Several observational studies have suggested that using bariatric surgery to achieve significant sustained weight loss may be an effective treatment for type 2 diabetes.
Dixon and colleagues conducted a 2-year, randomized, controlled trial involving 60 obese patients recently diagnosed as having type 2 diabetes (<2 years) to compare
the effects of weight loss from laparoscopic adjustable gastric banding (LAGB) with those of weight loss from conventional therapy. The patients, studied at an Australian obesity research center, were between the ages of 20 and 60 years with a body mass index (BMI) >30 and <40 kg/m2 (mean 37.1) and no evidence of renal impairment or diabetic retinopathy. Exclusions included previous bariatric surgery, type 1 diabetes, a recent major vascular event, or other contraindications. Each patient was assessed by a dietitian, a general physician, and an endocrinologist specializing in diabetes. The healthcare team suggested changes intended to maximize current diabetes management during a run-in period of 3 months or more, with the endocrinologist choosing the proper time for randomization.
Patients were randomly assigned to 2 treatment groups (n=30 each group), the conventional therapy group and the bariatric surgery group. The conventional group received best-practice treatment, education, and follow-up, which included open access to a healthcare professional, regular health visits every 6 weeks during the 2-year program, an individually designed lifestyle modification, including diet and exercise programs, as well as pharmaceutical therapy as needed. The surgery group received the same therapy as the conventional group in addition to placement of a LAG band via the pars flaccida technique by 1 of 2 surgeons within 1 month of randomization. Patient progress was reviewed by the surgery team every 4 to 6 weeks throughout the study and band volume adjustments were made as needed based on standard clinical criteria.
Primary endpoint was the proportion of participants achieving remission of their diabetes, defined as fasting plasma glucose levels <126 mg/dL and A1C values <6.2% without use of insulin or oral hypoglycemics. Secondary outcomes included percentage changes in measures of blood pressure, weight, A1C, fasting lipids, and waist circumference.
At the end of the 2-year program, 29 (97%) of the surgery group and 26 (87%) of the conventional group completed the trial. Twenty-two of 29 (76%) in the surgery group and 4 of 26 (15%) in the conventional group achieved diabetes remission. (Including the patients who withdrew prior to completion, 22 of 30 in the surgery group and 4 of 30 in the conventional group achieved remission; P<0.001.) Estimating the most conservative conclusion with the 1 surgical noncompleter not achieving remission and the 4 conventional noncompleters achieving remission, the results (22/30 [73%] and 8/30 [27%], respectively) remained significant (P<0.001).
Greater percentage of weight loss (R2=0.46; P<0.001) and, to a lesser degree, lower baseline A1C (R2=0.50 for both; P<0.001) were the only variables independently associated with remission. Only 4 of 34 patients (12%) who lost <10% of body weight achieved remission after 2 years, and only 4 of 26 (15%) who lost >10% of body weight did not achieve remission. The surgical group achieved a mean body weight loss of 20.0% (SD, 9.4%) after 2 years versus 1.4% (SD, 4.9%) in the conventional group (P<0.001), representing a loss of excess weight equaling 62.5% and 4.3%, respectively (using a BMI of 25 as ideal weight). Mean BMI dropped from 36.9 to 29.5 in the surgical group and 37.1 to 36.6 in the conventional group.
Compared with the conventional therapy group, the surgical group had significant reductions in A1C levels, use of metformin, use of other hypoglycemic agents (P=0.006), and use of antihypertensive drugs (P=0.005), as well as significantly greater improvements in insulin sensitivity, triglyceride levels (P=0.02), HDL-C levels, and waist circumference (P<0.001 for each measurement unless otherwise noted). The authors note that changes in secondary outcomes measures are for hypothesis-generating only because the study was not powered for a full examination of these measures. Adverse events were minor in both groups.
The study had several limitations. First, restricting the study to patients with recently (<2 years) diagnosed type 2 diabetes limits the application of these results to patients with a longer duration of diabetes. Second, the LAGB surgical team in this study was highly experienced in the technique; studies have shown an inverse relationship between the experience of LAGB surgeons and the occurrence of postoperative patient complications. Third, the study was not designed to measure such outcomes as cardiovascular events or mortality and follow-up was limited to 2 years, which limited the longterm extrapolations of the results. The authors concluded that greater weight loss from LAGB resulted in diabetes remission and other metabolic improvements in the majority of obese patients with recently diagnosed type 2 diabetes. They suggested the need for larger, longer studies to better determine long-term efficacy.
Dixon JB et al. Adjustable gastric banding and conventional therapy for type 2 diabetes: a randomized. controlled trial. JAMA. 2008;299(3):316-323.