Neurodegenerative dementia models, exemplified by Alzheimer’s disease posit the emergence of presymptomatic cognitive decline well after the earliest detectable biomarker changes and nearly proximate to the clinical presentation of MCI, but data supporting such models include cross sectional studies that fail to account for within-subject cognitive changes, or relatively insensitive psychometric tests as well as elderly individuals harboring advanced preclinical stage disease.

This is a longitudinal neuropsychological study of an apolipoprotein E (APOE) ε4 enriched cohort of cognitively normal individuals at entry that began in 1994 with enrollment age 21 years and above. Linear mixed models were fitted for each neuropsychological measure for MCI converters and nonconverters. The age at which the two group trajectories intersected was an inflection point from which pre- and post-inflection annual changes were compared between groups with piecewise linear mixed effects regression models.

Over a mean of 9.5 years 65 individuals developed amnestic MCI, mean age 73 (S.D. 6.8) years. 719 remained clinically normal. 34 of 35 cognitive measures and 9 of 18 behavioral measures showed a significantly steeper slope of decline post-inflection in the MCI converters relative to the nonconverters, and the earliest cognitive inflection point was 20 years in advance of MCI diagnosis.

In contrast to previously reported and widely accepted findings of an extreme time lag between the earliest reported amyloid biomarker changes and cognitive decline, we found the earliest cognitive changes occur 20 years before incident mild cognitive impairment, rivaling the earliest reported biomarker changes and challenging current disease models. Alternate pathophysiological models (e.g., loss of a critical APP-related function rather than a gain of abeta toxicity) that better fit this chronology should be further explored.