Human subjects with severe obesity are increasingly treated with bariatric surgery to promote weight loss via procedures that reduce the capacity of the stomach and/or the absorptive surface area of the small bowel, resulting in reduced food intake and/or energy malabsorption. Some of these surgical procedures may also be associated with changes in plasma levels of one or more gut hormones, due to anatomical alterations in gut motility, incomplete nutrient digestion, and disruption of neural innervation. Changes in the levels of circulating gut hormones have been commonly observed after gut surgery, and in some instances, relative changes in the numbers of specific enteroendocrine cell subsets have also been described. Diversion of nutrients away from the proximal gut and consequent exposure of the distal gut to a greater load of incompletely digested nutrients is often associated with a reductions in levels of circulating peptides derived from the proximal gut, and an increase in levels of peptide hormones derived from the distal gut, such as neurotensin, PYY and enteroglucagon (a surrogate for levels of glicentin, oxyntomodulin, GLP-1 and GLP-2) Gut hormone changes after jejunoileal (JIB) or biliopancreatic (BPB) bypass surgery for morbid obesity. Int J Obes 1981; 5:471-80. and Plasma enteroglucagon after jejunoileal bypass with 3:1 or 1:3 jejunoileal ratio. Scand J Gastroenterol 1979; 14:205-7. and Morphological and functional alterations to a sub-group of regulatory peptides in human pancreas and intestine after jejuno-ileal bypass. Int J Obes Relat Metab Disord 1993 17:109-113.
As many human subjects experience significant weight loss, and even more remarkable improvement or complete resolution of their diabetes within days of the surgical procedure, there is great interest in understanding the potential roles of gut hormones in the improvement of b-cell function, the amelioration of the diabetic state, and in the factors contributing to weight loss. An increase in the levels of anorectic hormones such as PYY or GLP-1, or a decrease in levels of orexigenic hormones such as ghrelin Plasma ghrelin levels after diet-induced weight loss or gastric bypass surgery. N Engl J Med 2002: 346:1623-1630 are associated with changes in appetite and body weight in subsets of patients following bypass surgery. Indeed, in some cross-sectional studies, a restored GLP-1 response is a key variable redictive of diabetes remission Roux-en-Y Gastric Bypass and Sleeve Gastrectomy: Mechanisms of Diabetes Remission and Role of Gut Hormones J Clin Endocrinol Metab. 2013 Nov;98(11):4391-9. As outlined below, ascertaining the precise individual contributions of specific hormones and metabolites in the glucoregulatory and weight loss responses after bariatric surgery is more challenging.
Gastric bypass surgery, GLP-1, weight loss, and improvement in glycemia
Many patients with obesity experience rapid weight loss together with striking amelioration of their diabetes often within days of gastric bypass surgery. As marked amelioration or complete resolution of the associated diabetes frequently precedes weight loss, a role for GI hormones secreted from the distal GI tract, such as GLP-1 has been invoked to explain these impressive clinical improvements. As different surgical procedures result in distinct anatomical rearrangement of the normal proximal-distal gut orientation and integrity, it is important to distinguish between the various surgical procures and associated changes in levels of gut hormones. Much of the published data correlates changes in gut hormone levels with improvements in glycemia and weight loss after GBS. Some studies use transient administration of antagonists to probe the role of specific gut hormones. It seems clear that a subset of patients develop hyperinsulinemic hypoglycemia secondary to a brisk exaggerated rise in GLP-1 secretion and in many patients, increased GLP-1 levels contribute to improved insulin secretion after meal ingestion post bariatric surgery. It is less clear how much GLP-1 contributes to the overall improvement in fasting glycemia, improvement in insulin sensitivity, complete resolution of diabetes, and weight loss in the majority of subjects undergoing bariatric surgery.
