Cognitive Trajectories and Alzheimer Disease Biomarkers: From Successful Cognitive Aging to Clinical Impairment

Objective Cross‐sectional definitions of successful cognitive aging have been widely utilized, but longitudinal measurements can identify people who do not decline. We performed this study to contrast maintenance with declining trajectories, including clinical conversion. Methods We included baseline cognitively unimpaired Alzheimer's Disease Neuroimaging Initiative participants with 3 or more cognitive testing sessions (n = 539, follow‐up 6.1 ± 3.5 years) and calculated slopes of an episodic memory composite (MEM) to classify them into two groups: maintainers (slope ≥ 0) and decliners (slope < 0). Within decliners, we examined a subgroup of individuals who became clinically impaired during follow‐up. These groups were compared on baseline characteristics and cognitive performance, as well as both cross‐sectional and longitudinal Alzheimer disease (AD) biomarker measures (beta‐amyloid [Aβ], tau, and hippocampal volume). Results Forty‐one percent (n = 221) of the cohort were MEM maintainers, and 33% (n = 105) of decliners converted to clinical impairment during follow‐up. Compared to those with superior baseline scores, maintainers had lower education and were more likely to be male. Maintainers and decliners did not differ on baseline MEM scores, but maintainers did have higher non‐MEM cognitive scores. Maintainers had lower baseline global Aβ, lower tau pathology, and larger hippocampal volumes than decliners, even after removing converters. There were no differences in rates of change of any AD biomarkers between any cognitive trajectory groups except for a higher rate of hippocampal atrophy in clinical converters compared to maintainers. Interpretation Using longitudinal data to define cognitive trajectory groups reduces education and sex bias and reveals the prognostic importance of early onset of accumulation of AD pathology. ANN NEUROL 2024;96:378–389