Remote and in‐clinic digital cognitive screening tools outperform the MoCA to distinguish cerebral amyloid status among cognitively healthy older adults

Abstract

AbstractBackgroundExisting cognitive screening measures fall short in capturing preclinical AD, and cognitive impairment is unrecognized or misdiagnosed in 27‐81% of older adult primary care patients. Digital assessment technology has the potential to deliver more efficient and sensitive cognitive screening, but requires rigorous validation first. We aimed to determine the accuracy of remote (M2C2 mobile app) and clinic‐based (TabCAT and DCTclockTM) digital tests to distinguish between older adults with and without AD pathologic change. We used the Montreal Cognitive Assessment (MoCA) as a reference standard comparison test.MethodsWe recruited 73 cognitively normal participants with Aß PET status (Aß+, n = 25; Aß‐, n = 48, determined by clinical read) from the Butler Hospital Alzheimer’s Prevention Registry (mean age = 69.2 and education = 16.5; 71% female; 89% White). Participants completed M2C2 tasks at home 3 times per day for 8 days, followed by the TabCAT tasks, DCTclockTM, and MoCA at an in‐person study visit. We calculated the area under the curve (AUC) to compare cognitive task accuracy to distinguish Aß status. Multi‐day learning curves were used to examine differences in M2C2 task performance by Aß status over time.ResultsAmong the M2C2 tasks, average performance on the Prices task (episodic memory) over 8 days showed the highest accuracy (AUC = .77) to distinguish Aß status. The Aß+ group tended to perform worse than the Aß‐ group on the Prices task over time (Figure 1). On in‐person screening measures (single time‐point), accuracy to distinguish Aß was greatest for the TabCAT Favorites task (AUC = .76), relative to the DCTclockTM (AUC = .72) and the MoCA (AUC = .71).ConclusionsWe showed that several brief digital screening approaches (memory‐specific tasks on the M2C2 and TabCAT) outperform the MoCA in distinguishing between cognitively healthy individuals with and without elevated cerebral Aβ. Although further validation in community and clinic‐based samples is needed, these results suggest that these digital cognitive assessments may be suitable for more widespread screening to detect early pathological changes in neurodegenerative disorders. Figure 1. M2C2 task performance over time (3 sessions daily for 8 days) by Aβ status.