Comparison of resting metabolic rate prediction equations in college-aged adults

Abstract

Prediction equations have been considered an accurate method for estimating resting metabolic rate (RMR) across multiple populations, but their accuracy for college-aged individuals not on an athletics team remains to be determined. Sixty-two college-aged (18–30 yrs) males (n = 31) and females (n = 31) had their RMR measured (RMRm), using indirect calorimetry, and body composition assessed via air-displacement plethysmography. The World Health Organization (WHO), Mifflin−St Jeor (Mifflin), Harris−Benedict (HB), Cunningham, and Nelson equations were used to estimate RMR. No difference was observed between the Cunningham and RMRm regardless of sex (p ≥ 0.05). All other prediction equations estimated a significantly lower RMR for males (p < 0.05). The Mifflin and Nelson equations predicted an RMR that was significantly lower than RMRm for females (p < 0.05). When compared with RMRm, no difference was detected for females using the WHO, HB, or Cunningham (p ≥ 0.05). Only the Nelson equation predicted an RMR that was outside of the clinically acceptable range (±10% of RMRm) regardless of sex. The Cunningham, WHO, and HB equations can accurately predict RMR for college-aged males and females. RMR prediction equations used in this study are less accurate for those with greater RMRs. Novelty: For adults 18–30 years old that are not on an athletics team, the Cunningham equation can accurately predict RMR. The Nelson equation should not be used to predict RMR for this population. There is a systematic bias for RMR prediction equations to underestimate higher measured RMR values.