Multicompartment model to assess body composition in professional water polo players

The aim of this study was to compare differences between skinfold thicknesses (SK), bioelectrical impedance analysis (BIA), a 2-compartment model (2C), and 2, 3-compartment models (3C) of percent body fat (%BF) assessment with a 4-compartment model (4C) in professional water polo athletes.

Ten male subjects, 18 to 29 years of age, participated in this study. Under water weighing, deuterium dilution, and dual-energy X-ray absorptiometry (DXA) were used to provide the required values for the models. The two, 3C models used were established by Withers et al. (3C-W) and by the DXA manufacturers (3C-DXA). 4C was also established by Withers et al.

There were no significant differences between 2C, 3C-W, and 3C-DXA when compared with 4C (criterion model). There were no significant differences between 2C and 3C-W. %BF was significantly greater using 2C compared with 3C-DXA. %BF derived from SK and BIA were significantly greater than all other

4C was best correlated with 3C-W (R2=0.998), followed by 2C (R2=0.806), 3C-DXA (R2=0.5071), SK (R2=0.2945), and BIA (R2=0.2377).

We conclude that 2C and 3C-W assess %BF equally as well as 4C; however, SK and BIA significantly over-estimated %BF in water polo athletes.