Comparing single-frequency bioelectrical impedance analysis against deuterium dilution to assess total body water

In this study, we aimed to validate the accuracy of single-frequency bioelectrical impedance analysis (SF-BIA) at 50 kHz to assess total body water (TBW) against the reference technique deuterium dilution (D(2)O) and to explore if the simple clinical parameters extracellular fluid (ECF) composition and body shape explain individual differences between D(2)O and SF-BIA (Diff(BIA-D(2)O)).

We assessed TBW with D(2)O and SF-BIA in 26 women and 26 men without known disease or anomalous body shapes. In addition, we measured body shape with anthropometry and ECF composition (osmolality, albumin, glucose, urea, creatinine, sodium and potassium).

On group average, SF-BIA to predict TBW agreed well with D(2)O (SF-BIA, 39.8 ± 10.1 l; D(2)O, 40.4 ± 10.2 l; and Diff(BIA-D(2)O) -0.7 l). In four individuals ('outliers'; 15% of the study population), Diff(BIA-D(2)O) was high (-6.8 to +3.8 l). Diff(BIA-D(2)O) was associated with individual variations in body shape rather than ECF composition. Using gender-specific analysis, we found that individual variability of waist circumference in men and arm length in women significantly contributed to Diff(BIA-D(2)O). When removing the four 'outliers', these associations were lost.

In the majority of our sample, BIA agreed well with D(2)O. Adjusting for individual variability in body shape by anthropometrical assessment could possibly improve the accuracy of SF-BIA for individuals who deviate from mean values with respect to body shape. However, further studies with higher subject numbers are needed to confirm our findings.