Evaluating the predictive validity of SARC-F cut-off scores for low muscle strength among older adults in a low-income community

Abstract

Background

With an aging population, muscle health, encompassing locomotion and metabolic function, has become a public health priority. Handgrip strength is a validated surrogate measure of general muscle strength, but measurement may not be feasible in low-resource settings. The SARC-F questionnaire (Strength, Assistance in walking, Rise from a chair, Climb stairs, and Falls) provides a simple, low-cost, and practical tool for sarcopenia screening, though its optimal cut-off remains debated.

Objective

To evaluate the predictive validity of SARC-F cut-off scores in identifying low muscle strength among community-dwelling older adults in a low-income population.

Methods

We included 733 participants (221 men, 512 women; aged 60–95 years) from the Amazonas region of Brazil. All completed the SARC-F and underwent handgrip strength testing. Low handgrip strength was based on EWGSOP2 criteria (<27 kg men, <16 kg women). Agreement, sensitivity, specificity, predictive values, and ROC curves were calculated for cut-offs ≥4 and ≥ 2, stratified by sex.

Results

Low handgrip strength was highly prevalent (47.1% in men, 94.1% in women). Agreement between SARC-F and muscle weakness was generally poor (κ <0.4), except for men at the ≥4 threshold (κ = 0.41). Sensitivity was higher in men than in women (≥4: 48% vs 37%; ≥2: 70% vs 69%). Lowering the cut-off to ≥2 improved sensitivity but reduced specificity (65.8% men, 56.7% women). ROC analysis identified ≥2 as the optimal threshold, with AUC 0.68 in men and 0.63 in women.

Conclusion

A SARC-F cut-off of ≥2 improves sensitivity for detecting probable sarcopenia and may be more suitable for screening in low-income settings. Longitudinal validation is warranted across diverse populations.