Should we still use the Harris and Benedict equations?

Resting metabolic rate (RMR) is commonly predicted using the Harris-Benedict (HB) equations, but an overestimation of 10% to 15% is normally found. More recent studies have proposed equations with a better predictive value. In this study, we explore the relationship between measured RMR and HB in 67 healthy volunteers and in a data set from the literature and compared measured RMR with six more recent equations. Mean differences between RMR and HB were 21%, 12%, 10%, and 4% for the lowest to the highest RMR quartile, respectively, and 20%, 8%, 6%, and -4% for Owen's subjects. Among the six recent equations, only the World Health Organization (WHO) equations predicted RMR within 10% in 100% of the cases. Our results suggest that overestimation of RMR by HB is not a homogenous finding but is inversely related to RMR. This may have important implications for predicting RMR in women and in patients with diminished lean body mass. In addition, the WHO equations appear more precise than the HB equations.