Jimenez and colleagues examined the importance of endogenous GLP-1 for glucose control and insulin secretion by studying 8 female subjects (mean age 54) who had experienced complete remission of their diabetes; patients were studied at least 2 years after their GBS. Subjects were studied on two different days with or without the GLP-1R antagonist exendin(9-39) administered as an iv bolus followed by a continuous infusion prior to and following a standardized liquid test meal. Glucose levels rose more rapidly and peaked at higher levels in the post GBS subjects. Ex(9-39) produced a small but significant increase in glycemic excursions in control subjects from 30-70 minutes. In contrast, the increased glycemic profile in post GBS patients was comparatively delayed at ~ 80-120 minutes after Ex(9-39). Notably, Ex(9-39) did not shift the glycemic profile of the GBS patients into the diabetic range. Ex(9-39) had no significant effect on fasting insulin/C-peptide levels in control or post GBS groups. Plasma GLP-1, C-peptide and insulin responses were significantly greater post test meal in the GBS groups and Ex(9-39) significantly decreased insulin and C-peptide excursions to a much greater extent in the GBS group compared to control, without compaable large increases in plasma glucose levels. Glucagon levels were also higher post GBS, and rose modestly but were not significantly different after Ex(9-39). GIP levels were also much greater post GBS and not different after Ex(9-39). The authors conclude that elevated GLP-1 levels post GBS do not contribute significantly to the major improvements in glycemic control in the majority of subjects. GLP-1 Action and Glucose Tolerance in Subjects With Remission of Type 2 Diabetes Mellitus After Gastric Bypass Surgery Diabetes Care. 2013 Jan 28
In contrast, Jorgensen and colleagues studied 9 patients with type 2 diabetes, mean age 50, BMI 39, duration of diabetes ~ 5 yrs, before and 1 week and 3 months after RYGB surgery. Acute administration of exendin(9-39) reversed the improvements in b-cell glucose sensitivity and glucose tolerance, and further increased already elevated levels of glucagon, leading the authors to conclude that the ~8-fold increase in GLP-1 levels was very important for the obsrved improvements in b-cell function and reduction of glycemia. Exaggerated glucagon-like peptide 1 response is important for improved β-cell function and glucose tolerance after Roux-en-Y gastric bypass in patients with type 2 diabetes Diabetes. 2013 Sep;62(9):3044-52
Similarly, Jorgensen and colleagues used exendin(9-39) infusion in 9 patients at 1 week and 3 months after RYGB to implicate a role for GLP-1R signaling in changes in insulin, glucagon, and beta cell glucose sensing after bypass surgery. β-GS decreased to preoperative levels, glucagon secretion increased, and glucose tolerance was impaired by Ex-9 infusion. Exaggerated glucagon-like peptide 1 response is important for improved β-cell function and glucose tolerance after Roux-en-Y gastric bypass in patients with type 2 diabetes Diabetes. 2013 Sep;62(9):3044-52. In contrast, similar studies by different investigators in 8 subjects with sustained remission of diabetes after RYGP surgery demonstrated detectable but modest effects of exendin(9-39) on parameters of glucose tolerance, leading these investigators to conclude that " The limited deterioration of glucose tolerance on blockade of GLP-1 action in our study suggests the resolution of T2DM after RYGBP may be explained by mechanisms beyond enhancement of GLP-1 action" GLP-1 action and glucose tolerance in subjects with remission of type 2 diabetes after gastric bypass surgery Diabetes Care. 2013 Jul;36(7):2062-9.
Shah and colleagues infused exendin(9-39) into human subjects after gastric bypass to probe the role of GLP-1R signaling. Although exendin(9-39) did increase glucose and reduce insulin levels, exendin(9-39) did not alter meal appearance, suppression of glucose production or stimulation of glucose disappearance after RYGB subjects. Hence, GLP-1 may contribute to but does not explain the complex metabolic improvements observed after RYGB. Contribution of endogenous glucagon-like peptide 1 to glucose metabolism after Roux-en-Y gastric bypass Diabetes. 2014 Feb;63(2):483-93
Naslund and colleagues assessed plasma levels of GIP and GLP-1 from 9 months to 20 years after jejunoileal bypass surgery. Both fasting and postprandial levels of GIP were markedly elevated in female obese human subjects 20 years after bypass surgery, in association with improvement in glucose tolerance as described in Importance of small bowel peptides for the improved glucose metabolism 20 years after jejunoileal bypass for obesity. Obes Surg 1998 8:253-260.
