Validation of a web-based cognitive test for early detection of Alzheimer's disease

BackgroundEarly detection of Alzheimer's disease (AD) is critical for effective disease management and treatment. Web-based assessment tools offer advantages by enabling broader accessibility and reducing reliance on specialized clinical infrastructure.ObjectiveThis study aimed to validate a self-administered, web-based cognitive assessment tool for AD screening.MethodsA total of 106 older adults aged 55 to 84 years were recruited and clinically classified as cognitively unimpaired (CU, n = 35), amnestic mild cognitive impairment (aMCI, n = 37), or AD (n = 34). Participants completed Cogscreen, a 10-min web-based cognitive test comprising verbal cued memory and digit symbol substitution tasks.ResultsBoth the verbal cued memory and digit symbol substitution tasks showed significant score differences among CU, aMCI, and AD (p < 0.001). The Cogscreen composite score yielded area under the curve (AUC) values of 0.876 for aMCI (cut-off = 0.64, sensitivity = 0.865, specificity = 0.657) and 0.994 for AD (cut-off = -0.59, sensitivity = 0.971, specificity = 0.971), and outperformed the Mini-Mental State Examination (MMSE) in diagnosing aMCI (AUC = 0.638, p = 0.001). The composite score significantly correlated with the Consortium to Establish a Registry for Alzheimer's Disease assessment packet total score (r = 0.765, p < 0.001) and MMSE score (r = 0.722, p < 0.001).ConclusionsCogscreen is a rapid, self-administered cognitive screening tool for detecting aMCI and AD. It outperforms the MMSE in identifying early cognitive decline and holds potential for detecting even subtler cognitive changes in the future.