Detecting cognitive decline in community-dwelling older adults using simple cognitive and motor performance tests.

Abstract

The objective of this study was to estimate the predictive accuracy of simple cognitive and motor performance tests to detect cognitive decline (CD) in community-dwelling older adults. In total, 102 community-dwelling older adults participated in this study. Cognitive function, gait performance and coordinated finger movements were assessed using the Montreal Cognitive Assessment, the 10-m walking test and the finger-tapping test, respectively. We classified the participants into either a CD (n = 60) or a healthy control (n = 42) group. Significant group differences were found in the visuospatial/executive function subscale score on the Montreal Cognitive Assessment, stride length and total finger tap count. The results of multivariate logistic regression analysis showed that visuospatial/executive function subscale score, stride length and total tap count were the significant predictors for the presence of CD (sensitivity 83.3%, specificity 82.9%, predictive accuracy 83.2%). We also constructed a decision tree model with these three variables to increase the usefulness of our model as a screening tool by assigning a cut-off value for each assessment. The sensitivity and specificity of the model were 88.1% and 85.2%, respectively, with an overall predictive accuracy of 86.4%. The results of the present study suggest that simple cognitive and motor performance tests have moderate-to-high discriminant validity for the presence of CD in community-dwelling older adults. Addition of such tests might lead to the more accurate detection of early cognitive decline. Geriatr Gerontol Int 2020; ••: ••-••.