Compiled by Christina Frangou

Evidence has firmly established that bariatric surgery quickly and dramatically improves type 2 diabetes mellitus, effective even before patients lose significant amounts of weight. A meta-analysis published this year found that the clinical and laboratory manifestations of type 2 diabetes resolved or improved in the great majority of patients after bariatric surgery, and these responses were more pronounced for procedures associated with a greater percentage of excess body weight loss and were maintained for two years or longer (Am J Med 2009;122:248-256.e5).

But why and how bariatric procedures affect diabetes is the subject of scientific debate. Experts weigh in on the discussion.

Walter J. Pories, MD
Professor of Surgery and Biochemistry
Chief, Metabolic Institute
East Carolina University School of Medicine
Greenville, North Carolina

This effect is due to both factors: by weight loss and by changes in signaling systems from the gut. We know that it’s due to weight loss in part because the adjustable gastric band, an operation that owes its effects solely to the restriction of [food] intake, produces remission of diabetes in 50% to 60%. Weight loss is also one of the major tools that internists use in treating diabetes.

We have to be careful in our considerations of the relationship of obesity and diabetes. Of the patients who have diabetes, 10% are actually lean individuals. Further, in our obese populations, only one-third have diabetes. If obesity and diabetes were direct cause and effect, neither of those observations would be true. On the other hand, operations that exclude a portion of the intestine from contact with food produce significantly higher rates of remission (i.e., 83% with the gastric bypass to 99% in biliopancreatic bypass). Clearly, there is a reason for that higher success rate. One explanation is that the foregut in diabetic patients responds abnormally to food. That response, mediated either by endocrine or neural signals, causes an excessive secretion of insulin and, secondarily, insulin resistance. The signal could also affect muscle directly and cause insulin resistance by interfering with insulin action, most likely in the mitochondria. There’s a major effort now around the world to determine if there are such signals, to define these molecules and, most important, to find ways to interrupt those signals. With luck, if we can dissect those signaling systems, we may be able to get remission of diabetes without the surgery. We need a nonoperative answer. After all, you can’t operate on the whole world. In the meantime, though, we need to increase access to the surgery; it is the best treatment we have at this time.

The most exciting finding is that the remissions produced by surgery are not temporary. We have followed patients who have remained in remission for one to two decades who also remain free of the complications of diabetes. I have a patient I operated on 23 years ago who is still off insulin and has both legs, both eyes, both kidneys. That’s pretty amazing, especially if we recall that bariatric surgery does not produce lean individuals. Almost all remain overweight, obese, an observation that also throws doubt on the cause-and-effect relationship between weight and diabetes.

The recent success of Byetta [Amylin/Lilly] (exenatide), a GLP-1 [glucagon-like peptide-1] analogue, provides additional support for the hypothesis that the gut plays a role in the pathogenesis of diabetes.

John Dixon, MBBS, PhD
Associate Professor
Head of the Obesity Research Unit
Monash University
Senior Clinical Scientist
Baker IDI Heart and Diabetes Institute
Melbourne, Australia

Diabetes is so strongly associated with weight and weight gain that the most impressive and overwhelming reason that diabetes improves is weight loss. There’s just no debate. That is an overwhelming effect. The question is: Is there an additional effect with some operations? It appears that there certainly is an additional effect with diversionary operations. Those operations produce a more immediate effect, an effect that’s faster than what one would expect for the weight loss. However, we don’t at this stage know the basis or the underlying mechanism of that effect and we don’t know, in fact, whether it’s having a fundamental underlying effect on the pathogenesis of diabetes. Therefore, we’re not sure how beneficial it is in the long run. So is it a 60/40, an 80/20 or a 90/10 relationship? We don’t know, but we do know that if we don’t see the weight loss, we don’t see the remarkable long-term improvement.

I take issue with the [use of the] phrase “resolution of diabetes.” Resolution is something we have not defined. All we’re really looking at with bariatric surgery and diabetes is improvement in glycemic control, so terms such as resolution or “cure” really can’t be entertained. Endocrinologists and diabetologists would find such terms almost offensive in the sense that no surgery to date has been able to show any resolution or cure of diabetes. What we’re talking about is excellent glycemic control or “remission.” Remission is reasonable because if there is weight regain, diabetes will reemerge. We see it with all surgery, restrictive or satiating procedures, gastric bypass and even biliopancreatic diversion. Diabetes increases in likelihood as we get older anyway, so what we’re really doing with surgery is excellently treating diabetes, even putting it into remission for many years. How many years we don’t know, but the best data come from the Swedish Obese Subjects study, which suggests that the surgery is incredibly useful to delay the onset of diabetes or put diabetes back in its box for 10 or 15 years.

The other point that I would like to make is if there is an effect outside of weight loss, then we should be able to demonstrate this powerful effect in people who don’t have a weight problem. When they’ve tried this surgery as anti-diabetes procedures in people with BMIs [body mass indexes] of 25 to 30 kg/m², in other words major, major surgery in people who are overweight and don’t have obesity, they have not seen the stunning results we see with weight loss or gastric bypass in obese people. This is putting a cautionary note on how much of this is beyond weight loss. Some people like to dismiss weight loss and get on to the other effects of bariatric surgery, but we must put them together and understand that they work together.

