Abstract
BACKGROUND: EWGSOP2 criteria for sarcopenia recommends the use of either handgrip strength (GS) or 5-times repeated chair stand test (RCS) as a muscle strength measure. We aim to compare the impact of different muscle strength definitions on sarcopenia prevalence and predictive validity for 2-year outcomes, using the EWGSOP2 clinical algorithm. METHODS: We studied 200 community-dwelling older adults, comparing sarcopenia prevalence using three muscle strength definitions: 1) maximum GS (Asian Working Group cutoffs); 2) RCS-1 (standard cutoff >15s); and 3) RCS-2 (ROC-derived cutoff >12.5s). Two-year outcomes include: 1) Incident frailty (modified Fried criteria); 2) Physical performance [Short Physical Performance Battery (SPPB) score <10]; and 3) Quality of life [EuroQol-5 dimension (EQ-5D) <25th percentile]. We performed logistic regression on 2-year outcomes adjusted for age, gender, cognition and mood. RESULTS: Prevalence of confirmed sarcopenia was 14.5%, 4% and 9% for GS, RCS-1 and RCS-2 respectively. For 2-year outcomes (N=183), RCS-2 predicted incident frailty (OR: 5.7, 95% CI 1.4-22.8, p=0.013), low SPPB (OR: 4.4, 95% CI 1.4-13.1, p=0.009), and trended towards predicting low QOL (OR: 2.1, 95% CI 0.9-4.9, p=0.095). In contrast, GS and RCS-1 did not predict frailty nor low QOL, but predicted low SPPB only (GS: OR 3.8, 95% CI 1.3-10.6, p=0.01; RCS-1: OR: 8.8, 95% CI 2.2-35.0, p=0.002). CONCLUSIONS: Sarcopenia prevalence varies with muscle strength definitions, with GS being significantly higher vis-à-vis RCS definitions. Our results also support the use of population-specific over standard cutoffs for RCS to obtain intermediate estimates of sarcopenia prevalence and the best predictive validity for two-year outcomes.