Urea space and total body water measurements by stable isotopes in patients with acute renal failure

Background. Knowledge of urea volume of distribution (V-urea) in patients with acute renal failure (ARF) is critical in order to prescribe and monitor appropriate dialytic treatment. We have recently shown that in ARF patients, V-urea estimation by urea kinetic modeling is significantly higher than total body water (TBW) by anthropometric estimation. However, these estimates of V-urea and TBW have not been validated by isotopic methods, considered as reference measurement standards. Methods. In this study, we measured V-urea by ([13C]) urea and TBW by deuterium oxide (D2O) in 21 patients with ARF (14 males, 7 females, age 62.0 +/- 10.6 years old, 83% Caucasian, 17% African American) at three different centers. These measurements were compared to TBW estimates from anthropometric and bioelectrical impedance (BIA) measurements. Results. Our results show that V-urea by ([13C]) urea (51.0 +/- 11.7 L) is significantly higher than TBW estimated by all other methods (TBW by D2O: 38.3 +/- 9.8 L, P < 0.001; TBW by BIA: 45.7 +/- 15.7 L, P = 0.08; TBW by Watson formula: 38.3 +/- 7.3 L, P < 0.001; TBW by Chertow formula: 39.3 +/- 7.8 L, P = 0.002, all versus V-urea). Despite significant overestimation of the absolute value and considerable variation, V-urea significantly correlated with TBW by BIA (r = 0.66, P < 0.01) and TBW by D2O (r = 0.5, P = 0.04). There was also significant correlation between D2O and BIA determined TBW (r = 0.8, P < 0.001). Conclusion. In terms of useful guidelines to prescribe a specific dose of dialysis in patients with ARF, conventional estimates of TBW as surrogates for V-urea should be used with caution. We propose that these conventional estimates of TBW should be increased by approximately 20% (a factor of 1.2) to avoid significant underdialysis.