Validity Of BIA For The Measurement Of Body Composition In Collegiate Gymnasts

PURPOSE: Body composition (BC) is a frequently assessed component of health-related fitness. A variety of laboratory methods including Dual-Energy X-Ray Absorptiometry (DXA) and field methods including bioelectrical impedance analysis (BIA) can be used to measure BC. The purpose of this study was to determine the validity of BIA for estimating percent body fat (% fat) in female collegiate gymnasts. METHODS: Participants’ % fat was estimated using two different BIA analyzers, the InBody 230 (INB) and the Omron HBF-306 (OMR), and then compared to DXA, which served as the reference value. Participants were 67 female collegiate gymnasts [age 19.54 (1.9) yrs., height 1.61 (0.07) m, body mass 70.0 (7.1) kg, and BMI 24.0 (2.4) kg/m2]. All participants were tested in the hydrated state [1.016 (0.01), Urine Specific Gravity]. The validity of the BIA % fat estimates was based on the evaluation of each BIA analyzer versus the reference value from the DXA by calculating the mean difference, standard error of the estimate (SEE), and coefficient of determination (R2) from linear regression analysis. Repeated measures ANOVA was used to determine overall and pairwise differences between INB, OMR, and DXA. RESULTS: The mean % fat results were as follows: INB = 22.20 (4.51), OMR = 20.92 (2.83), and DXA = 25.25 (4.18). Simple linear regression analysis for INB using DXA as the reference variable indicated an R2 value of 0.11 and SEE of 0.751. Simple linear regression for OMR using DXA as the reference variable indicated an R2 of 0.27 and SEE of 0.617. A repeated measures ANOVA indicated a statistically significant difference in % fat values between INB, OMR, and DXA (F (2, 132) = 67.56, p < 0.001). Post-hoc analyses indicated significant differences between INB and DXA [-3.06 (95% confidence interval: -3.86, -2.25) % fat, p < 0.001]; between OMR and DXA [-4.33 (95% confidence interval: -5.06, -3.60) % fat, p < 0.001]; and between INB and OMR [1.27 (95% confidence interval: 0.52, 2.03) % fat, p = 0.001]. CONCLUSIONS: Neither the INB nor OMR % fat estimates provided a valid assessment of BC in this sample of female collegiate gymnasts. When compared to the reference (DXA), both BIA analyzers significantly underestimated % fat. Neither BIA analyzer can be recommended for estimating % fat in this population based on this analysis.