Comparison of Overall and Segmental Body Composition in Collegiate Track Athletes Using BIA and DXA

Abstract

Anthropometric and body composition measurements can offer insight into athletes’ health as well as assist in developing training or diet regimens to enhance competitive potential. Valid and reliable assessments of these measures are vital. PURPOSE: The purpose of this analysis was to compare overall and segmental body composition data of collegiate track athletes using bioelectrical impedance analysis (BIA) and dual energy X-ray absorptiometry (DXA). METHODS: Participants visited the Human Performance Laboratory once for measurement of anthropometric data (ht via stadiometer) and body composition assessment by BIA (via InBody 570) and DXA (via GE Healthcare Lunar Prodigy Advance). Contraindications were addressed and jewelry was removed prior to testing. Athletes were instructed to remove their socks; thereafter, they followed verbal instructions provided by the InBody 570. Body weight (lbs.) determined by the InBody 570 was converted to kilograms (kg) and used in the DXA analysis, for consistency. Relative body fat (%BF) and regional lean mass in the arms (ArmsLean), legs (LegsLean), and trunk (TrunkLean) (kg) were compared between methods, by sex, using paired- samples t-tests. RESULTS (females): Twenty-four females (age: 19.6±1.0 yr, ht: 168.8±7.9 cm, and wt: 65.0±13.7 kg) volunteered for testing. There were statistical differences (ordered by BIA and DXA) for %BF [19.8±6.4 vs. 25.0±7.9%, p < 0.0001] and TrunkLean [22.10±3.72 vs. 21.11±2.42 kg, p = 0.038]. No statistical differences were found for ArmsLean [5.18±1.23 vs. 5.03±0.92 kg, p = 0.163] or LegsLean [15.73±2.33 vs. 16.32±2.35 kg, p =0.070]. RESULTS (males): Thirty males (age: 20.5±1.9 yr, ht: 179.9±5.7 cm, and wt: 80.2±16.8 kg) volunteered for testing. There was a statistical difference (ordered by BIA and DXA) for ArmsLean [8.02±1.54 vs. 8.65±2.17 kg, p = 0.026], while %BF was at the statistical cut-point [11.2±4.7 vs. 13.0±6.7%, p = 0.05]. No statistical differences were found for LegsLean [21.23±2.53 vs. 22.31±4.95 kg, p = 0.115] or TrunkLean [30.36±4.42 vs. 29.87±3.65 kg, p = 0.231]. CONCLUSIONS: Clearly, differences exist in the estimation of overall and segmental body composition depending on the method utilized. If pre-post or serial evaluations are to occur, switching assessment methods would not be appropriate.