Anthropometric estimation of bone mineral content in young adult females

It was hypothesized that improved prediction of bone mineral content (BMC) would be achieved by an expanded battery of anthropometric variables beyond height and the four extremity diameters used by previous investigators to predict skeletal mass (r = 0.74). Accordingly, sitting height, trunk diameters, chest depth, and head circumference, as well as whole body BMC and bone mineral density (BMD) via dual‐energy X‐ray absorptiometry (DEXA), were also measured on 52 young adult females (26 each Blacks and Whites), aged 18 to 30 years. The product of height and the squared sum of four extremity breadths, utilized by previous investigators, resulted in r values of 0.74 and 0.70 for the Black and White groups, respectively. Separate stepwise multiple regressions to predict BMC for each group resulted in a collectively different array of independent variables (though there was some overlap between groups). The multiple regression formula for the Black group (R = 0.88; SEE = 156 g) included height, chest depth, chest diameter at the sixth rib level, and knee bicondylar diameter. That for the White group (R = 0.83; SEE = 193 g) included two diameters (bicondylar and biepicondylar) and head circumference. When anthropometrically estimated values for upper‐arm muscle area or upper‐leg lean tissue area (which were better correlated to BMD than BMC) were added to these equations, prediction of BMC was further improved (R = 0.91; SEE = 144 g; and R = 0.85; SEE = 187 g, respectively, for the Black and White groups). The major observation in this study was a substantial improvement in the predictive precision of BMC, as assessed by DEXA, via easily obtainable anthropometric data. Further, the SEE values in this totally anthropometric prediction of whole body BMC closely approximate that recently reported for adult females, aged 18–49 years, when including the mid‐shaft radial BMC (assessed by single‐photon absorptiometry) with height, weight, and body mass index. It is concluded that anthropometric estimation of BMC may be a potentially useful tool to improve the accuracy of percentage body fat estimation assessed by hydrostatic weighing, and that it should be examined in other populations differing significantly in fat‐free mass density from the 1.10 g/ml value of the reference man. © 1992 Wiley‐Liss, Inc.