Introduction: Body fat percentage (BF%) has been shown to be an important item for talent identification (Reilly et al., 2000) and injury prevention in elite-standard youth soccer (Kemper et al., 2015). Our aim was to find out what are the most appropriate alternative methods to identify changes in BF% ( BF%) when dual-energy X-ray absorptiometry (DXA) is used as the reference method in elite youth soccer players. Methods: The sample was composed of 20 highly-trained youth soccer players (17.1±0.5 y). BF% was measured in the preseason and in the middle of season (5 months after) by DXA [Hologic Serie Discovery QDR, USA], 2 bioelectrical impedance analyzers [Tanita BC- 418 (TA), JP; and InBody770 (IB), KR], and skinfold thickness measurement using a Holtain caliper and estimating BF% through several formulas proposed in the literature [Slaughter (SLA), Faulkner (FAU), Carter (CAR), Durnin-Wormersley (DW), Brook (BR), Withers (WIT), and Lohman (LOH)]. T test for dependent samples, intraclass correlation coefficient (ICC) and Bland-Altman analysis were used. Results: BF% measured by DXA was significantly decreased after 5 months of soccer competition (11.4±2.6% vs. 10.7±2.2%; p<0.05). ICCs between dBF% by DXA and alternative methods were: 0.53 (TA), 0.44 (IB), 0.83 (SLA), 0.70 (FAU), 0.74 (CAR), 0.79 (DW), 0.75 (BR), 0.87 (WIT), and 0.58 (LOH) (all p<0.05). Bland-Altman analysis revealed a significant bias [95%CI] between dBF% by DXA and dBF% by IB (1.9% [5.7, -1.9]), FAU (-0.6% [1.1, -2.3]), CAR (-0.8% [1.0, -2.6]), DW (-0.8% [1.0, -2.6]), BR (-1.1% [1.1, -3.3]), WIT (-0.5% [0.9, -2.0]), and LOH (-0.7% [1.3, -2.7]) (all p<0.01), but not for TA (-0.5% [2.3, -3.2]) and SLA (-0.3% [1.8, -2.4]. The difference between DXA and IB, FAU, CAR, and LOH in estimating dBF% was affected by the amount of dBF% (r²=0.25-0.80; p<0.05). Conclusion: The equation of Slaughter et al., which includes triceps and medial calf skinfold thicknesses, seems to be the most recommended alternative method to DXA to assess dBF% in young soccer players due to its moderate-high ICC and lack of significant bias with DXA.
© Copyright 2016 21st Annual Congress of the European College of Sport Science (ECSS), Vienna, 6. -9. July 2016. Published by University of Vienna. All rights reserved.