Objective: To compare the consistency in the measurement of percentage of body fat (PBF) by multi-frequency bioelectrical impedance analysis (MF-BIA) and dual energy X-ray absorptiometry (DXA) in overweight and obese adults in China, and provide evidence for the accurate MF-BIA application in China.
Methods: A total of 1 323 overweight/obese adults aged 22-55 years were recruited in this study. All the subjects received PBF measurement by both MF-BIA and DXA. The consistency in PBF measurement by MF-BIA and DXA was evaluated by using interclass correlation coefficients (ICC), then the correction prediction models was established.
Results: The differences in PBF measurement in male subjects and female subjects between MF-BIA and DXA were statistical significant (all P < 0.01), the mean difference values were -6.5% for overweight males and -4.3% for obese males, -2.5% for overweight females and 0.5% for obese females, respectively. The difference in ICC of PBF between MF-BIA and DXA measurement were statistically significant in all subgroups (P < 0.01). The ICC was 0.746 for overweight males, 0.807 for obese males, 0.628 for overweight females and 0.674 for obese females, respectively. The correction prediction models included: PBF (DXA) = 13.425 + 0.719 × PBF (MF-BIA) for overweight males; PBF (DXA) = 12.572 + 0.741 × PBF (MF-BIA) for obese males; PBF (DXA) = 9.785 + 0.802 × PBF (MF-BIA) for overweight females; PBF (DXA) = 20.348 + 0.532 × PBF (MF-BIA) for obese females.
Conclusion: The consistency in PBF measurement in overweight/obese adults by MF-BIA and DXA was poor in China. Correction should be conducted when MF-BIA is used in the measurement of PBF.