Factors associated with complete and partial remission, improvement, or unchanged diabetes status of obese adults 1 year after sleeve gastrectomy

Background Laparoscopic sleeve gastrectomy (SG) achieves type 2 diabetes (T2D) remission to various extents, and reasons for such variations are unknown. Objectives We assessed patients' characteristics associated with T2D remission 1 year post SG. Setting University hospital. Methods Retrospective study of 230 T2D patients (18–64 yr) who underwent SG at our institution. We examined pre- and postoperative demographic, anthropometric, biochemical, and clinical characteristics associated with T2D complete remission, partial remission, improvement, or unchanged status. Independent predictors of T2D complete remission were assessed by binary logistic regression and then included in 7 predictive models. Logistic regression assessed the pre- and postoperative predictors of T2D complete remission and their predictive performance was measured with the area under the curve of the receiver operating characteristic curve. Results A total of 230 patients were included in the study, females comprised 69%, and mean age was 45.66 ± 8.84 years. Mean preoperative weight and body mass index were 115.69 ± 20.76 kg and 43.53 ± 6.98 kg/m 2 , respectively. Approximately two thirds (64.4%) of the sample had diabetes for >5 years. Insulin therapy users comprised 36.9% of the sample and 29.6% of patients were on ≥2 oral hypoglycemic agents (OHA). At 1 year, mean body mass index was 32.77 ± 6.09 kg/m 2 , percent excess weight loss (%EWL) was 62.29 ± 23.60% and glycosylated hemoglobin (HbA1C) improved from 8.1% to 6.18%. Approximately 42.2% of the sample achieved T2D complete remission. Compared with those with no remission, patients with complete remission were significantly younger, had shorter duration of diabetes, were not on insulin therapy, took fewer OHA, had higher C-peptide, lower preoperative HbA1C, were less likely to have had hypertension or dyslipidemia, and more likely to have achieved higher %EWL. Seven proposed models for prediction of complete remission showed the most useful model comprised diabetes duration + pre-HbA1C + %EWL + insulin therapy + age + OHA (area under the curve=.81). Independent predictors of complete remission were preoperative HbA1C, %EWL, insulin therapy, age, and OHA (but not diabetes duration). Conclusion SG results in significant weight reduction and various extents of T2D remission. HbA1C, %EWL, insulin therapy, age, and OHA were independent predictors of complete remission. Assessing these factors before bariatric surgery is important to identify any modifiable characteristics that can be altered to increase the likelihood of remission.