Evaluating handgrip strength and functional tests as indicators of gait speed in older females

With aging, females often experience greater declines in functional capacity [e.g., gait speed (GS)] compared to males, highlighting the need for sex-difference considered in screening and intervention planning. In certain contexts, assessing GS may not be feasible. Handgrip strength (HGS) commonly used as a surrogate measure for physical performance, also serves as an indirect indicator of muscle strength in the lower limbs. This cross-sectional study aims to investigate the associations between HGS and common functional tests and to determine the optimal cut-off values for these tests in assessing GS.

142 community-dwelling older females aged 60-80 years old (mean age: 75 ± 6 years) were evaluated with HGS, the 30-second arm curl (30 s-AC), 30-second chair stand (30 s-CS), the Short Physical Performance Battery (SPPB), and the 8-foot Up & Go (8-UG) test. Pearson's correlation (r) was used to assess the strength of associations between HGS and functional variables, while multiple linear regression models identified determinants of GS. Receiver operating characteristic (ROC) curves were employed to evaluate the effectiveness of various tests in detecting slow GS (<1.0 m/s), by means of the area under the curve (AUC), sensitivity, and specificity.

HGS showed positive significant (p < 0.001) associations with 30 s-AC (r = 0.499), SPPB (r = 0.447), and 30 s-CS (r = 0.329). Standardised coefficients of the linear models were: 30 s-AC (β=0.593), 30 s-CS (β=0.513), 5-CS (β=-0.431), and HGS (β=0.475) (all p < 0.001). ROC analysis revealed the following results: 30 s-AC (AUC = 0.80, cut-off=∼16 repetitions, sensitivity 83%, specificity 36%), 30 s-CS (AUC = 0.74; cut-off=∼13 repetitions, sensitivity 78%, specificity 64%), and 5-CS (AUC = 0.75, cut-off = 10.0 s, sensitivity 81%, specificity 57%), HGS (AUC = 0.73, cut-off=∼20 kg, sensitivity 79%, specificity 46%).

We found that HGS was moderately-to-weakly associated with functional outcomes in older females, indicating that it may not reflect the overall body functional capacity. Despite similar AUCs across all tests, the 30 s-CS and 5-CS showed a better balance of sensitivity and specificity, making them potential indicators of slow GS compared to HGS and 30 s-AC.