Demographically adjusted normative study of everyday cognition in the ACTIVE sample.

Abstract

Objective: The goals of this project were to (1) provide demographically adjusted normative data for three performance-based tests of everyday cognition: The Everyday Problems Test, Observed Tasks of Daily Living-Revised, and Timed Instrumental Activities of Daily Living and (2) examine the relationships between test performance and traditional cognitive test scores and relevant self-report measures. Method: A sample of 2,767 Black (n = 726) and White (n = 2,041) older adults (aged 65-94) in the ACTIVE baseline sample were included in this study. Normed scores adjusting for age, education, gender, and race were created using multivariable fractional polynomial regressions. Adjusted scores were unrelated to age, education, gender, and race. A Poisson regression was performed to predict participants' number of demographically adjusted low (<16^th percentile) test scores. Results: Higher intellectual self-efficacy (coef = -0.20), immediate memory (-0.21), reasoning (-0.25), recognition vocabulary (-0.04), and digit-symbol substitution (-0.01) significantly predicted fewer low test scores while higher physical health related QOL (0.21) and daily activity limitations (0.10) significantly predicted more low test scores (p < .01). Conclusions: Generally, persons with more widespread impairment on the tests evinced more cognitive, emotional, and functional problems. The three demographically adjusted scores fit excellently on a single factor, which uniquely accounted for 19%-36% of variance in the three test scores, in excess of what was already explained by the covariates, indicating the presence of reliable shared variance among the three tests that was not attributable to demographics nor any of the other covariates.