Body composition analysis by dual X-ray absorptiometry:: in vivo and in vitro comparison of three different fan-beam instruments

Background. Dual X-ray absorptiometry (DXA) is the preferred method for measuring body composition in clinical practice, but interchange between devices may pose problems with the interpretation of results. Objective. To establish conversion equations for body composition variables between three fan-beam DXA systems. Methods. Body composition was assessed in 21 subjects using Lunar Expert ( Expert), Lunar Prodigy ( Prodigy) and Hologic Delphi W ( Delphi). Weekly measurements of Hologic whole body phantom 164 were performed. Results. There were no significant differences between DXA-measured means of body weight, fat mass and lean body mass. Bland-Altman analysis revealed that Lunar Expert increasingly overestimated fat mass with increasing total mass (p < 0.001) relative to Delphi and Prodigy, while Delphi produced a constant underestimation of fat mass. Correlations between scale weights and DXA-measured body weights, and between DXA-measured body weights and the sum of fat masses, lean body masses and bone mineral contents (BMC)between the three instruments, were excellent (Rsqr 0.998-0.910; p < 0.001). Conversion factors to Prodigy for Expert and Delphi were respectively 1.003 and 1.011 for total body mass, 0.954 and 1.079 for fat mass, 1.018 and 0.967 for lean body mass and 1.049 and 1.136 for BMC ( Rsqr 0.999-0.991; p < 0.001). Standard error of estimate ( SEE) for the slopes ranged from 0.20% to 2.10%. Phantom studies revealed stable instrument function with CV% commonly < 2%, except for lean mass for Delphi (5.5%). Conclusions. Despite the significant differences in measurement of body composition between DXA fan-beam instruments, clinically relevant conversion factors can be established.