Skip to main content
ARS Home » Plains Area » Grand Forks, North Dakota » Grand Forks Human Nutrition Research Center » Healthy Body Weight Research » Research » Publications at this Location » Publication #316732

Research Project: Dietary Guidelines Adherence and Healthy Body Weight Maintenance

Location: Healthy Body Weight Research

Title: Cross-validation of resting metabolic rate prediction equations

Author
item Flack, Kyle
item Siders, William
item Johnson, Luann - University Of North Dakota
item Roemmich, James

Submitted to: American Journal of Clinical Nutrition
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/18/2016
Publication Date: 4/30/2016
Publication URL: http://handle.nal.usda.gov/10113/62825
Citation: Flack, K.D., Siders, W.A., Johnson, L., Roemmich, J.N. 2016. Cross-validation of resting metabolic rate prediction equations. American Journal of Clinical Nutrition. doi:10.1016/j.jand.2016.03.018.

Interpretive Summary: Knowledge of the resting metabolic rate (RMR) is necessary for determining how many calories a person needs. Measuring RMR is time consuming and requires expensive equipment. Prediction equations provide an easy method to estimate RMR. This study tested the accuracy of 3 prediction equations (Harris-Benedict, WHO, Mifflin-St.Jeor) that have been used for many years and 5 contemporary equations (3 meta equations of Sabounchi and colleagues, and equations developed by Nelson and colleagues and Wang and colleagues) in healthy adult subjects. We also sought to determine whether individual differences in body composition were associated with prediction bias of the equations. We found that on average predictions from the older Harris-Benedict and WHO equations did not differ from directly measured RMR. The Harris-Benedict equation under-estimated RMR by a mean of 14 kcal/24 hrs, whereas the WHO equation under-estimated RMR by 25 kcal/24 hrs. More recently developed equations by Nelson and colleagues (190 kcal/24 hrs) and Wang and colleagues (183 kcal/24 hrs) produced the greatest under-predictions of RMR. Prediction errors were associated with the amount of fat-free mass (FFM) a person has. Greater FFM was associated with greater under-estimation of RMR. In healthy adults the contemporary equations (Sabounchi, Wang, Nelson and collegues) as well as the Mifflin-St.Jeor did not improve upon those of the traditional Harris-Benedict or WHO equations. As FFM increased, the prediction equations further underestimated RMR.

Technical Abstract: Background: Knowledge of the resting metabolic rate (RMR) is necessary for determining individual total energy requirements. Measurement of RMR is time consuming and requires specialized equipment. Prediction equations provide an easy method to estimate RMR; however, the accuracy of these equations likely varies at the individual level. Objective: To test the validity of 3 established prediction equations (Harris-Benedict, WHO, Mifflin-St.Jeor) and 5 contemporary equations in healthy adult subjects (3 meta equations of Sabounchi and colleagues, and equations developed by Nelson and colleagues and Wang and colleagues). We also sought to determine whether individual differences in body composition were associated with prediction bias of the equations. Design: Prediction equation-derived estimates of RMR were tested against values obtained from indirect calorimetry in 30 adults. Results: Predicted values from the Harris-Benedict and WHO equations did not differ from indirect calorimetry. The Harris-Benedict equation under-predicted RMR by a mean ± SD of 13.8 ± 192.8 kcal/24 hrs, whereas the WHO equation under-predicted RMR by 24.8 ± 201.0 kcal/24 hrs. Equations by Nelson and colleagues (189.5 ± 159.9) and Wang and colleagues (182.6 ± 169.2) produced the greatest under-predictions of RMR. Prediction bias was associated with the magnitude of fat-free mass (FFM). For all equations except that of Nelson and colleagues, greater FFM was associated with greater under-prediction of RMR. Conclusions: In healthy adults the contemporary equations (Sabounchi, Wang, Nelson and collegues) as well as the Mifflin-St.Jeor did not improve upon those of the traditional Harris-Benedict or WHO equations. As FFM increased, the prediction equations further underestimated RMR.