Validity of predicted total body water and extracellular water using multifrequency bioelectrical impedance in an Ethiopian population.

Total body water (TBW) and extracellular water (ECW) were measured by deuterium oxide dilution and bromide dilution, respectively, in a group of 24 male and 20 female healthy Ethiopians, living in the capital Addis Ababa. Body weight, body height, skinfolds and total body impedance at 1 and at 100 kHz were also measured. TBW and ECW were predicted from impedance values at 1 and 100 kHz, respectively, using prediction equations developed in a Dutch adult population. ECW was overestimated by 1.3 +/- 1.0 kg (p < 0.05) and 0.6 +/- 0.8 kg (p < 0.05) in males and females, respectively. TBW was accurately predicted in males (0.1 +/- 1.9 kg, n.s.), but overestimated in females (1.0 +/- 1.3 p < 0.05). TBW/height and ECW/height were substantially lower in the Ethiopians compared to values recently published in Dutch and Italian adult subjects, indicating a different, more slender body build of the Ethiopians. After correcting for these differences in body build and for the slight differences in body water distribution (ECW/TBW), the differences between measured and predicted TBW and ECW decreased and were no longer significant. The results indicate that the validity of predicted body water from impedance depends on the body build of the subjects, which should be taken into account to avoid systematic errors when applying prediction formulas from a reference population to another population under study.