Valverde and colleagues studied plasma levels of GLP-1, together with serial analysis of glucose tolerance in two groups of patients; after Larrad's pancreaticobiliary diversion (BPD) or following vertical banded gastroplasty (VBG). Basal and glucose-stimulated plasma GLP-1 increased after surgery, with GLP-1 levels comparatively greater in subjects following BPD. See Changes in glucagon-like peptide-1 (GLP-1) secretion after biliopancreatic diversion or vertical banded gastroplasty in obese subjects. Obes Surg. 2005 Mar;15(3):387-97. Similarly, the GLP-1 response to meal ingestion was flat or absent prior to surgery for obesity, but significantly improved in obese subjects 6 weeks following Roux-en-Y Gastric Bypass (RYGBP) GLP-1, PYY, Hunger and Satiety Following Gastric Bypass Surgery In Morbidly Obese Subjects. J Clin Endocrinol Metab. 2006 Feb 14; [Epub ahead of print]
Le Roux and colleagues examined changes in plasma levels of gut hormones and glucose tolerance following different gastric bypass procedures in rats and human subjects. Patients with Roux-en-Y gastric bypass (RYGB) had increased postprandial levels of plasma PYY and GLP-1 together with early and exaggerated insulin responses, and improved glycemic control. In contrast, these hormonal changes were not seen in subjects after gastric banding.Gut hormone profiles following bariatric surgery favor an anorectic state, facilitate weight loss, and improve metabolic parameters. Ann Surg. 2006 Jan;243(1):108-14.
Sequential changes in glucose tolerance, gastric emptying and levels of gut hormones in response to a liquid test meal were also assessed by Falken and colleagues in 12 obese subjects at 3 days, 2 months, and one year after gastric bypass. Change in body mass was highly significant (from 45 to 30 over one year). Plasma levels of several gut hormones, including GLP-1 increased progressively over time, glucose tolerance and b-cell function improved, coincident with satiety and weight loss. However, significant improvements in plasma levels of GLP-1 were also noted as early as day 3. Transit through the GI tract was rapidly improved as early as day 3. Changes in Glucose Homeostasis after Roux-en-Y Gastric Bypass Surgery for Obesity at Day Three, Two Months, and One Year after Surgery: Role of Gut Peptides. J Clin Endocrinol Metab. 2011 May 4. [Epub ahead of print]
Guidone and colleagues studied glucose tolerance, b-cell function, insulin sensitivity, and plasma levels of beta-cell function was observed even 1 week after surgery, prior to the development of significant weight loss. Plasma levels of GIP fell, whereas levels of GLP-1 increased, in association with complete resolution of the diabetes. Whether these changes in gut hormones were simply associated with or contributed to improvement in the diabetic state remains unclear. See Mechanisms of recovery from type 2 diabetes after malabsorptive bariatric surgery. Diabetes. 2006 Jul;55(7):2025-31. Indeed Dutia and colleagues studied 16 severely obese patients with type 2 diabetes, up to 3 years post RYGB, 11 severely obese normal glucose tolerant controls and 7 lean controls. Weight loss was substantial, ~31% by year 1, and sustained after 3 years of F/U. Although the incretin effect and insulin secretion in response to oral glucose improved after RYGB, in association with increased levels of GLP-1, assessment of b-cell function with intraveous glucose infusion revealed limited improvement in insulin secretion and b-cell function, desite complete diabetes remission after bariatric surgery. Hence, although increased GLP-1 levels after meals clearly contributes to meal-related improvements in insulin secretion and glycemia after GBS, the lack of improvement of b-cell function suggests that other mechanisms, including weight loss and improved insulin sensitivity, likely contribute to diabetes remission. Limited recovery of β-cell function after gastric bypass despite clinical diabetes remission Diabetes. 2013 Dec 2.
