment due to AD versus PD. We used controlled learning and cued recall paradigm based on the Memory Binding Test (MBT) in 'clinically cognitively normal' controls (CN; n = 161), in patients with amnestic mild cognitive impairment due to AD (AD-aMCI; n = 50) and due to PD (PD-MCI; n = 22), and in PD with normal cognition (n = 18) as based on performance in the neuropsychological battery to prevent circularity in diagnostic decision-making. We applied analysis of covariance (ANCOVA) and Receiver Operating Characteristic (ROC) analysis to determine between-group differences and detection potential of the MBT. We found statistically large between-group differences with worse memory performance in paired cued recall conditions in AD-aMCI<PD-MCI; AD-aMCI<PD-NC; AD-aMCI<CN (p < .001 after Bonferroni correction), and to a lesser extent in PD-MCI<CN (p = .039). However, PD-NC did not differ from PD-MCI, and PD-NC did not differ from CN (p > .050). The detection potential of MBT paired cued recall for differentiating memory impairment in AD-aMCI from CN yielded an AUC of 90% (95% CI, 85-96) and an AUC of 91% (95% CI, 81->99) between AD-aMCI and PD-MCI. Associative memory and binding impairment are most pronounced in AD-aMCI in comparison to PD-MCI and controls. Overall, the MBT is an efficient tool for the differential diagnosis of memory impairment due to the two most common neurodegenerative diseases.