Accuracy of total energy expenditure predictive equations after a massive weight loss induced by bariatric surgery

Abstract

Background and aims: Predictive equations remain the clinical tool of choice to estimate the energy expenditure, however, poor accuracy has been found when applied in patients with severe obesity. The aim of this study was to test the accuracy of the total energy expenditure (TEE) those obtained by predictive equations of resting energy expenditure (REE) times individual estimates of metabolic equivalents (MET), taking as reference the TEE measured by doubly labeled water (DLW), before, six and twelve months after bariatric surgery. Subjects and methods: Twenty class III obese women (age: 29.4 ± 5.1 years; BMI: 44.9 ± 2.5 kg/m2), approved for Roux-en-Y gastric bypass participated in this study. TEE and body composition was measured after administration of a fixed dose of DLW. Predictive equations of REE were selected: Dietary Reference Intake (DRI), World Health Organization, Oxford, Harris–Benedict, Mifflin, De Lorenzo, Lazzer and Muller and their collaborators. The MET values were obtained individually by triaxial accelerometer. Results: The patients showed 65% of excess weight loss. The body fat mass decreased 17% after 1 year of surgery. TEE (2930 ± 525 kcal.day−1) decreased by 20% (p < 0.05) by the sixth postoperative month (2319 ± 430 kcal.day−1), increasing 10% by the twelfth month (2538 ± 336 kcal.day−1). The Harris and Benedict (accuracy: 65%) and DRI equations (accuracy: 60%) yielded better results in the prediction of TEE values at pre-surgery. In the sixth month after surgery, only the equation of Harris and Benedict kept accuracy above 50%. At twelve months post-surgery, only the equation of Lazzer et al. considering body composition showed better prediction (accuracy: 50%) in this period. Conclusion: None of the prediction equations tested was accurate for estimating TEE for the 3 periods evaluated; however, while there are no reports of specific equations for class III obese women, the Harris and Benedict x MET and DRI equations, can describe the TEE with acceptable accuracy. After surgery, the best equation to be used will depend if the patient has been treated for weight loss and in which post-treatment period it is used. Registration of clinical trial as an observational study in Brazilian Clinical Trials Registry: RBR-8k5jsj. Universal Trial Number: U1111-1206-0858.