Validating predictive factors for resting energy expenditure of adolescents in Indonesia

Our objectives were to (1) compare the measured and predicted resting energy expenditure (REE) using 18 equations in normal-weight and obese adolescents, and (2) examine potential predictive factors for accurately estimating REE in obese adolescents. The subjects chosen were 41 obese and 35 normal-weight adolescents living in an urban area in Indonesia. REE was measured using indirect calorimetry and compared with predicted REE determined via 18 equations. New predictive equations were developed via stepwise multiple regression analysis. Mean differences between the predicted and measured REE were within ± 5% with 13 equations tested in normal-weight subjects, and only six equations in obese subjects. New predicted REE derived from our new equations was significantly correlated with measured REE (P < 0.0001, R(2) = 0.5-0.7). Measured REE was overestimated by existing predictive equations, especially in obese subjects. New predictive equations need to be developed that take age, climate and ethnicity into account, in addition to physical characteristics.