Agreement between predictive equations and indirect calorimetry in estimating resting metabolic rate among free‐living older adults with sarcopenia: A prospective observational study

Resting metabolic rate (RMR) accounts for the largest proportion of daily total energy expenditure. The purpose of this study is to systematically evaluate the agreement of predictive equations for RMR and indirect calorimetry (IC) in a population with sarcopenia. The study included 43 individuals with sarcopenia aged 60 and above (diagnosed according to the AWGS 2019 criteria for sarcopenia). One-on-one measurements and questionnaires were used to obtain information including height, weight, body composition, gait speed, and grip strength. A total of 11 different predictive equations were used to estimate RMR and were compared with the IC in multiple dimensions. The RMR of the study participants ranged from 652 to 1244 kcal/day, with an average of 920.5 ± 150.3 kcal/day. All predictive equations exhibited an accuracy rate of less than 40%, with the Mifflin equation based on body weight (Mifflin_W) showing the lowest bias (16.54%), the lowest root mean square error (RMSE) value (216.51 kcal/day), the highest Lin's Concordance Correlation Coefficient (CCC) value (0.40) and the lowest median absolute percentage error (MAPE) (16.52%). The predictions of the Porter equations were not influenced by age or body mass index (BMI), making its predictions more stable, whereas the biases of other equations were significantly affected by age or BMI. Despite overall poor performance by all equations, the Mifflin_W equation showed the best agreement with IC measurements and may be used with caution when IC is unavailable, noting its tendency to overestimate RMR.