Journal Article

Are prediction equations reliable for estimating resting energy expenditure in chronic kidney disease patients?

Maria A. Kamimura, Carla M. Avesani, Ana P. Bazanelli, Flavia Baria, Sergio A. Draibe and Lilian Cuppari

in Nephrology Dialysis Transplantation

Published on behalf of European Renal Association - European Dialysis and Transplant Assoc

Volume 26, issue 2, pages 544-550
Published in print February 2011 | ISSN: 0931-0509
Published online July 2010 | e-ISSN: 1460-2385 | DOI:
Are prediction equations reliable for estimating resting energy expenditure in chronic kidney disease patients?

Show Summary Details


Background. The determination of resting energy expenditure (REE) is the primary step for estimating the energy requirement of an individual. Although numerous equations have been formulated for predicting metabolic rates, there is a lack of studies addressing the reliability of those equations in chronic kidney disease (CKD). Thus, the aim of this study was to evaluate whether the main equations developed for estimating REE can be reliably applied for CKD patients.

Methods. A total of 281 CKD patients (124 non-dialysis, 99 haemodialysis and 58 peritoneal dialysis) and 81 healthy control individuals were recruited. Indirect calorimetry and blood sample collection were performed after a 12-h fasting. Two most traditionally used equations for estimating REE were chosen for comparison with the REE measured by indirect calorimetry: (i) the equation proposed by Harris and Benedict, and (ii) the equation proposed by Schofield that is currently recommended by the FAO/WHO/UNU.

Results. Schofield’s equation exhibited higher REE [1492 ± 220 kcal/day (mean ± SD)] in relation to Harris and Benedict’s equation (1431 ± 214 kcal/day; P < 0.001), and both prediction equations showed higher REE in comparison with the reference indirect calorimetry (1352 ± 252 kcal/day; P < 0.001). In patients with diabetes, inflammation or severe hyperparathyroidism, the REE estimated by the Harris and Benedict equation was equivalent to that measured by indirect calorimetry. The intraclass correlation of the REE measured by indirect calorimetry with the Schofield’s equation was r = 0.48 (P < 0.001) and with the Harris and Benedict’s equation was r = 0.58 (P < 0.001). According to the Bland and Altman analysis, there was a large limit of agreement between both prediction equations and the reference method. Acceptable prediction of REE (90–110% adequacy) was found in 47% of the patients by using the Harris and Benedict’s equation and in only 37% by using the Schofield’s equation.

Conclusions. The most traditionally used prediction equations overestimated the REE of CKD patients, and the errors were minimized in the presence of comorbidities. There is a need to develop population-specific equations in order to adequately estimate the energy requirement of these patients.

Keywords: chronic kidney disease; energy requirement; indirect calorimetry; prediction equations; resting energy expenditure

Journal Article.  4346 words.  Illustrated.

Subjects: Nephrology

Full text: subscription required

How to subscribe Recommend to my Librarian

Users without a subscription are not able to see the full content. Please, subscribe or purchase to access all content.