the dataset of our formerly published Sarcopenia in Geriatric Elderly (SAGE) study includes 144 geriatric inpatients (86 women, 58 men, mean age 80.7±5.6 years) with measurements of gait speed, handgrip strength and appendicular muscle mass by dual x-ray absorptiometry (DXA). We analysed the agreement between EWGSOP and EWGSOP2 algorithms in identifying patients as sarcopenic/non-sarcopenic. Differences in the distribution sarcopenic vs. non-sarcopenic were assessed by Chi²-test.
sarcopenia prevalence according to EWGSOP1 (41 (27.7%)) was significantly higher than with EWGSOP2 (26(18.1%), p<0.05). The sex-specific sarcopenia prevalence was 22.1% (EWGSOP1) and 17.4% (EWGSOP2), respectively, for women (difference not significant) and 37.9% vs. 19.4% for men (p<0.05%). The overall agreement in classifying subjects as sarcopenic/non-sarcopenic was 81.25% (81.4% for women, 81.0% for men). However, among the 41 sarcopenia cases identified by EWGSOP1, only 20 (48.8%) were diagnosed with sarcopenia by EWGSOP2 (9/19 w (47.4%), 11/22 m (50.0%)). Ten of 19 women (52.6%) and 11 of 22 men (50.0%) diagnosed with sarcopenia by EWGSOP1 were missed by EWGSOP2, while 6 of 15 women (40.0%) and 0 of 11 men (0.0%) were newly diagnosed.
there is a substantial mismatch in sarcopenia case finding according to EWGSOP1 and EWGSOP2. The overall prevalence and the number of men diagnosed with sarcopenia are significantly lower in EWGSOP2. While the absolute number of women identified as sarcopenic remains relatively constant, the overlap of individual cases between the two definitions is low.