Comparing different methods of assessing body composition in end-stage renal failure

Background. Accurate measurement of nutritional status in patients with end-stage renal disease is important because of its clear association with prognosis. Total body water (TBW) has additionally been recently recognized as an independent prognostic value because of its relationship with hypertension and cardiac morbidity. The current study was designed to assess the utility of surrogate markers of nutritional state and TBW in patients with end-stage renal disease. Methods. Fifty-four patients with renal disease were studied. TBW obtained using the deuterium dilution technique was compared with estimates derived from anthropometric measures of TBW, including 58% body weight, Watson equations, and bioelectrical impedance analysis (BIA). Anthropometrically derived fat-free mass (FFM) was compared with BIA-derived estimates. Total body nitrogen (TBN) measurements were correlated with TBW estimates and BIA-derived resistance. Results. TBW was significantly underestimated by the Watson equation (mean difference, -1.751 L, P = 0.01) and the 58% body weight approximation significantly overestimated it (mean difference, 1.792 L, P = 0.04). The Kushner BIA estimation of TBW did not significantly differ from that of the gold standard determined from D2O dilution (mean difference, -1.221 L, P = 0.12) and was also the method that showed the best agreement with the D2O estimate. However, the limits of agreement were large. Accurate prediction equations for FFM (FFM = -21.768 + 0.001 x ht(2) + 6630.669 x 1/R + 0.312 x wt, R-2 = 0.95) and TBN (TBN = -668.324 - 3.963 x age + 10.133 x wt + 0.035 x ht(2) + 32141.457 x 1/R, R-2 = 0.91) were derived from BIA obtained resistance. Conclusions. The estimation of TBW varies significantly depending on the method of calculation. BIA is the most accurate surrogate marker for the measurement of both TBW and other parameters of body composition.