ABSTRACT
BACKGROUND:In chronic kidney disease (CKD), multiple metabolic and nutritional abnormalities contribute to the impairment of skeletal muscle mass and function thus predisposing patients to the condition of sarcopenia. Herein, we investigated the prevalence and mortality predictive power of sarcopenia, defined by three different methods, in non-dialysis-dependent (NDD) CKD patients.METHODS:We evaluated 287 NDD-CKD patients in stages 3-5 [59.9 ± 10.5 years; 62% men; 49% diabetics; glomerular filtration rate (GFR) 25.0 ± 15.8 mL/min/1.73 m(2)]. Sarcopenia was defined as reduced muscle function assessed by handgrip strength (HGS <30th percentile of a population-based reference adjusted for sex and age) plus diminished muscle mass assessed by three different methods: (i) midarm muscle circumference (MAMC) <90% of reference value (A), (ii) muscle wasting by subjective global assessment (B) and (iii) reduced skeletal muscle mass index (<10.76 kg/m² men; <6.76 kg/m² women) estimated by bioelectrical impedance analysis (BIA) (C). Patients were followed for up to 40 months for all-cause mortality, and there was no loss of follow-up.RESULTS:The prevalence of sarcopenia was 9.8% (A), 9.4% (B) and 5.9% (C). The kappa agreement between the methods were 0.69 (A versus B), 0.49 (A versus C) and 0.46 (B versus C). During follow-up, 51 patients (18%) died, and the frequency of sarcopenia was significantly higher among non-survivors. In crude Cox analysis, sarcopenia diagnosed by the three methods was associated with a higher hazard for mortality; however, only sarcopenia diagnosed by method C remained as a predictor of mortality after multivariate adjustment...
