Reliability and agreement between DXA-derived body volumes and their usage in 4-compartment body composition models produced from DXA and BIA values

Two research groups recently produced equations for estimation of body volume from dual-energy x-ray absorptiometry (DXA) scans. These body volume estimates can be used for body composition evaluation in modified 4-compartment models. In the present analysis, the reliability of body volume calculations, as well as their usage in 4-compartment models, was explored while employing precise scheduling of assessments and dietary standardization. Forty-eight recreationally active males and females completed two pairs of identical assessments, which included a DXA scan and single-frequency bioelectrical impedance analysis. Each assessment within a pair was separated by 24 hours, during which participants were provided a standardized diet. Body volume and 4-compartment equations were applied to the data, and metrics of reliability and agreement were calculated for body volume and 4-compartment components. While both body volume equations demonstrated excellent reliability individually, substantial disagreement between equations was present when utilized in 4-compartment equations. The magnitude of this disagreement was 4.3 kg for lean mass and fat mass and 6.9% for body fat percentage. At present, the large discrepancies in body composition components when using existing body volume equations preclude their interchangeability and demonstrate the need for continued exploration of the utility of body volume estimates.