Which Baseline Weight Should Be Preferred as Reference for Weight Loss Results? Insights in Bariatric Weight Loss Mechanisms by Comparing Primary and Revision Gastric Bypass Patients

Bariatric weight loss essentially is expressed with reference to the baseline weight, for example, as relative percentages or as absolute body mass index (BMI) points lost from baseline. A different definition of baseline weight would therefore affect all weight loss results. We try to determine which value to prefer for baseline weight in weight loss surgery: the accidental weight at time of operation or the patient-specific steady weight, reflecting a steady personal craving for calories that is independent of the operation. Nadir percentage alterable weight loss (%AWL) outcome of all primary gastric bypass patients in our hospital with a 2-year follow-up is compared twice with nadir %AWL outcome of all revision gastric banding-to-bypass patients: relative to their BMI before their banding and before their banding-to-bypass (Mann-Whitney; p < 0.05). Out of 713 gastric bypass patients with a 2-year follow-up, 82 had revision banding-to-bypass. Total mean baseline BMI is 44.1 kg/m(2); nadir BMI is 29.2 kg/m(2). Difference in mean nadir weight loss between primary (49.4 %AWL) and revision patients is not significant if compared to baseline BMI before gastric banding (47.4 %AWL) but significant if compared to baseline BMI before revision banding-to-bypass (37.7 %AWL). Revision gastric bypass with removal of gastric banding does not affect the "new" weight after the gastric banding but the "old" weight before the banding. Gastric bypass effectiveness was not added to the gastric banding effectiveness; it replaced it. Therefore, the patient-specific steady weight should be preferred for baseline BMI, reflecting an underlying personal craving for calories that remains constant over time and independent of a bariatric procedure. Baseline BMI can be standardized by using the measured weight at first visit before the primary bariatric procedure, also in revision cases.