Abstract: Background：The accuracy of bioelectrical impedance analysis (BIA) in measuring body composition of children and adolescents is controversial, and there is a lack of correction formulas that do not rely on the brand and model of BIA instrument. Objective：To evaluate the consistency of BIA and dualenergy Xray absorptiometry (DXA) in measuring body composition in children and adolescents aged 619 years. Design：Crosssectional survey. Methods：The subjects were selected from the 20132023 China Life Cycle Body Composition Study, and 1,841 schoolage children and adolescents were tested using the same BIA and DXA instruments. Intragroup correlation coefficient (ICC), relative error frequency (RE), and BlandAltman analysis were used to evaluate the consistency of measurements of fat mass (FM), nonfat tissue mass (FFM), and limb skeletal muscle mass (ASM). With DXA as the dependent variable and resistance index (RI) as the independent variable, a prediction model for BIA correction was established. Main outcome measures：ICC, RE, and BlandAltman analysis. Results：In different sexes, compared with DXA, BIA underestimated FM (RMSE=2.60, 2.27 kg, P<0.05), underestimated ASM (RMSE=1.02, 0.91 kg, P<0.05), and overestimated FFM (RMSE=3.07, 2.66 kg, P<0.05) for male and female students. The ICC of FM, FFM and ASM of male and female students were all >0.90 (P<0.05). With the increase of age, the proportion of ICC and RE with absolute value <15% in FM increased gradually, while the proportion of ICC and RE with absolute value <15% in ASM first increased and then decreased. With the increase of BMI, the proportion of ICC and RE with absolute value <15% of FM increased gradually. The BIA corrected prediction model is: FM= 2.336+0.871×RI50-1.739×RI500+0.517×SexM1F2-0.478×Age+0.753×Weight (R2=0.940)；FFM=2.795-1.406×RI50+2.177×RI500-0.329×SexM1F2+0.471×Age+0.243×Weight (R2=0.985); ASM=-1.937-0.697×RI50+1.132×RI500-0.389×SexM1F2+0.119×Age+0.095×Weight (R2=0.976). Conclusion：BIA measured ASM, FFM and DXA in schoolage children and adolescents with good agreement, FM and DXA with poor agreement. The corrected model of FM, FFM and ASM measured by BIA has good fitting effect, and can be used to evaluate the body composition of schoolage children and adolescents with different types of multifrequency BIA instruments.

Key words: Bioelectrical impedance analysis, Dualenergy Xray absorptiometry, Fat mass appendicular, Skeletal muscle mass, Consistency analysis