Gastric bypass surgery and postprandial hypoglycemia
Six patients with hyperinsulinemic hypoglycemia following meal ingestion were detected 0.5-8 years following Roux-en-Y gastric bypass surgery were found to have histological evidence for nesidioblastosis following resection of pancreatic tissue-one patient was found to have multiple insulinomas as described in the July 21 2005 New England Journal of Medicine Hyperinsulinemic hypoglycemia with nesidioblastosis after gastric-bypass surgery. N Engl J Med. 2005 Jul 21;353(3):249-54. The authors speculated, as further discussed in an accompanying editorial, that excessive secretion of gut hormones such as GLP-1 may have contributed to the development of islet proliferation in these human subjects. Indeed, many of the patients presenting with hypoglycemia post bypass have increased GLP-1 levels and enhanced beta cell sensitivity to GLP-1 action Exaggerated release and preserved insulinotropic action of glucagon-like peptide-1 underlie insulin hypersecretion in glucose-tolerant individuals after Roux-en-Y gastric bypass Diabetologia. 2013 Dec;56(12):2679-87
A similar clinical picture was reported in 3 subjects following gastric bypass, and plasma levels of GLP-1 were markedly elevated in these studies Severe hypoglycaemia post-gastric bypass requiring partial pancreatectomy: evidence for inappropriate insulin secretion and pancreatic islet hyperplasia. Diabetologia. 2005 Sep 30; [Epub ahead of print]. However, a subsequent re-analysis of the pancreas histology, together with control slides obtained from 31 obese subjects and 16 lean control subjects, yielded somewhat modified conclusions. Meier and colleagues reported that b-cell area was not increased in the subjects with gastric bypass-associated hypoglycemia and no evidence of increased islet neogenesis or b-cell proliferation was detected in this group. These findings further emphasis the importance of functional defects, namely changes in gut motility and the acute b-cell response to nutrients/gut hormones in the pathogenesis of the hyperinsulinemic hypoglycemia syndrome seen in some subjects. See Hyperinsulinemic hypoglycemia after gastric bypass surgery is not accompanied by islet hyperplasia or increased beta-cell turnover. Diabetes Care. 2006 Jul;29(7):1554-9. Indeed, many of the patients presenting with hypoglycemia have increased GLP-1 levels and enhanced beta cell sensitivity to GLP-1 action Exaggerated release and preserved insulinotropic action of glucagon-like peptide-1 underlie insulin hypersecretion in glucose-tolerant individuals after Roux-en-Y gastric bypass Diabetologia. 2013 Dec;56(12):2679-87
Salehi and colleagues examined the importance of GLP-1/the GLP-1R in glucose-stimulated insulin secretion in asymptomatic subjects after gastric bypass (GB) and in subjects with recurrent hypoglycemia using acute administration of the GLP-1R antagonist, exendin(9-39). Fasting blood glucose and insulin levels were comparable in the two groups. The insulin response to IV glucose and insulin sensitivity was also similar between groups. Administration of Ex9 produced a comparatively greater suppression of postprandial insulin levels in the GB groups however the rate of gastric emptying was not modified by Ex9. The postmeal rise in plasma GLP-1 was significantly higher in the patients experiencing hypoglycemia. Intriguingly, Ex9 infusion produced a further rise in plasma GLP-1 levels, consistent with a negative feedback loop. The control of glucagon secretion was also abnormal and increased in GB patients fater meal ingestion. Whether increased levels of GLP-1 are the predominant factor producing hypoglycemia in these subjects remains uncertain. Gastric Bypass Surgery Enhances Glucagon-Like Peptide 1-Stimulated Postprandial Insulin Secretion in Humans Diabetes. 2011 Sep;60(9):2308-2314
In a subsequent study, Salehi studied 24 subjects, 9 (2 with prior diabetes) with recurrent hypoglycemia (recurrent episodes with symptoms of neuroglycopenia) and 7 individuals (2 with prior diabetes) without hypoglycemia after gastric bypass, and 8 control subjects who did not undergo previous bariatric surgery. The magnitude of peak GLP-1 levels correlated inversely with the glucose in hypoglycemic subjects. Complaints compatible with dumping syndrome were common in subjects after GBS. Total weight loss and time since surgery were comparable across groups. Subjects were studied with a mixed meal, tracer infusion, with and without an intravenous infusion of the GLP-1R antagonist Ex9. Blockade of the GLP-1R with Ex9 increased both fasting and postprandial glucose levels in all three groups, however the increase in glucose was greater in subjects with hypoglycemia and infusion of Ex9 largely prevented the previously demonstrated meal-related hypoglycemia in this cohort. Insulin responses were more rapid and greater in the hypoglycemia group, and Ex9 attenuated in the insulin responses to meal ingestion in all groups. Glucagon levels were higher after meals in the GBS groups and increased further after Ex9. The authors conclude that accelerated glucose absorption together with increased GLP-1 secretion substantially contribute to hypoglycemia after bariatric surgery in susceptible individuals Blockade of Glucagon-like Peptide 1 Receptor Corrects Post-prandial Hypoglycemia After Gastric Bypass Gastroenterology. 2013 Dec 3. doi:pii: S0016-5085(13)01725-3
Salehi and colleagues also noted that plasma levels of GLP-1 were not greater in subjects presenting with severe hypoglycemia after GBS, and insulin clearance was delayed, hence perhaps enhanced beta cell sensitivity to GLP-1 together with other delayed insulin action and failure to increase glucagon in the setting of hypoglycemia (levels of plasma glucagon are elevated and often rise briskly after meal ingeston in these subjects) may partially explain the hypoglycemia in some patients. Altered Islet Function and Insulin Clearance Cause Hyperinsulinemia in Gastric Bypass Patients With Symptoms of Postprandial Hypoglycemia J Clin Endocrinol Metab. 2014 Mar 10:jc2013268
How important is GLP-1 for the multiple benefical effects of bariatric surgery in mice? Remarkably, although GLP-1 levels rise in all preclinical models of GBS, the available evidence from multiple labs is quite consistent, demonstrating that improvement in glycemia and reduction in body weight in these experiments does not require a functional GLP-1R.
