Agreement between anthropometry and bioelectrical impedance for measuring body fat in nonobese and obese nondialyzed chronic kidney disease patients.

OBJECTIVES

We investigated reproducibility and agreement between anthropometry (ANT) and bioelectrical impedance (BIA) in nonobese and obese nondialyzed chronic kidney disease (CKD) patients, and examined the factors influencing the agreement between these two methods.

PATIENTS AND METHODS

One hundred and five clinically stable CKD patients, comprising stages 3 and 4 of CKD, had their body fat (BF) assessed by ANT and BIA. Patients were stratified into nonobese (body mass index [BMI] <25 kg/m(2), n = 48, 26 males, aged 64.5 +/- 13.6 years) and overweight/obese (BMI > or =25 kg/m(2), n = 57, 35 males, aged 63.1 +/- 12.1 years; +/-SD in all values) groups.

RESULTS

In the nonobese group, BF as assessed by ANT was similar to body fat as assessed by BIA (11.4 +/- 3.8 kg vs 11.7 +/- 4.1 kg, respectively, no significance). The concordance correlation coefficient (CCC) was indicative of good reproducibility for males (0.67; 95% confidence interval [CI], 0.41 to 0.83) and females (0.88; 95% CI, 0.71 to 0.95), and the mean inter-method difference between ANT and BIA and limits of agreement were -0.9 (-6.8 to 4.9) kg for males, and 0.4 (-3.4 to 4.3) kg for females. In the overweight/obese group, a weak CCC was observed for males (0.46; 95% CI, 0.26 to 0.62) and females (0.53; 95% CI, 0.29 to 0.70), and the mean inter-method difference between both methods was 5.7 (-3.2 to 14.7) kg for males, and 6.4 (-1.8 to 14.7) kg for females. In the multiple regression analysis (adjusted r(2) = 0.42, n = 105), BMI, sex, body water, and age were independent determinants of the agreement between ANT and BIA.

CONCLUSIONS

For nonobese patients, ANT and BIA can be applied equally well for assessing BF. For overweight/obese patients a weak agreement was observed. Body mass index, sex, body water, and age were the factors influencing the agreement between ANT and BIA.