- Accurate determinations of body composition, fat mass (FM) and fat-free
mass (FFM) are of interest to scientists as well as many individuals
who serve as health and fitness practitioners. Currently, researchers and
practitioners use a variety of indirect methods to determine body
composition. Traditional methods of estimating percent body fat include
hydrostatic weighing, skinfold anthropometry and bioelectrical
impedance. Dual energy x-ray absorptiometry (DEXA), utilized most
commonly to determine bone mineral content, has recently been proposed
to be an accurate measure in the assessment of body composition. Hologic
Inc., manufacturers of the QDR-1000/W bone densitometer, have recently
developed tissue composition software which can be used to estimate
percent body fat in humans. Although the Hologic QDR-1000/W has been
extensively evaluated for its accuracy in measuring tissue composition in
"vitro", "in vivo" validation studies of body composition have been few.
Therefore, the purpose of this study was to evaluate the accuracy of the
Hologic QDR-1000/W in determining human body composition by
comparing values for percent fat from DEXA to values derived using a multicomponent criterion measure of body composition in a group of 51 women and 50 men aged 19-82 years. Additionally, it was of interest to make comparisons of percent body fat determinations between the multicomponent criterion model and hydrostatic weighing, skinfold measures and bioelectrical impedance.
All subjects completed the various body composition procedures used to estimate percent body fat: DEXA, hydrostatic weighing (2-component model), skinfold anthropometry and bioelectrical impedance. The multicomponent (4-component) criterion model procedures induded hydrostatic weighing (body density), DEXA whole body scanning (bone mineral content), and deuterium oxide (D₂0) dilution in respiratory water (total body water).
Results of this study failed to reveal statistically significant mean percent body fat differences between hydrostatic weighing (2-component model) and the 4-component criterion model (25.2 ± 9.4 vs. 26.7 ± 8.4%,
p > 0.05) for all 101 subjects. However, differences in percent body fat were found between the 4-component model (26.7 ± 8.4%) and dual energy x-ray absorptiometry (23.2 ± 7.9%), skinfold anthropometry (24.0 ± 8.5%) and bioelectrical impedance (23.4 ± 7.5%).
Among women, hydrostatic weighing (2-component model), DEXA, skinfold anthropometry, and bioelectrical impedance all provided accurate estimates of percent body fat. However, each of the various indirect methods used to predict body composition underestimated percent body fat in men.