Acceso usuarios
José Ramón Alvero Cruz 
, Michael Marfell-Jones, Fernando Alacid Cárceles, Pedro Artero Orta, Lorena Correas-Gómez, Fernando Santonja Medina , Elvis Álvarez Carnero
The purpose of the present study was to investigate percentage body fat (%BF) differences in three Spanish dance disciplines and to compare skinfold and bioelectrical impedance predictions of body fat percentage in the same sample.
Seventy-six female dancers, divided into three groups, Classical (n=23), Spanish (n=29) and Flamenco (n=24), were measured using skinfold measurements at four sites: triceps, subscapular, biceps and iliac crest, and whole body multi-frequency bioelectrical impedance (BIA). The skinfold measures were used to predict body fat percentage via Durnin and Womersley�s and Segal, Sun and Yannakoulia equations by BIA. Differences in percent fat mass between groups (Classical, Spanish and Flamenco) were tested by using repeated measures analysis (ANOVA). Also, Pearson�s product-moment correlations were performed on the body fat percentage values obtained using both methods. In addition, Bland-Altman plots were used to assess agreement, between anthropometric and BIA methods.
Repeated measures analysis of variance did not found differences in %BF between modalities (p<0.05). Fat percentage correlations ranged from r= 0.57 to r=0.97 (all, p<0.001). Bland-Altman analysis revealed differences between BIA Yannakoulia as a reference method with BIA Segal (-0.35 ± 2.32%, 95%CI: -0.89to 0.18, p=0.38), with BIA Sun (-0.73 ± 2.3%, 95%CI: -1.27 to -0.20, p=0.014) and Durnin-Womersley (-2.65 ± 2,48%, 95%CI: -3.22 to -2.07, p<0.0001). It was concluded that body fat percentage estimates by BIA compared with skinfold method were systematically different in young adult female ballet dancers, having a tendency to produce underestimations as %BF increased with Segal and Durnin-Womersley equations compared to Yannakoulia, concluding that these methods are not interchangeable.
