Prediction of body composition in premenopausal females from dual-energy X-ray absorptiometry
Niels Hansen, T. G. Lohman, Scott B. Going, M. C. Hall, Richard W. Pamenter, L. A. Bare, Thomas W. Boyden, Linda Houtkooper
Abstract
Niels Hansen, T. G. Lohman, Scott B. Going, M. C. Hall, Richard W. Pamenter, L. A. Bare, Thomas W. Boyden, Linda Houtkooper
Abstract
Four methods for predicting body composition were compared in premenopausal females (n = 100), 28-39 yr old, by using underwater weighing (UWW) as the criterion method. The four methods were dual energy X-ray absorptiometry (DEXA), skinfolds, bioelectrical impedance, and body mass index. The sample had a mean percent fat (%fat) of 29.7 +/- 6.8% (SD) by DEXA and 29.9 +/- 5.8% measured by UWW. DEXA yielded a standard error of estimate (SE) of 2.4% (r = 0.91) for the prediction of %fat from UWW. When %fat was estimated from other methods, larger SEs were obtained: 3.0% for skin-folds, 3.3% for body mass index, and 2.9% for bioelectrical impedance (height2/resistance) plus weight. Individual body density values derived from UWW were corrected for bone mineral variation. DEXA predicted the corrected body density with a lower SE (0.0040 vs. 0.0053 g/ml) than the original density values. We conclude that DEXA was a precise method and correlated highly with fat-free body weight and %fat from UWW in this homogeneous female sample.
OpenAlex reports 53 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Four methods for predicting body composition were compared in premenopausal females (n = 100), 28-39 yr old, by using underwater weighing (UWW) as the criterion method. The four methods were dual energy X-ray absorptiometry (DEXA), skinfolds, bioelectrical impedance, and body mass index. The sample had a mean percent fat (%fat) of 29.7 +/- 6.8% (SD) by DEXA and 29.9 +/- 5.8% measured by UWW. DEXA yielded a standard error of estimate (SE) of 2.4% (r = 0.91) for the prediction of %fat from UWW. When %fat was estimated from other methods, larger SEs were obtained: 3.0% for skin-folds, 3.3% for body mass index, and 2.9% for bioelectrical impedance (height2/resistance) plus weight. Individual body density values derived from UWW were corrected for bone mineral variation. DEXA predicted the corrected body density with a lower SE (0.0040 vs. 0.0053 g/ml) than the original density values. We conclude that DEXA was a precise method and correlated highly with fat-free body weight and %fat from UWW in this homogeneous female sample.
Key concepts: Bioelectrical impedance analysis, Hydrostatic weighing, Dual-energy X-ray absorptiometry, Dual energy, Bone mineral, Body adiposity index, Body water, Densitometry