Agreements between indirect calorimetry and prediction equations of resting energy expenditure in end-stage renal disease patients on continuous ambulatory peritoneal dialysis

Equations are frequently used to estimate resting energy expenditure (REE) in a clinical setting. However, few studies have examined their accuracy in end-stage renal disease (ESRD) patients. To investigate agreement between indirect calorimetry and several REE estimating equations in 38 ESRD patients on peritoneal dialysis, we performed indirect calorimetry and compared the results with REEs estimated using 5 equations [Harris-Benedict (HBE), Mifflin, WHO, Schofield, and Cunningham]. Measured REE was 1393.2+/-238.7 kcal/day. There were no significant differences between measured and estimated REEs except Mifflin (1264.9+/-224.8 kcal/day). Root mean square errors were smallest for HBE, followed by Schofield, Cunningham, and WHO, and largest for Mifflin (171.3, 171.9, 174.6, 175.3, and 224.6, respectively). In Bland-Altman plot, correlation coefficients between mean values and differences were significant for HBE (r=0.412, p=0.012) and tended to be significant for Cunningham (r=0.283, p=0.086). In DM patients and patients with overhydration, HBE showed significant underestimation when REE increased. In ESRD patients on continuous ambulatory peritoneal dialysis (CAPD), REE-estimating equations have no significant differences from indirect calorimetry, except Mifflin. However, HBE showed greater bias than others when REE was high.