Accuracy of resting metabolic rate prediction equations among healthy adults in Trinidad and Tobago

Background: Over 50% of adults in Latin America and the Caribbean have a body mass index (BMI) >= 25 suggesting excess energy intakes relative to energy expenditure. Accurate estimation of resting metabolic rate (RMR), the largest component of total energy requirements, is crucial to strategies aimed at reducing the prevalence and incidence of overweight and obesity. Aim: We evaluated the accuracies of established and locally developed RMR prediction equations (RMRP) among adults. Methods: Four hundred adult volunteers ages 20 to 65 years had RMR measured (RMRM) with a MedGem (R) indirect calorimeter according to recommended procedures. RMRP were compared to RMRM with values + 10% of RMRM deemed accurate. Anthropometry was measured using standard procedure. Linear regression with bootstrap analyses was used to develop local RMRP equations based on anthropometric and demographic variables. The University of the West Indies Ethics Committee approved the study. Results: Males had higher mean absolute RMR (p < 0.001) but similar mean age-adjusted measured RMR per kg of body (20.9 vs. 21.5 kcals/day; p = 0.1) to females. The top performing established anthropometry-based RMRP among participants by sex, physical activity (PA) level and BMI status subgroups were Mifflin-St Jeor, Owen, Korth, Harris-Benedict, and Livingston, while Johnstone, Cunningham, Muller (body composition (BC)), Katch and McArdle, Mifflin-St Jeor (BC) were the most accurate BC-based RMRP. Locally developed RMRP had accuracies comparable to their top-ranked established RMRP counterparts. Conclusions: Accuracies of established RMRP depended on habitual PA level, BMI status, BC and sex. Furthermore, locally developed RMRP provide useful alternatives to established RMRP.