A randomized prospective trial of gastric bypass versus vertical banded gastroplasty for morbid obesity and their effects on sweets versus non-sweets eaters.

Vertical banded gastroplasty (VBGP) was compared with Roux-en-Y gastric bypass (RYGBP) in a randomized prospective trial that included preoperative dietary separation of "sweets eaters" versus "non-sweets eaters." Randomization was stopped at 9 months after 20 patients had undergone each procedure because a greater weight loss (p less than 0.05) was noted after RYGBP than VBGP. This difference became more significant (p less than 0.001) at each 3-month interval through 3 years, when patients who had VBGPs had lost 37 +/- 20% of excess weight compared with 64 +/- 19% for patients who had RYGBPs. The members of the groups were comparable with regard to age, sex, eating habits, morbidity rates before surgery, ideal body weight, and weight before surgery. Although there was no significant difference between the loss of excess weight in "sweets eaters" (69 +/- 17%) or "non-sweets eaters" (67 +/- 17%) after RYGBP at 1 year, "sweets eaters" who had VBGPs lost significantly less excess weight (36 +/- 13%) than did "non-sweets eaters" who had VBGPs (57 +/- 18%), p less than 0.02, or "sweets eaters" who had RYGBPs, p less than 0.0001. No significant differences were noted for electrolytes, renal or liver function tests, and most vitamins between patients who had VBGPs and RYGBPs; however, patients who had RYGBPs had lower (p less than 0.05) serum vitamin B12 levels (286 +/- 149 pg/dl) than did patients who had VBGPs (461 +/- 226 pg/dl) at 2 years. By 3 years, the vitamin B12 levels were equal in members of the two groups. Five patients who had RYGBPs required endoscopic stomal dilatation for stomal stenosis and one had a marginal ulcer develop, which responded to cimetidine. RYGBP was clearly superior to VBGP for "sweets eaters," probably because of the development of dumping syndrome symptoms. However, RYGBP was associated with a larger number of correctable problems. Thus, it is important to evaluate a patient's eating habits before surgery for morbid obesity; "non-sweets eaters" probably should be treated with VBGP and "sweets eaters" with RYGBP.