RESUMO
INTRODUCTION: Assessment of body composition is paramount in early assessment of nutritional status impairments due to excess or deficit. There are, however, few field reliable methods for this objective for patients with chronic renal failure (CRF.). OBJECTIVE: To assess the reliability of the estimations of body composition by different methods as compared to dual energy X-ray absorptiometry (DEXA) as the gold standard method in patients with CRF and on regular chronic haemodialysis. PATIENTS AND METHODS: We assessed body composition in 30 haemodialysis patients (46.9 +/- 15.1 years (18-76); BMI 25.9 +/- 5.7 kg/m(2) (18.1-41.5)), observing agreement in the percentage of fat mass (%FM) between the sum of the 4 folds (SP; calibrator Lange) and bioimpedantiometry by using different equations (BIA; Biodynamics 450) versus DEXA (Lunar DPX-L). RESULTS: (X +/- SD) By BMI, 3 subjects had low weight (10%), 14 normal weight (46.7%), 7 overweight (23.3%), and 6 obesity (20%). The %FM with SP (30.7 +/- 7.1%) significantly differed from DEXA (27.3 +/- 10.3%; p < 0.001). With BIA there was a significant difference in %FM with the Deurenberg and Formica equations. The %FM obtained with the manufacturer's equations (Segal, Lukaski and Kyle) did not show a significant difference from DEXA. With Kyle's equation we observed a better agreement (difference with DEXA: -0.58 +/- 4.2%). CONCLUSIONS: We found a low percentage of patients with low weight as compared to previous studies. The skin folds show low reliability to estimate the fat mass. The bioimpedantiometry, using Kile's equation may be a good filed method to assess haemodialysis patients.