Wilson-Perez and colleagues examined the extent of weight loss and improvement in glucose homeostasis arising in two different models of GLP-1R knockout mice subjected to vertical sleeve gastrectomy. Weight loss, body composition, food preferences, and improvements in glucose tolerance were comparable in all groups of mice after VSG independent of the presence or absence of the GLP-1R. Vertical sleeve gastrectomy is effective in two genetic mouse models of glucagon-like Peptide 1 receptor deficiency Diabetes. 2013 Jul;62(7):2380-5.
Mokadem and colleagues assessed the importance of GLP-1 and the GLP-1R using two different knockout mouse lines, Glp1r-/- mice, with disruption of the classical GLP-1R, and a-gustducin-/- mice, which exhibit defective secretion of GLP-1 and the intestinal proglucagon-derived peptides, including oxyntomodulin. Hence, Glp1r-/- mice provide a model for inactivation of the known GLP-1R signaling pathway, whereas a-gustducin-/- mice do not increase the secretion of GLP-1, oxyntomodulin, and metabolites such as GLP-1(9-36) and GLP-1(28-36) after nutrient ingestion or bariatric surgery. Remarkably, RYGB bariatric surgery was similarly effective in producing weight loss, enhanced insulin secretion, and improved glycemia, independent of whether GLP-1 levels were elevated, or the GLP-1R signaling pathway was disrupted. These findings provide further evidence that multiple overlapping signals, beyond a single hormone like GLP-1, contribute to weight loss and improved glucose homeostasis after bariatric surgery. Effects of Roux-en-Y gastric bypass on energy and glucose homeostasis are preserved in two mouse models of functional glucagon-like peptide-1 deficiency Molecular Metabolism, 2013 Dec 11;3(2):191-201.
Similarly, Ye and colleagues asked whether the outcomes of bariatric surgery, with a focus on weight loss, required the GLP-1R signaling pathways. Central GLP-1R signaling pathways are preserved in rats after RYGB, and high fat fed obese Glp1r-/- mice lost the samount amount of weight after RYGB surgery as controls, further suggesting that GLP-1 receptor signaling is not required for reduced body weight after RYGB in rodents GLP-1 receptor signaling is not required for reduced body weight after RYGB in rodents AmJ Physiol Regul Integr Comp Physiol. 2014 Mar;306(5):R352-62
Consistent with these findings, Salinari et al studied the correlation between plasma levels of GLP-1 and outcomes in both Wistar and GK rats. Incretin levels were not increased but insulin sensitivity increased markedly after experimental bypass surgery Duodenal-jejunal bypass and jejunectomy improve insulin sensitivity in goto-kakizaki diabetic rats without changes in incretins or insulin secretion. Diabetes. 2014 Mar;63(3):1069-78.
Does GBS restore the sensitivity of central neuronal circuits that contribuye to control of appetite and body weight. After 12 weeks of high fat feeding rat DMV neurones were also less responsive to cholecystokinin and glucagon-like peptide 1. Roux-en-Y gastric bypass reversed all of these DIO-induced effects and restored neuronal sensitivity to the neuropeptides. Roux-en-Y gastric bypass reverses the effects of diet-induced obesity to inhibit the responsiveness of central vagal motoneurones J Physiol. 2013 May 1;591(Pt 9):2357-72.
