Use of both quantitative and qualitative methods to improve assessment of resting energy expenditure equation performance in hospitalized adults

Clin Nutr ESPEN. 2018 Apr:24:120-126. doi: 10.1016/j.clnesp.2018.01.001. Epub 2018 Jan 12.

Abstract

Objective: To introduce the use of qualitative assessment in energy expenditure (EE) equation research to improve the understanding of performance of the equations in the clinical setting.

Patients and methods: Hospitalized individuals who had an indirect calorimetry (IC) measurement during their hospital stay from 2010 to 2012 were included in the study (n = 59). An additional 1000 patients hospitalized during this time were used to limit the IC cohort to a more "clinically relevant" BMI range (n = 46). The following estimation equations were assessed: Harris-Benedict, 25 kcal/kg using actual body weight, Mifflin St. Jeor, Ireton-Jones, Penn State, and Owen. Bland-Altman plots with Loess curves were generated to compare estimated basal caloric needs between EE equations and IC values.

Results: This study found a large amount of variability with all EE equations. As the mean calorie level increased, the Harris Benedict, Mifflin St. Jeor, Penn State, and Owen equations all tended to increasingly under-predict caloric need.

Conclusion: In a research setting a qualitative assessment of EE equations can provide a more comprehensive understanding of equation performance by complementing traditional quantitative methods. The addition of a Loess curve to the Bland-Altman plot further enhances qualitative assessment.

Keywords: Calories; Clinical nutrition; Energy expenditure; Energy expenditure equation; Indirect calorimetry; Metabolism.

MeSH terms

  • Adult
  • Aged
  • Basal Metabolism / physiology*
  • Body Mass Index
  • Calorimetry, Indirect
  • Energy Metabolism / physiology*
  • Evaluation Studies as Topic
  • Female
  • Hospitalization*
  • Humans
  • Male
  • Middle Aged
  • Nutrition Assessment
  • Nutritional Requirements / physiology*
  • Qualitative Research
  • Reproducibility of Results
  • Rest