EFFECT OF BONE-DENSITY ON BODY-COMPOSITION ESTIMATES IN YOUNG-ADULT BLACK-AND-WHITE WOMEN

Bone mineral content (BMC) and density (BMD) by dual x-ray absorptiometry, total body water (TBW) by the deuterium oxide (D2O) dilution technique, and body density (B(d)) by hydrostatic weighing were measured in 26 black (B) and 26 white (W) young adult women. Both groups were similar in age, height, weight, and total skinfolds; however, black subjects had significantly higher BMC and BMD. Formulas to estimate percent body fat (%BF) from B(d) included Siri's two-component equation for the reference man, which assumes a fat free body density (FFB(d)) of 1.100 g-ml-1, and an adjusted two-component formula that assumes a lower FFB(d) of 1.095 g-ml-1. Percent body fat was also predicted from TBW and by several multicomponent models that corrected for individual subject variation in measured BMC and TBW. The two groups did not differ significantly in %BF predictions by any of the methods. However, the difference in %BF between the groups was halved with the four-component model (B = 21.9%; W = 23.6%) as compared with that calculated from the Siri two-component densitometric model (B = 21.2%; W = 24.2%). Within each racial group, %BF was not significantly different when predicted by two-component or multicomponent models. However, %BF of individuals with the highest and lowest BMD was substantially under- and overpredicted, respectively, by Siri's equation. It is concluded that a wide range of BMD exists in the young adult female population, and that while densitometric predictions of %BF may not differ significantly from multicomponent models for most young adult women, care should be taken in interpreting the %BF of individuals who vary greatly from reference populations in BMC/FFM and/or BMD. It is suggested that these individuals' %BF may be predicted more accurately by multicomponent models.