Dr. Peter S. W. Davies replies
Abstract
174 Letter to the Editor Dr. Peter S. W. Davies replies I am grateful to Martorell and his colleagues for stressing the importance of considering simple anthropometric predictors of total body water in bioelectrical impedance validation studies. Indeed, in some such studies (Holier et al. 1969) the impedance technique has been deemed 'useful' if the correlation between height2/resistance (H2/R) and total body water was greater than that found between body weight and total body water. The analyses suggested by Martorell et aL was carried out at the time of submission of our work. Specifically, the use of age as a single predictor variable yielded a standard error of 3.2 litres of water, whilst the use of Quetelet's index (W/H 2) gave a standard error of 4" 2 litres. Furthermore, the use of multiple predictor variables (height, weight, age, Quetelet's index) failed to improve upon the standard error of 1.69 litres obtained when using H2/R. Therefore I feel that our conclusion that bioelectrical impedance is a potentially valuable technique for assessing body composition in the paediatric population is valid. References HOFFER,E. C., MEADOR,C. K. and SIMPSON,D. C., 1969, Correlation of whole-body impedance with total body water volume. Journal of