Robert Cooney, MD
Professor of Surgery, Cellular and Molecular Physiology
Chief, Division of General Surgery
Penn State M.S. Hershey Medical Center
Hershey, Pennsylvania

Although weight loss and exercise certainly improve diabetes, I’d like to focus my comments on how bariatric procedures that alter gastrointestinal anatomy, like gastric bypass, improve glucose homeostasis. Bariatric surgeons have known for years that glucose control is dramatically improved following gastric bypass surgery, in most cases before significant weight loss has occurred. Decreased food intake, weight loss, changes in adipokine production, foregut bypass and enhanced exposure of the ileum to undigested food have all been identified as potential mechanisms for improved glucose control after gastrointestinal surgery. Type 2 diabetes mellitus (T2DM) is a heterogeneous disease caused by the combination of insulin resistance resulting in decreased glucose uptake in muscle, and pancreatic β-cell dysfunction, eventually causing loss of β-cells. At present, we do not completely understand how gastrointestinal procedures improve T2DM. However, clinical and experimental studies suggest surgical procedures that bypass the foregut and/or enhance ileal exposure to nutrients (e.g., gastric bypass, biliopancreatic diversion) improve glucose tolerance by altering intestinal metabolism and endocrine function. Endocrine cells lining the intestine secrete peptide hormones in response to luminal nutrients that regulate appetite and energy metabolism. The incretins, glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1), are gut peptides that regulate glucose metabolism by improving insulin sensitivity and pancreatic insulin secretion. Patients with T2DM demonstrate resistance to GIP and decreased circulating levels of GLP-1. Within weeks of gastric bypass surgery, postprandial GLP-1 secretion is increased and insulin sensitivity is improved. These changes are not caused by weight loss alone and are not seen in experimental studies which control for post–gastric bypass reductions in food intake by pair-feeding a sham surgery group. Although postoperative changes in incretins are posited as an important mechanism for improvements in glucose control after gastric bypass, other gut hormones that regulate appetite and satiety (including ghrelin, apolipoprotein AIV and peptide YY) are altered as well, and likely contributory.

Philip D. Schauer, MD
Director of the Bariatric and Metabolic Institute
Cleveland Clinic
Cleveland, Ohio

It is clearly both. We know that weight loss plays an important role. We know also that it is not just about weight loss because gastric bypass surgery causes a remarkable improvement in type 2 diabetes mellitus that is independent of weight loss. There is a normalization of glucose metabolism that occurs long before substantial weight loss takes place. The rapid time and degree of improvement after gastric bypass compared with equivalent weight loss from other interventions suggest there is a surgery-specific, weight-independent effect. We know gastric banding primarily improves diabetes only by weight loss. The gastric bypass and other bypass-type procedures have an enhanced effect that is caused by an elaboration of gut hormones that actually stimulate insulin secretion. So, it looks like thus far—and the jury’s still out on this one—the bypass procedures have an additional effect that’s additive to the weight loss effect that we see with weight loss operations.

Francesco Rubino, MD
Chief of Gastrointestinal Metabolic Surgery
Assistant Professor of Surgery
Weill Cornell Medical College
New York, New York

There is now enough evidence that weight loss alone cannot explain the effect of surgery on diabetes. We need to revise completely the way we look at obesity and weight loss.

Much of the scientific community and the society at large still tend to neglect the modern knowledge on gastrointestinal (GI) physiology, and think that obese people can lose weight by just a voluntary decision to eat less and exercise more. We might be making it too simplistic by thinking that way. Body weight and glucose homeostasis (and the related diseases of obesity and diabetes) are regulated by an incredibly powerful biological mechanism that is finely balanced. This mechanism might not be so simple to override when it is dysfunctional.

In my opinion, there also is a widespread misunderstanding that weight loss itself induces diabetes improvement. That may very well be wrong. In fact, weight loss is not an intervention. It is not an operation nor a drug. Weight loss is a consequence of another intervention, for instance, of diet, intense physical exercise, drugs or surgery. So maybe it is not the weight loss itself that is improving diabetes after diet or surgery, but weight loss and diabetes improvement could be two consequences of the intervention’s effect on metabolic mechanisms. Studies that show that diabetes improves after weight loss from diet and intensive lifestyle changes are usually interpreted as evidence that weight loss per se improves diabetes. In fact, what these studies really show is only that diet and lifestyle changes can, at the same time, improve diabetes and cause weight loss; they do not provide scientific proof that there is a cause–effect relationship between these two outcomes. I know that most of my colleagues will raise their eyebrows when I say this, but if we want science to make inroads into metabolic surgery, we should use scientifically rigorous thinking and accurate methods of investigations even when this means challenging a commonly held belief.

Any clinical study where one uses interventions that influence energy homeostatic mechanisms (like diet, or surgery or drugs) is likely to produce both weight loss and improvement of diabetes. For instance, the findings from the recent meta-analysis by Buchwald and co-workers showing that opeartions that have the most effect on diabetes also are the ones that induce more weight loss do not necessarily mean that diabetes improves because of weight loss; in fact, it is completely plausible that the operations that have greater influence on neuroendocrine mechanisms of energy homeostasis yield more profound effects on both body weight and diabetes. You need mechanistic studies to answer the question of how surgery improves diabetes.

In the past, the GI tract was considered as just a conduit for nutrients; accordingly, restriction and malabsorption were the only two mechanisms that one could think of to explain weight loss after bariatric procedures. Modern science has shown that the GI tract has a powerful physiologic role in regulation of appetite/satiety and insulin secretion/sensitivity. Therefore, changing the anatomy of the digestive tract may change the way GI signals are produced and, as a consequence, induce both diabetes improvement and body weight loss. Also, different sites of the GI tract have distinct physiologic effects on energy homeostasis; it is therefore very likely that the hormonal consequences of a gastric resection are different from those of an intestinal bypass, and that the different lengths of intestinal limbs in different operations may have distinct hormonal effects. This may explain why degree of diabetes control, rapidity of diabetes remission and recurrence rates of disease can vary with the type of surgical procedures.