Evaluating the Associations of Adiposity, Functional Status, and Anthropometric Measures with Nutritional Status in Chronic Hemodialysis Patients: A Cross-Sectional Study.

BACKGROUND: Malnutrition is common in chronic hemodialysis (HD) patients and often remains underdiagnosed. While body composition, functional status, and anthropometric measures can support nutritional assessment, their associations with nutritional status are not fully established in this population. This study aimed to evaluate the diagnostic performance of various measures for assessing malnutrition in chronic HD patients, using the Subjective Global Assessment (SGA) as the reference standard. METHODS: This cross-sectional study involved chronic HD patients, stratified by nutritional status using the SGA. Data collection consisted of clinical interviews, anthropometric and functional measurements, bioelectrical impedance analysis (BIA), and biochemical analyses. Statistical analysis included Spearman's correlation, logistic regression, receiver operating characteristic (ROC) curve analysis with area under the curve (AUC) to assess predictive accuracy, standardized effect sizes to show the magnitude of differences, and kappa statistics to evaluate concordance between variables. RESULTS: This study included 103 chronic HD patients. Malnutrition was diagnosed in 50.5% of patients based on the SGA. Phase angle (PA) was the strongest single predictor of malnutrition (AUC = 0.79; specificity 0.88; sensitivity 0.58). PA ≤ 5.1° was significantly associated with higher malnutrition risk (OR: 10.23; 95% CI: 3.93-30.61; p < 0.001). Handgrip strength (HGS) also demonstrated good diagnostic value (AUC = 0.71; specificity 0.84; sensitivity 0.59). A multivariable model incorporating eight parameters-gender, post-dialysis ECW/ICW ratio, post-dialysis lean and fat mass, serum albumin, normalized protein catabolic rate (nPCR), arm circumference (AC), and HGS-achieved an AUC of 0.88 (95% CI: 0.81-0.95) and pseudo-R2 of 0.46, demonstrating improved predictive performance. CONCLUSIONS: An integrated panel of anthropometric, bioimpedance, functional, and biochemical markers provides superior diagnostic accuracy compared to individual predictors, supporting a holistic diagnostic approach in HD patients.