Can GLP-1 responsivity predict the results of bariatric surgery? Habegger and colleagues assessed the weight loss and glucoregulatory responses to 4 consecutive days of exendin-4 administration, 50 ug/kg/d, with treatment of male Long Evans rats after 8 weeks of high fat feeding. Groups of rats were stratified according to their exendin-4 responses, first by body weight loss, then by glucose lowering efficacy. After exendin-4 washout, the test was repeated to ensure reproducibility. Groups of responders and non-responders were then subject to RYGB surgery, and monitored for food intake, weight loss and improvements in glucose tolerance. No major differences in weight loss after RYGB surgery were apparent in the responder vs. non-responder groups. However glucose tolerance after surgery was preferentially improved in responders relative to non-responders, even when corrected for changes in weight loss. This increased responsivity correalted with increased levels of active GLP-1 and postprandial insulin in the responders. See GLP-1R responsiveness predicts individual gastric bypass efficacy on glucose tolerance in rats Diabetes. 2013 Nov 1.
Chambers and colleagues assessed glucose tolerance, insulin sensitivity, and incretin responses in several groups of rats subjected to either sleeve gastrectomy (VSG; 80% of the stomach is removed) or Roux en Y gastric bypass (RYGB) in which a small remnant gastric pouch was connected to the distal jejunum. Glucose clamps were done 2 weeks after surgery and glucose tolerance was assessed 5 weeks and 5 months after surgery in separate groups of animals. RYGB and VSG animals exhibited comparable reductions in food intake, weight loss, and improvements in glucose and insulin tolerance. Hepatic insulin sensitivity improved in both groups of bypassed rats independent of weight loss. Nutrient-stimulated levels of GLP-1 were higher in bypassed rats, glucose tolerance was improved and insulin levels post meal were higher; the majority of these changes in glucose tolerance were acutely attenuated by co-administration of the GLP-1R antagonist exendin(9-39). Weight-Independent Changes in Blood Glucose Homeostasis after Gastric Bypass or Vertical Sleeve Gastrectomy in Rats Gastroenterology. 2011 Sep;141(3):950-8. .
Abegg and colleagues acutely administered the GLP-1R agonist exendin-4, or the antagonist exendin(9-39) to adult male Wistar rats after RYGB surgery. Ex-9 increased food intake but had no effect on energy expenditure, whereas Ex-4 reduced food intake to a greater extent in RYGB rats, but had no acute effect on energy expenditure. Acute peripheral GLP-1 receptor agonism or antagonism does not alter energy expenditure in rats after Roux-en-Y gastric bypass Physiol Behav. 2013 Apr 3. doi:pii: S0031-9384(13)00103-0
Salinari and colleagues assessed changes in insulin sensitivity and GLP-1 levels in GK rats and Wistar rats after two different types of bariatric surgery, duodenal-jejunal bypass and jejunal resection. Both DJB and jejunal resection normalized insulin sensitivity in diabetic rats without changes in plasma GLP-1 levels Duodenal-Jejunal Bypass and Jejunectomy Improve Insulin Sensitivity in Goto-Kakizaki Diabetic Rats Without Changes in Incretins or Insulin Secretion Diabetes. 2013 Nov 15.
A role for PYY(3-33) as a regulator of GLP-1 secretion has been proposed. Plasma GLP-1 levels are lower after bariatric surgery in Pyy-/- mice, and levels of portal, but not systemic intact GLP-1, were higher in WT mice after intraperitoneal administration of PYY(3-36), findings blocked by the Y2R antagonist BIIE0246. Furthermore, the ability of PYY(3-36) to improve glucose tolerance was abrogated by co-administration of exendin(9-39). These findings delineate a portal neural PYY(3-36)-GLP-1 insulinotropic axis in mice. Peripheral activation of the Y2-receptor promotes secretion of GLP-1 and improves glucose tolerance Molecular Metabolism 2(3) 142-152, August 2013
Are there sexually dimorphic differences in mechanisms and outcomes of gastric bypass surgery that depend on sex steroids? Asarian and colleagues examined the effects of estrogen administration in ovariectomized rats after RYGP surgery. Estrogen potentiated the anorectic actions of GLP-1 and CCK and enhanced body weight loss over several weeks. Estradiol increases body-weight loss and gut-peptide satiation after Roux-en-Y gastric bypass in ovariectomized rats Gastroenterology. 2012 May 15.