Cross‐Validation of Recent and Longstanding Resting Metabolic Rate Prediction Equations

Abstract

Resting metabolic rate (RMR) measurement is time consuming and requires specialized equipment. Prediction equations provide an easy method to estimate RMR; however, their accuracy likely varies across individuals. Understanding the factors that influence predicted RMR accuracy at the individual level will help to revise existing or develop new equations that are more accurate. The objective of the present study was to test the validity of established (Harris‐Benedict, WHO, Mifflin‐St. Jeor) and contemporary (3 meta equations of Sabounchi and colleagues, 1 equation of Nelson and colleagues, 1 by Wang and colleagues) RMR prediction equations in healthy adults.DesignPredicted RMR was tested for agreement with indirect calorimetry. Men and women (n=30) age 18–65y from Grand Forks, ND were recruited and included for analysis during Spring/Summer 2014. Participants were nonobese to obese, primarly Caucasian and a representive sample of a healthy, free‐living population. The main outcome measure was bias between measured (indirect calorimetry) and predicted RMR. The methods of Bland and Altman were employed to determine mean bias (predicted minus measured RMR, kcal/day) and limits of agreement between predicted and measured RMR. Repeated measures ANOVA was utilized to test for differences in RMR predicted from each equation versus the measured RMR. Mean bias was lowest for the Harris‐Benedict (4 ± (2 SD) 386 kcal/24 hrs) and WHO (25 ± 402 kcal/24 hrs) equations and these mean predicted RMR did not differ from measured. Mean RMR predictions from all other equations significantly differed from indirect calorimetry. The 2 SD limits of agreement were moderate or large for all equations tested ranging from 282 to 445 kcal/24 hrs. Prediction bias was positively associated with the magnitude of RMR and with FFM. In conclusion, the traditional Harris‐Benedict and WHO equations were the most accurate. However, even these equations did not perform well at the individual level. As FFM increased, the prediction equations further underestimated RMR.Support or Funding InformationSupport: USDA‐ARS 3062‐51000‐051‐00D