RESUMO
OBJECTIVE: To determine the most potent bioelectrical impedance analysis (BIA) marker of malnutrition and to adjust its application to hemodialysis (HD) patients. DESIGN: An observational study. SUBJECTS: A total of 99 patients on maintenance HD were enrolled in the study. INTERVENTION: The nutritional state of the patients was examined before and after the HD procedure using Subjective Global Assessment Scale (SGA), serum albumin, body mass index and BIA-derived fat-free mass index, reactance, resistance, and phase angle (PA). Malnutrition defined by the SGA questionnaire was used to detect the most potent marker of malnutrition. This marker was further analyzed and corrected for the excess fluid, age, and gender producing the nutritional state-specific cutoffs. RESULTS: The SGA rates of nutritional state were as follows: 57.6% (57) well nourished, 28.3% (28) moderately malnourished, and 14.1% (14) severely malnourished. Multivariate forward logistic regression analysis of the nutrition markers revealed PA as the most potent malnutrition predictor (odds ratio = 3.69; 95% confidence interval [CI]: 1.59-8.62; P = .002). PA was adjusted for the excess fluid (5.00 ± 0.97 vs 4.87 ± 1.08 P < .001). Patients were assigned into groups with adjusted PA values below the 5th through the 50th percentile of the mean PA reference value. The moderately malnourished patients were most accurately identified by the percentile group of <25th (area under the curve = 0.70; 95% CI: 0.60-0.81; P = .001), and the severely malnourished patients were most accurately identified by the percentile group of <15th (area under the curve = 0.74; 95% CI: 0.62-0.85; P = .005). CONCLUSION: Malnutrition is present almost in a half of the HD patients. BIA-provided PA is the most potent predictor of malnutrition. PA can be adjusted for the excess fluid after HD, age, and gender and used accordingly.
