Percent Body Fat Determination: Influence of Bone Mineral in Female Athletes

Assessment of body composition is widely used in athletics to determine individually focused training goals. The accuracy of the assessment relies on certain assumptions for the fat‐free body. For female athletes, there is a dearth of research to validate the methods and test the assumptions for validity. Violation of one assumption, a fixed density of the fat‐free body of 1.1 g/mL, may occur with bone mineral content that varies depending on the type of athlete (Prior et al 2001). The study examined the effect of bone mineral on the prediction of percent fat. We hypothesized that by accounting for bone mineral content by using a three‐compartment model, the percent fat will change from that determined with a two‐compartment model. More specifically, correcting for bone mineral content would increase the percent body fat in gymnasts, those in a high‐impact sport, and lower percent fat values in swimmers, those in a low‐impact sport. Seven collegiate female swimmers (age, 19.0 ± 1.4 y; height, 171.2 ± 9.8 cm; weight, 73.6 ± 12.4 kg) and 11 collegiate female gymnasts (age, 19.6 ± 1.1 y; height, 161.3 ± 4.3 cm; weight, 58.9 ± 7.6 kg) participated; the two groups averaged ≥ 12 y of training in their sport (p>0.05). Each subject underwent a DEXA scan (GE Healthcare Lunar iDXA, Madison, WI) for bone mineral density and had body density measured using air displacement plethysmography (BodPod, Cosmed, Chicago, IL). Subjects also completed a survey on sport‐specific history. The two‐compartment approach used the Siri equation (Siri 1961) and the three‐compartment model used the Lohman equation (Lohman 1986). Independent t tests were used to compare the athletes for body composition and physical characteristics. A paired t test was used to compare percent fat for the two models. The results revealed sport‐specific differences in body density (1.0322 ± 0.0151 vs 1.0586 ± 0.0131 g/mL), bone mineral density (1.203 ± 0.707 vs. 1.283 ± 0.083 g/cm2), and percent body fat (p<0.05 for swimmers vs. gymnasts). For both groups, a significant difference in percent body fat was found for methods of determination: swimmers, 29.6 ± 7.1% vs. 21.9 ± 8.0% and gymnasts, 17.7 ± 5.8% vs. 10.4 ± 7.1% for 2‐compartment vs. 3‐compartment model, respectively. In contrast to our hypothesis, both groups decreased in percent body fat when bone mineral was taken into account. These data demonstrate the importance of compartmentalizing the fat‐free body when assessing percent fat and providing an accurate criterion variable to which field methods may be calibrated.