Relationship between gait parameters and cognitive indexes in elderly adults

Abstract

PurposesTo select the summary cognitive index that is best predicted from spatio-temporal gait features (STGF) extracted from gait patterns.Methods125 participants were recruited, including 40 young and 85 elderly adults. Their performances in different cognitive domains were evaluated through 12 cognitive measures obtained from five neuropsychological tests. Four summary cognitive indexes were calculated in each case: 1) the z-score of Mini-Mental State Examination (MMSE) from a population norm (MMSE z-score); 2) the sum of the absolute z-scores of the patient’s neuropsychological measures from a population norm (ZSum); 3) the patient’s scores for the first principal component (PC) of the set of neuropsychological test scores (PCCog); and 4) the Mahalanobis distance of each patient’s score from a population norm (MDCog). The gait patterns were recorded while they executed four walking tasks (normal, fast, easy- and hard-dual tasks). Sixteen STGF were measured using a body-fixed Inertial Measurement Unit. Dual-task costs were computed. The proportion of variances (R2) that PCA-STGF scores accounted for the four summary cognitive indexes and for the 12 cognitive variables across individuals were measured in multiple linear regressions. The confidence intervals for each R2 were estimated by bootstrapping the regression 1000 times.ResultsThe mean values of R2 for the summary cognitive indexes were: 0.0831 for MMSE z-score, 0.0624 for ZSum, 0.0614 for PCCog, and 0.4751 for MDCog. The mean values of R2 for the 12 cognitive values ranged between 0.0566 and 0.1211. The multivariate linear regression was only statistically significant for MDCog, with the highest value of the R2 estimation.ConclusionsIndividual cognitive variables and most of the summary cognitive indices showed a weak association with gait parameters. However, the MDCog index showed a stronger association, explaining about 50 % of the variance. This suggests that this index can be used to study the relationship between gait patterns and cognition.