This study was designed to assess the usefulness of the skinfold equations developed by 1-Jackson (JP) et al., (1978) and 2-Evans (Ev) et al. (2005) in tracking body composition changes [absolute and relative fat mass (%FM), and fat-free mass (FFM)] of elite Portuguese male athletes before a competition, using a 4 compartment (4C) model as the reference method. METHODS: A total of 14 top-level male judoist and karateist (age: 22.9 ± 3.1yrs; stature: 1.75 ± 0.06m) volunteered for the study. Subjects were evaluated at baseline (Weight: 73.7 ± 8.8 kg; FM-4C: 5.6 ± 2.7 kg; %FM-4C: 7.6 ± 3.5; and FFM-4C: 68.1 ± 8.2 kg) and before a competition, approximately one month apart (Weight: 73.4 ± 8.3; FM-4C: 5.1 ± 2.6 kg; %FM-4C: 7.0 ± 3.4; and FFM-4C: 68.3 ± 8.0 kg). Measures of body density assessed by air displacement plethysmography (BOD POD), bone mineral content by DXA, and total body water by bioelectrical impedance spectroscopy were used to estimate FM, %FM, and FFM with the 4C model. Seven skinfold (SKF) were measured to estimate FM, %FM, and FFM using the J and Ev 7SKF-based models, and the simplified Ev 3SKF-based equation. Comparison of means, linear regression analysis, and analysis of slope and intercept were performed. RESULTS: Changes in FM, %FM, and FFM were observed ranging from -2.6 to 3.5 kg of FM, -2.5 to 2.7 %FM, , and -2.1 to 2 kg of FFM, though the mean differences were not significant (p>0.05). The FM, %FM, and FFM changes obtained by the SKF models were not different from those obtained with the 4C model (p>0.05). The body composition changes obtained using the selected SKF equations were highly related with those obtained from the reference method in tracking FM and %FM changes, with coefficients of correlation ranging from 0.5 to 0.58, and FFM changes, ranging from 0.74 to 0.79. All the models explained 44 to 51% of the variance in the observed FM changes from the 4C model with a standard error of measurement (SEE) between 1.27 to 1.35 kg. For %FM changes, the models explained 51 to 57% with a SEE ranging from 0.91 to 0.99 kg. For FFM, the explained variance ranged from 55 to 56 % with a SEE of 0.89 to 0.91. The models did not differ from the line of identity in estimating FM, %FM, and FFM changes (p>0.05). CONCLUSION: Considering the performance criteria, the equation using 3SKF and the two models using 7SKF presented a similar accuracy in assessing small changes of %FM, FM, and particularly FFM when compared to the reference method. These data highlight the usefulness of the skinfold-based models in tracking body composition changes in highly trained athletes.
© Copyright 2007 12th Annual Congress of the European College of Sport Science, Jyväskylä, Finland - July 11-14th 2007. All rights reserved.