Observational cohort study.
Geriatric unit in an academic medical department.
One hundred nineteen acutely ill persons (34.4% female), with mean age 80.4 卤 6.9 years and body mass index 26.3 卤 4.9 kg/m2.
Assessment of muscle mass by bioimpedence analysis, muscle strength by handheld dynamometer, and gait speed with the 4-meter walking test.
Using the EWGSOP classification for sarcopenia, 5.0% presented with sarcopenia and 21.0% with severe sarcopenia. Combining gait speed test and handgrip strength measurement, the highest predictive power in detecting subjects with low muscle mass was observed (sensitivity and specificity, 80.6% and 62.5%, respectively). Subjects presenting with both normal gait speed and handgrip showed normal values of muscle mass as assessed with bioimpedence analysis. By using the ROC method, when the 2 tests were combined, the AUC was statistically higher than when using each test separately (0.740; P聽= .018).
Our study shows that 1 of 4 patients admitted to the acute care department were recognized to be sarcopenic. When a modifived version of the EWGSOP flow chart, obtained combining both gait speed and handgrip was used, sensitivity and specificity of algorithm to identify subjects with low muscle mass was improved.