The purpose of this study is to quantify the relative contribution of improved insulin sensitivity and improved insulin secretion to the glycaemic effect of bariatric surgery.

In 1987 Pories et al. (1) published that after gastric bypass 99% of morbidly obese patients became and remained euglycaemic over time. However, the most interesting observation of these authors (1) was that the patients were converted to euglycaemia within 10 days of the operation, even if before they required large doses of insulin. Thereafter, we ourselves (2-6) as well as other Authors (7,8) have found that both mainly malabsorptive bariatric and mainly restrictive surgery are effective in producing early improvement/resolution of type 2 diabetes, also in normal weight (9) or overweight-obese subjects (10).

Until now the mechanisms through which bariatric surgery operate in improving insulin sensitivity and, consequently, diabetes control have not been elucidated. To complicate the picture there is the fact that after bariatric surgery patients are not allowed to eat for a time that is longer as the entity of the abdominal trauma increases. In fact, after laparoscopic bariatric surgery the patients are re-alimented in the 3rd post-operative day while after open surgery the time length before re-alimentation is about 7 days. Although in most centres the patients receive a parenteral nutrition covering the energy needs while they remain at rest, in some other centres they are only re-hydrated. Thus, it might be hypothesised that fasting can contribute to the improvement of glucose disposal. The data in the literature counteract this possibility as many Authors (11-15) have demonstrated that a very-low calorie diet (VLCD) is essentially unable to improve insulin sensitivity at least in the short run, when the weight loss is minimal. In this regard, Jazet et al. (11) found that the glucose control amelioration in diabetic subjects undergoing a 2-day VLCD was a reduction in the hepatic glucose production rather than an improvement in the whole-body glucose disposal. Similarly, Christiansen et al (12) observed the increase in the glucose metabolic clearance rate (MCR) was not apparent before day 20 of energy restriction, when the weight loss became significant. In another study, diabetic patients under a 4-day energy-restricted diet (providing 4620 ± 1050 kJ/d) had a deterioration of both basal glucose MCR and insulin-stimulated glucose disposal (13).

Recently, Laferrère et al (16) compared the late (1 month) effect of an equivalent weight loss by RYGB or by diet in two groups of matched morbidly obese patients with type 2 diabetes mellitus. The glucose area under-the-curve (AUC) after an oral glucose tolerance test (OGTT) and glucose levels at 120' were significantly lower after RYGB compared with diet (P = 0.014 and P = 0.001, respectively).

Another important aspect of the weight loss is the modulation of lipolysis. Many studies have shown that obese subjects are insulin resistant at the level of adipose tissue, ie, despite high plasma insulin levels lipolysis and FFA release are not suppressed. It is therefore interesting to study the changes in lipolysis before and after surgery and also the ability of insulin (e.g., during the euglycemic, hyperinsulinemic clamp) to suppress peripheral lipolysis. By infusing deuterated glycerol we will be able to estimate glycerol rate of appearance, which is the best available index of lipolysis. In fact intracellular triglyceride (TG) hydrolysis results in the release of one mole of glycerol and 3 moles of FFA. While the FFA can be re-esterified intracellularly, and thus do not appear into the systemic circulation, glycerol cannot be re-esterified since the adipocyte lacks glycerol kinase. Thus, all glycerol derived from intracellular TG hydrolysis is released into the circulation and can be traced by deuterated glycerol. We have used this technique previously in diabetic subjects (17), in subjects with NASH as well as in obese subjects under VLCD (18). In the latter study we have shown that VLCD improved adipose tissue IR.

As far as we know, no data are available in the literature comparing the effect of RYGB with a purely restrictive bariatric procedure in order to take into account simultaneously the effect of the surgical trauma and that of the reduced food intake in the early post-operative period.

The main goal of this study is to determine the immediate effect of Roux-en-Y gastric bypass surgery on glucose homeostasis, factoring out the confounding effects of the surgical approach (incl. laparoscopy) and of the post-operative fasting.

To this purpose, patients of the two bariatric surgery groups (Roux-en-Y gastric bypass and gastric banding) are matched as far as feasible. Primary parameter is insulin resistance (as assessed by Matsuda index). Further the same BMI inclusion criteria apply for both groups (BMI > 30 kg/m2 and < 55 kg/m2). While age, gender and BMI within the specified range are also expected to influence the read-outs, no matching for these baseline characteristics is foreseen due to feasibility issues.

Patients receive oral food typically from day 3 or 4 after laparoscopic bariatric surgery. This post-operative food intake is restricted to 800 - 1000 kcal/day, usually divided in 6 small meals, with the following composition 55% carbohydrates, 30% lipids and 15% proteins; saturated lipids ~ 10% . A similar caloric intake restriction will be imposed during the 1-week diet period prior to Visit 3, which will be done 10 +/- 2 weeks before the operation in order to allow the patients a complete recovery of his/her body weight which must be kept constant before the next study in order to avoid methodological biases.

Inclusion Criteria:

• Morbidly obese subjects of both sexes with a BMI >40 kg/m2 who, for their obesity disease, are eligible for bariatric surgery and have accepted to undergo RYGB or LAGB
• Confirmed insulin resistance as assessed by Matsuda index
• Age 25-55 years
• Weight stable for at least 6 months before the study (+/- 5 kg within the previous 6 months)
• Stable medication
• Provision of informed consent, statistical analysis, and publications of obtained results

Exclusion Criteria:

• Patients not eligible for laparoscopic RYGB or LAGB

• Incapacity to give a valid informed consent or unwilling to give the consent

• Pregnancy or lactating

• Patients eligible for RYGB/LAGB, but with:

  • Type 2-diabetes mellitus
  • Significant illness within the two weeks preceding surgery, as judged by the physician.
  • Obvious infection (bacteria, virus etc)
  • Major cardiovascular disease

Twenty morbidly obese subjects scheduled for RYGB or LAGB for will be included in the study.