Introduction:
Nutritional impairment in patients with chronic kidney disease (CKD) is due to decreased body stores of both protein and fat. We need a tool that can be used in clinics to determine and monitor fat composition with a special focus on normalizing fat measurements to height in these children. Bio-impedance analysis (BIA), a portable and simple tool, has been used to estimate body fat in children with CKD but needs validation against the reference tool dual energy X-ray absorptiometry (DXA). The purpose of the cross-sectional study was to estimate the prevalence of low body fat in children with stages 2-5 CKD (non-dialysis) and CKD 5D (dialysis), and to compare fat measures from two different methods namely BIA and DXA.
Method:
Children in stages 2–5 CKD (
n
= 19) and in CKD 5D (
n
= 14) were recruited for assessment of fat mass (FM, Kg) by BIA and DXA, from which percent body fat (BF %) and fat mass index (FMI, Kg/M
2
) were obtained. Low body fat was defined as <5
th
age and gender centile for BF% or FMI by DXA and BF% by BIA.
Results:
Low body fat was detected equally using BF% and FMI in 18% of children by DXA while only 12% were detected using BF% by BIA. In children with CKD2–5, a good degree of reliability was found with FMI measurements (ICC 0.76 CI [0.48,0.9]) and poor reliability in children with CKD 5D (ICC 0.58 CI [0.1,0.84]). BF% had poor to fair reliability in the children with CKD 2-5 and CKD 5D (ICC 0.64 [0.28,0.84] and 0.53 [0.02,0.82]), respectively. Comparing BF% and FMI obtained by BIA and DEXA, BIA overestimated BF% by 3.5% in comparison to DXA.
Conclusion:
In children with CKD, body fat is preserved in the majority. Among the two measures of fat, BF% estimated by BIA did not compare well with DXA while FMI measure was comparable with a lower bias. However, due to lack of reference values in Indian children for FMI obtained by BIA, BIA cannot be used to measure fat in this population.