Accuracy of four resting metabolic rate prediction equations: Effects of sex, body mass index, age, and race/ethnicity

• Published in: Journal of science and medicine in sport Vol. 14; no. 4; pp. 344 - 351
• Main Authors: Hasson, Rebecca E, Howe, Cheryl A, Jones, Bryce L, Freedson, Patty S
• Format: Journal Article
• Language: English
• Published: Australia Elsevier Ltd 01.07.2011; Elsevier Limited
• Subjects: Accuracy; Adolescent; Adult; Age Distribution; Air leakage; Analysis of Variance; Basal Metabolism; Body Mass Index; Calorimetry, Indirect - methods; Calorimetry, Indirect - standards; Cardiovascular disease; Data collection; Ethnic Groups - statistics & numerical data; Ethnicity; Female; Humans; Male; Metabolic disorders; Middle Aged; Models, Biological; Obesity; Physical Medicine and Rehabilitation; Prediction equations; Race; Regression Analysis; Resting metabolic rate; Sex Distribution; Sports Medicine; Studies; Young Adult; Obesity; Resting metabolic rate; Prediction equations
• ISSN: 1440-2440; 1878-1861
• DOI: 10.1016/j.jsams.2011.02.010

Abstract Objective: This study compared the accuracy of four commonly used RMR prediction equations to measured RMR obtained from the MedGem® metabolic analyzer. Design and Methods: Height, weight and RMR were measured in 362 healthy individuals [51% female; body mass index (BMI): 17.6–50.6 kg m−2 ; ages: 18–60 years; 17.4% non-white]. Following a 4 h fast, participants rested in the supine position after which RMR was measured. RMR was estimated using four commonly used prediction equations: Harris–Benedict, Mifflin–St. Jeor, Owen, and WHO/FAO/UNU. Accuracy was determined by calculating the percentage of predicted RMR values that were within ±10% of measured RMR values. Main effects of sex, BMI, age, and race/ethnicity were assessed using repeated measures ANCOVAs. Results: For all participants combined, the Harris–Benedict, Mifflin, and WHO/FAU/UNU equations similarly predicted RMR values within ±10% of measured RMR values (57.5, 56.4, and 55.2% of the sample, respectively). When participant data were stratified by sex, BMI, age, and race/ethnicity, the accuracy of each regression equation varied dramatically. The Harris–Benedict equation over-predicted RMR in 18–29 year olds. The Owen equation under-predicted RMR in both sexes, all three BMI categories, 18–49 year olds and White participants. The Mifflin under-predicted RMR in both sexes, normal weight individuals, 40–60 year olds, and non-Hispanic White participants. The WHO/FAO/UNU over-predicted RMR in males, overweight participants, and 50–60 year olds. Conclusions: When examining the entire sample, the Harris–Benedict, Mifflin, and WHO/FAU/UNU equations yielded similar levels of agreement with the MedGem® measured RMR. However, clinical judgment and caution should be used when applying these prediction equations to special populations or small groups.