Roux-en-Y gastric bypass (RYGB) results in sustained weight loss and amelioration of comorbid conditions in patients with clinically severe obesity. The mechanism of weight loss after RYGB is not well defined. The objective of this study was to document the changes in measured resting energy expenditure (MREE) over time in patients with clinically severe obesity after RYGB.

We prospectively studied MREE in 70 patients (11 male, 59 female; body mass index [BMI], 40 to 80 kg/m²) treated by RYGB. MREE was measured by indirect calorimetry before operation and at 6 weeks and 3, 6, 12, 18, and 24 months after operation. Patients were stratified to hypometabolic ([HM] MREE less than 85% of Harris-Benedict [HB] predicted; n = 22) or normal metabolic rate ([NM] MREE ±15% HB predicted; n = 48) before operation; mean BMIs were HM, 53.4 ± 11.0 kg/m²; NM, 51.4 ± 9.8 kg/m²; p = not significant. MREE, weight loss, percent excess body weight loss (EWL), and energy intake were determined at each time point.

Overall, MREE was significantly less than HB-predicted REE before operation (90 ± 28%), but rose to become equal to the HB-predicted REE by 6 weeks (96 ± 15%) and remained so. When stratified by initial metabolic rate, MREE increased significantly in the HM patients by 6 weeks, from 1329 ± 604 kcal/day (55% of HB predicted) to 1882 ± 398 kcal/day (88% of HB predicted) (p < 0.001), and MREE remained normal (2332 ± 484 kcal/day to 2029 ± 410 kcal/day) in the NM patients. Percent EWL was similar in both groups at each time. Energy intake was 2603 ± 982 kcal/day before operation and fell to 815 ± 196 kcal/day at 3 months, 969 ± 241 kcal/day at 6 months, 1095 ± 307 kcal/day at 12 months, 1259 ± 466 kcal/day at 18 months, and 1373 ± 620 kcal/day at 24 months, and was similar between the groups at each time point. Percent HB-predicted REE increased significantly after operation despite a significant decrease in energy intake.

RYGB is associated with significant changes in MREE over time. In NM patients MREE fell over time consistent with weight loss but remained normal, whereas patients who were hypometabolic exhibited increases in MREE toward normal. These changes in MREE occurred despite reduced energy intake comparable to a very low calorie diet. This paradoxical effect on MREE may contribute to the enhanced weight loss associated with RYGB.