Predicting the glycemic response to gastric bypass surgery in patients with type 2 diabetes

OBJECTIVEdTo find clinically meaningful preoperative predictors of diabetes remission and conversely inadequate glycemic control after gastric bypass surgery. Predicting the improvement in glycemic control in those with type 2 diabetes after bariatric surgery may help in patient selection. RESEARCH DESIGN AND METHODSdPreoperative details of 154 ethnic Chinese subjects with type 2 diabetes were examined for their influence on glycemic outcomes at 1 year after gastric bypass. Remission was defined as HbA1c #6%. Analysis involved binary logistic regression to identify predictors and provide regression equations and receiver operating char-acteristic curves to determine clinically useful cutoff values. RESULTSdRemission was achieved in 107 subjects (69.5%) at 12 months. Diabetes duration,4 years, body mass.35 kg/m2, and fasting C-peptide concentration.2.9 ng/mL provided three independent preoperative predictors and three clinically useful cutoffs. The regression equation classification plot derived from continuous data correctly assigned 84 % of participants. A combination of two or three of these predictors allows a sensitivity of 82 % and specificity of 87 % for remission. Duration of diabetes (with different cutoff points) and C-peptide also pre-dicted those cases in which HbA1c #7 % was not attained. Percentage weight loss after surgery was also predictive of remission and of less satisfactory outcomes