Measurement of total body water in children using bioelectrical impedance: a comparison of several prediction equations

Background Body composition evaluation by bioelectrical impedance analysis in children makes use of different group-specific population-derived equations. The present study was conducted to attempt to validate the use of population-independent physical model–derived equations in children. Methods The validity of bioelectrical impedance analysis for the measurement of total body water in children was evaluated by comparing results of two physical model–derived and two population-derived equations with those of deuterium dilution as reference method in a group of 38 heterogeneous children. Results Means ± standard deviation (in liters) for total body water measured with deuterium dilution and the physical model 1–derived equation were 18.4 ± 4.7 L and 18.1 ± 4.4 L, respectively. This difference is not significant, whereas significant differences were found for all other tested equations. Significant smaller absolute differences between the model 1 equation and deuterium reference results were found when compared with the results of the other three tested equations. Conclusion When compared with results of the reference deuterium method the physical model 1–derived equation was the only one that provided reliable total body water results by bioelectrical impedance analysis in children.