# Predictive equations commonly used in the clinic underestimate resting energy expenditure compared with whole-room indirect calorimetry in colorectal cancer survivors

## TL;DR

Common formulas used to estimate resting energy expenditure in colorectal cancer survivors are found to be inaccurate compared to a more precise measurement method.

## Contribution

The study evaluates the accuracy of predictive equations for resting energy expenditure in colorectal cancer survivors using whole-room indirect calorimetry.

## Key findings

- Most predictive equations underestimated resting energy expenditure compared to whole-room indirect calorimetry.
- Harris–Benedict, Henry, and FAO/WHO/UNUBIA equations showed the best agreement with measured resting energy expenditure.
- Only 62–68% of predicted values were within ±10% of measured resting energy expenditure.

## Abstract

Accurate methods for estimating resting energy expenditure (REE) are important to ensure adequate nutritional treatment in colorectal cancer (CRC) survivors.

This study aims to determine the agreement between REE estimated by commonly used predictive equations and by whole-room indirect calorimetry (WRIC).

This cross-sectional study included 31 CRC survivors {age: 53–78 y; mean [standard deviation (SD)]; body mass index 28.7 [4.28] kg/m2}, who underwent curative surgery. Body composition was measured by dual-energy X-ray absorptiometry (DXA). Predicted REE from equations in clinical use and derived from DXA and bioelectrical impedance analysis (BIA) were compared against REE measured by 30-min WRIC. Equations included Harris–Benedict, Mifflin–St. Jeor, Food and Agriculture Organization (FAO)/World Health Organization (WHO)/United Nations University (UNU), Henry, Mifflin–St. JeorDXA, and FAO/WHO/UNUBIA. Paired sample t-test, Lin’s concordance correlation coefficient, and Bland–Altman analysis were used to determine the agreement between measured REEWRIC and predicted REE. Accuracy was defined as the percentage of predicted REE values that fell within ± 10% of REEWRIC.

Mean (SD) REEWRIC was 1710 kcal/d (353), and respiratory quotient was 0.79 (0.05). Most equations underestimated REE. Overall, Harris–Benedict, Henry, and FAO/WHO/UNUBIA showed the best overall agreement with REEWRIC. However, these equations showed low accuracy with 65%, 68%, and 62% of predicted REE values within ± 10% of REEWRIC, respectively.

Most predictive equations tended to underestimate REE in CRC survivors compared with REEWRIC. The Harris–Benedict, Henry and FAO/WHO/UNUBIA equations showed both best accuracy and agreement with WRIC. They were still inaccurate, with individual variability for a relevant part of the sample. Future studies need to develop improved predictive equations for CRC survivors. This study was registered at clinicaltrials.gov as NCT01570010 (https://clinicaltrials.gov/study/NCT01570010?locStr=Norway&country=Norway&cond).