Madhu Sudhan Karri, Lesa Dawman, Deepthika Sadasivam, Pujitha Vallabhaneni, Shailza Mahajan, Tulika Singh, Harvinder Kaur, Jaivinder Yadav, Karalanglin Tiewsoh
Despite strong correlations, BIA showed poor concordance with DXA for every parameter assessed, with clinically unacceptable limits of agreement; correlation does not imply interchangeability. BIA may retain a role in population-level nutritional screening in resource-limited settings, although its ability to track change was not tested here; DXA remains necessary for individual-level assessment in paediatric CKD. What is known • Children with CKD experience progressive lean tissue loss, fat redistribution, and impaired bone mineralisation contributing to growth failure. • DXA is the reference standard for body composition assessment, but it is limited by cost, radiation, and availability. What is new • BIA shows a strong to very strong correlation with DXA but poor concordance (Lin's CCC < 0.90) for every body composition and bone parameter in paediatric CKD stages 3-5. • BIA cannot detect stage-wise differences in body composition that DXA identifies, limiting its diagnostic precision in this population.
INTRODUCTION: Children with chronic kidney disease (CKD) exhibit progressive body composition alterations. While dual-energy X-ray absorptiometry (DXA) is the reference standard, its routine use is limited by cost, radiation, and availability. Bioelectrical impedance analysis (BIA) offers a portable, radiation-free alternative, but its accuracy relative to DXA in paediatric CKD remains insufficiently evaluated.
METHODS: This cross-sectional study enrolled 102 children aged 5-14 years with CKD stages 3-5. Same-day BIA and DXA measurements were performed. Agreement between methods was assessed using Pearson's correlation coefficient, Lin's concordance correlation coefficient (CCC), and Bland-Altman analysis. Associations with CKD stage and sex were examined.
RESULTS: BIA showed strong to very strong correlations with DXA for fat mass (FM; r = 0.93), fat-free mass (FFM; r = 0.91), and total bone mineral content (BMC; r = 0.84). Lin's CCC indicated poor concordance for all parameters (all < 0.90): FFM 0.87, total BMC 0.81, total bone mineral density (BMD) 0.74, FM 0.66 and body fat percentage (BF%) 0.29. BIA systematically underestimated FM and overestimated FFM relative to DXA. DXA detected significant stage-wise reductions in FM, FMI, FFM and FFMI across CKD stages, whereas BIA detected only the decline in FFMI (p = 0.012).
CONCLUSIONS: Despite strong correlations, BIA showed poor concordance with DXA for every parameter assessed, with clinically unacceptable limits of agreement; correlation does not imply interchangeability. BIA may retain a role in population-level nutritional screening in resource-limited settings, although its ability to track change was not tested here; DXA remains necessary for individual-level assessment in paediatric CKD. What is known • Children with CKD experience progressive lean tissue loss, fat redistribution, and impaired bone mineralisation contributing to growth failure. • DXA is the reference standard for body composition assessment, but it is limited by cost, radiation, and availability. What is new • BIA shows a strong to very strong correlation with DXA but poor concordance (Lin's CCC < 0.90) for every body composition and bone parameter in paediatric CKD stages 3-5. • BIA cannot detect stage-wise differences in body composition that DXA identifies, limiting its diagnostic precision in this population.