Background: Gastroenterology is a beneficial treatment of morbidly obese type 2 diabetes mellitus (T2DM). We aimed to identify the predictors for the treatment of T2DM obese patients. Methods: A retrospective study consisting of 531 patients undergoing laparoscopic gastric banding (LGB), laparoscopic mini-gastric bypass (LMGB) and laparoscopic sleeve gastrectomy (LSG) from January 2004 to May 2007 was performed. Patients with preoperative fasting serum glucose concentration of more than 126 mg/dl were diagnosed as T2DM. A postoperatively fasting serum glucose level of less than 110 mg/dl was considered to be remission of T2DM. Results: Of the 531 patients, 62 (11.6%) were diagnosed as T2DM, including 23 men and 39 women, with a mean age of 31.8 ± 9.2 years, and a mean body mass index (BMI) of 40.0 kg/m2. The mean glucose at 3, 6, and 12 months after surgery were 100.1 mg/dl, 95.1 mg/dl and 91.8 mg/dl, respectively. The mean body weight loss one year after surgery was 9.4% for LGB, 31.4% for LSG and 37.1% for LMGB, respectively. Among these operation methods (LGB, LMGB and LSG), the BMI, body weight, waist circumference, serum lipid profile and serum factors associated with glucose metabolism were significantly different during the one-year postoperative follow-up. Remission rate of T2DM was achieved in 84.8%, 58.8% and 58.3% of patients for LMGB, LGB and LSG, respectively. The best operative method for the remission of T2DM was LMGB. Using an artificial neural network (ANN) data mining technique, waist circumference, operative methods and C-peptide were significantly predictors for the remission of T2 DM. Conclusion: One year after gastrointestinal surgery, improvement of serum lipid profiles and serum data related to glucose metabolism in the different operative methods were noticed. LMGB seems to be the most effective procedure for the reduction of serum glucose levels compared with LAGB and LSG.
ASJC Scopus subject areas