Aims: To compare estimations of body composition using two different methods of bioelectrical impedance analysis (BIA) with dual X-ray absorptiometry (DXA) for estimating body composition during weight loss in overweight and obese young females.
Methods: Twenty-four overweight or obese females (age 29.5+/-6.1 years, BMI 36.4+/-4.3 kg/m(2)) had body composition assessed using single-frequency (Tanita Ultimate Scale; SF-BIA) and multi-frequency (Impedimed SFB7; MF-BIA) BIA and DXA before and after a 10-week weight loss intervention.
Results: MF-BIA estimates of body composition showed good absolute agreement with DXA, as evidenced by the small biases in the estimation of fat free mass (FFM), fat mass (FM) and percentage body fat (BF%); however, the limits of agreement for each variable were wide (bias +/-1.96 standard deviation; FFM -1.6+/-6.5 kg, FM 1.6+/-6.5 kg, BF% 1.4+/-6.3%). SF-BIA exhibited a larger bias with wide limits of agreement (FFM 3.8+/-9.1 kg, FM -3.8+/-9.1 kg, BF% -4.37+/-10.3%). During weight loss the values provided by MF-BIA and SF-BIA were not significantly different from DXA (p> or =0.89) due to small bias and the limits of agreement were narrow (MF-BIA: FFM -0.01+/-3.74 kg, FM 0.01+/-3.74 kg, BF% 0.22+/-3.40%; SF-BIA: FFM 0.40+/-3.92 kg, FM -0.40+/-3.92 kg, BF% 0.25+/-3.40%).
Conclusion: Compared with DXA, both the MF-BIA and SF-BIA accurately assessed changes in body composition with weight loss but, compared with SF-BIA, MF-BIA provided superior cross-sectional estimates of body composition.