Accuracy of Equations Used to Predict Resting Metabolic Rate (RMR) Requirements Among Healthy and Acutely Ill Females in Trinidad and Tobago; Aged 18 to 65 Years

Abstract

Abstract Objectives To determine the level of accuracy of the recommended resting metabolic rate (RMR) prediction equations among healthy and acutely ill females: aged 18 to 65 years, in Trinidad and Tobago. Methods Following informed consent and enrollment, sixty female (acutely ill 30; healthy 30) volunteers had anthropometry and RMR (MedGem® indirect calorimeter, Micro life, USA) measured using recommended procedures. RMR from prediction equations were compared to RMR measured by indirect calorimetry with values between ±10% of measured RMR being considered accurate. The university's ethics committee approved the study. Results The top-two ranked recommended RMR prediction equations for acutely ill females, in decreasing order of accuracy were Owen (46.7%) and Bernstein (40%). Among the healthy females, the top-two ranked recommended RMR prediction equations with a similar level of accuracy (46.7%) were Livingston and Kohlstadt, and DeLorenzo. The population-derived RMR prediction equations had 56.7% and 70% accuracies among the acutely ill and healthy females respectively. These were significantly higher than the top-ranked recommended prediction equations for both groups. Notably, limiting the risk of malnutrition by at least 5%: through diet quality by way of accurate energy predictions could improve health-related quality of life. Increasing the predictability of energy needs within any population can also ensure accuracy of RMR per weight in kilogram (kcal/kg) needed daily and energy balance. On average, this female population utilized 18.4 kcal of energy per kilogram body weight (kcal/kg). Conclusions All the recommended RMR prediction equations resulted in biases > 50%, thus our population-derived RMR equations can be used as a superior alternative among participants to determining the energy-needs of acutely ill individuals as well as healthy females. Substituting the commonly used prediction equations with population-specific equations can increase the level of accuracy by at least 10%, thus limiting the risk of malnutrition by at least 5% and improving health-related quality of life. Funding Sources NIL.