Background: There exists a need for more sensitive measures capable of detecting subtle
cognitive decline due to Alzheimer's disease.
Objective: To advance the literature in Alzheimer’s disease by demonstrating that performance on a
cued-Stroop task is impacted by preclinical Alzheimer's disease neuropathology.
Method: Twenty-nine cognitively asymptomatic older adults completed a computerized, cued-Stroop
task in which accuracy rates and intraindividual variability in reaction times were the outcomes of interest.
Cerebrospinal fluid biomarkers of Aβ42 and tau were measured and participants were then grouped
according to a published p-tau/Aβ42 cutoff reflecting risk for Alzheimer’s disease (preclinical Alzheimer's
disease = 14; control = 15).
Results: ANOVAs indicated that accuracy rates did not differ between the groups but 4-second delay
incongruent color-naming Stroop coefficient of variation reaction times were higher in the preclinical
Alzheimer’s disease group compared to the control group, reflecting increased within-person variability.
Moreover, partial correlations showed no relationships between cerebrospinal fluid biomarkers and accuracy
rates. However, increases in coefficient of variation reaction times correlated with decreased
Aβ42 and increases in p-tau and the p-tau/Aβ42 ratio.
Conclusion: Results supported the ability of the computerized, cued-Stroop task to detect subtle Alzheimer’s
disease neuropathology using a small cohort of cognitively asymptomatic older adults. The
ongoing measurement of cued-Stroop coefficient of variation reaction times has both scientific and
clinical utility in preclinical Alzheimer’s disease.