Digital Stroop Test as a neurocognitive marker in a community-based sample of older adults: Data from the Framingham Heart Study.

BackgroundDigital cognitive testing allows for assessment of more granular aspects of cognition that may enhance the ability to detect cognitive decline and Alzheimer's disease earlier.ObjectiveTo assess cognition using a smartphone-based Stroop Task in older adults.MethodsThe smartphone-based Stroop Task consisted of four subtests including two with minimal cognitive demand, i.e., color matching (subtest 1) and color-word matching (subtest 2), and two with greater cognitive demand i.e., inhibition (subtest 3); and inhibition/switching (subtest 4). Each subtest consisted of five trials. Repeated measures ANOVA were conducted to examine mean completion times within and between the four test conditions. Completion times were also compared to traditional neuropsychological tests.ResultsAmong 478 iPhone users, 429 (89.7%) used the app-based Stroop Test (mean age 73.5 years (6.4), 58% female, 87% non-Hispanic White, mean MMSE score 28.75 ± 1.4). Error-free performance occurred in 395 participants on subtest 1, 404 on subtest 2, 320 on subtest 3, and 183 on subtest 4. Mean completion times differed between the four subtests (all p < 0.0001) with faster subtest 2 completion (1.45 to 1.19 s across the five trials) and slower subtest 4 completion (2.95 to 2.75 s across the 5 trials) than the other subtests. Completion times were positively associated with paper/pencil measures of processing speed but negatively associated with measures of episodic and working memory and language.ConclusionsWe demonstrate the feasibility and construct validity of administering a fully self-administered smartphone-based Stroop cognitive test in older adults completed outside a clinical setting.