Comparisons between fat measurements by dual-energy X-ray absorptiometry, underwater weighing and magnetic resonance imaging in healthy women.

OBJECTIVE

To compare the results of measurements of total and regional fat by dual-energy X-ray absorptiometry (DXA), underwater weighing (UWW) and magnetic resonance imaging (MRI) in women.

DESIGN

13 volunteer pre-menopausal women measured with a Norland XR26 HS DXA instrument, hydrodensitometry equipment and a 0.08 Tesla MRI scanner.

SETTING

Aberdeen Royal Infirmary.

RESULTS

There were high correlations between measurements of total-body fat by the different techniques, but poor agreement. The mean values of fat as a percentage of the total body mass were UWW = 28.6%, DXA = 40.0%, MRI = 23.0%, although the MRI excluded the head, forearms and feet, estimated from the DXA measurements to contain 8% of the body fat. Although the correlations were apparently linear in the ranges examined, those including the Norland DXA had high intercepts. Bland and Altman plots showed no dependence of the differences on the degree of fatness. Evidence was obtained from in vivo comparisons and phantom measurements that the fat calibration of the Norland DXA had changed during a three year period, suggesting that fat calibration errors contributed to the MRI/DXA differences. Previous comparisons between the Norland DXA and other brands indicate that Hologic and Lunar instruments would show a closer agreement with MRI and UWW. Profiles of fat distribution along the body showed variations in the DXA/MRI ratio, particularly in the chest, with the DXA pattern thought to be the less accurate. The DXA measurements of bone mineral, fat and lean tissue were used to demonstrate the variable proportion of bone in fat-free tissue, leading to errors in UWW.

CONCLUSIONS

Each method has its assumptions, which are often not sufficiently appreciated. Each method may be useful for testing and improving the assumptions in the other methods. No method can yet be regarded as a satisfactory reference technique.