Association of gait, balance, and handgrip strength with cognitive performance in hospitalized older adults: a retrospective analysis

TL;DR

This study found limited evidence that sensorimotor functions like gait and handgrip strength are reliable indicators of cognitive performance in hospitalized older adults.

Contribution

The study provides real-world evidence from a hospital setting on the relationship between sensorimotor functions and cognitive performance in older adults.

Findings

Only handgrip strength showed a weak positive correlation with cognitive performance.

Gait speed was associated with cognitive impairment status in logistic regression models.

Education level was the strongest predictor of cognitive performance.

Abstract

Sensorimotor impairments—such as reduced gait speed, diminished balance, and lower muscle strength—are common in older adults and have been suggested as early markers of cognitive decline. However, evidence from real-world hospital settings remains limited. This study investigated the associations between multiple sensorimotor functions and cognitive performance in hospitalized older adults.

A retrospective cross-sectional study was conducted among 548 inpatients aged ≥ 60 years. Sensorimotor measures included gait speed, Timed Up and Go (TUG), handgrip strength, balance score, and activities of daily living (ADL). Cognitive performance was assessed using the Mini-Mental State Examination (MMSE). Correlations were analyzed using Pearson coefficients, followed by multivariable linear and logistic regression models adjusting for demographic, clinical, and laboratory covariates.

Only handgrip strength showed a significant positive correlation with MMSE score (r = 0.085, P = 0.046), whereas gait speed, TUG, balance score, and ADL were not significantly associated with cognitive performance. In multivariable linear regression, none of the sensorimotor measures independently predicted MMSE score after covariate adjustment. Education level was the strongest independent predictor of cognitive performance (β = 0.41, P < 0.001). In the fully adjusted logistic regression model, gait speed was significantly associated with cognitive impairment, whereas other sensorimotor indicators were not independently associated.

In this real-world hospital cohort, sensorimotor measures were not independently associated with continuous MMSE performance after adjustment. However, gait speed was associated with cognitive impairment status in the fully adjusted logistic regression model, suggesting limited utility of routine sensorimotor assessments for cognitive screening during acute hospitalization. These findings should be interpreted cautiously given the retrospective, single-center design and potential measurement variability during hospitalization.

Figures