Objective: To compare the validity of skinfold anthropometry (SFA) and bioelectrical impedance (BIA) for the measurement of body composition in patients with chronic renal failure (CRF).
Design: We compared measurement of % total body fat (%TBF) and fat-free mass (FFM) by BIA and SFA, with dual energy X-ray absorptiometry (DEXA) as the criterion method to determine their value in groups of patients with CRF and a group of normal controls.
Setting: Renal Unit, Leeds General Infirmary and Centre for Bone and Body Composition Research, University of Leeds, UK.
Results: 95% levels of agreement of BIA and SFA with DEXA (using the method of Bland & Altman) showed considerable variation, which was greater for both techniques in CRF patients than in a normal control group. In normal subjects, BIA derived values for % TBF and FFM showed a closer agreement with DEXA than did SFA (% TBF BIA-DEXA -8.2 to +3.8%; %TBF SFA-DEXA -4.6 to +12.1%; FFM BIA-DEXA -2.5 to +5.8kg; FFM SFA-DEXA -7.9 to +3.8 kg). However, in CRF patients no differences in the 95% levels of agreement were observed for % TBF or FFM derived from BIA or SFA compared with DEXA (%TBF BIA-DEXA -13.7 to +8.3%; %TBF SFA-DEXA -13.0 to +9.4%; FFM BIA-DEXA -5.1 to +9.6 kg; FFM SFA-DEXA -5.6 to +9.1 kg).
Conclusions: In CRF, errors of both BIA and SFA in comparison with DEXA are greater than in normal subjects. The magnitude of the limits of intermethod difference are relatively much greater for measurement of % TBF than FFM for both BIA and SFA in CRF patients and control subjects.