Appetite, Glycemia, and Entero-Insular Hormone Responses Differ Between Oral, Gastric-Remnant, and Duodenal Administration of a Mixed-Meal Test After Roux-en-Y Gastric Bypass

Roux-en-Y gastric bypass (RYGB) is an effective treatment for obesity and associated diseases (1). RYGB rapidly improves glycemic control, reduces hunger, and enhances satiation (2). The underlying mechanisms remain unclear, but weight loss–independent gut hormone–mediated responses to intestinal nutrient sensing may play a crucial role (3,4).

Traditionally, two competing theories were hypothesized (5). The “foregut theory” proposes that food bypassing the duodenum leads to decreased stimulation of currently unidentified hormones (“anti-incretins”), possibly contributing to improved glucose metabolism (6). The “hindgut theory” states that enhanced exposure of the distal intestine to undigested food triggers increased incretin secretion (e.g., glucagon-like peptide 1 [GLP-1] and possibly glucose-dependent insulinotropic peptide [GIP]), with consequently improved glycemia (7).

Few case studies compared metabolic responses to meals administered orally versus per gastrostomy into the bypassed stomach after RYGB (8–12). Consistent findings included an earlier rise in glycemia accompanied by a massive peak of insulinemia at 30 min after oral intake, compared with gastrostomy administration leading to lower glycemic excursions and negligible increases in insulin and GLP-1. All authors inclined to the hindgut theory as a plausible explanation (8–12).

These studies are limited by not having considered slower gastric emptying rates of the stomach remnant regulated by pyloric relaxations. In other words, differences between oral versus intragastric routes may result from differences in exposure rates of enteroendocrine cells to nutrients entering the small bowel.

We therefore hypothesized that nutrients delivered directly into the duodenum at similar rates lead to comparable changes in appetite, glucose metabolism, and gut hormones as when nutrients enter the gastrointestinal (GI) tract orally after RYGB. We tested this hypothesis by administering equivalent nutrients orally versus into the stomach remnant (intragastric; prepyloric) versus directly into the duodenum (intraduodenal; postpyloric) in a post-RYGB patient.

A 37-year-old Caucasian male without diabetes underwent laparoscopic distal RYGB (alimentary limb ∼500 cm, biliary limb 65 cm, and common limb 100 cm) for severe obesity (preoperative BMI = 43 kg/m²). Four months later, a reoperation for internal hernia became complicated by intestinal perforations that led to septic shock, prompting transfer to our bariatric referral center. Upon arrival, a laparotomy was performed to repair the intestinal perforations and to insert a double-lumen percutaneous gastrostomy to the remnant stomach, allowing gastric decompression and postpyloric enteral nutrition. An abdominal computed tomography located the tip of the postpyloric feeding tube in the third duodenal segment. The patient’s condition and weight stabilized (BMI = 35 kg/m²). Two months after the last intra-abdominal operation (6 months after the initial RYGB), in a noninflammatory and good general state, the patient voluntarily consented to participate in the current study. The enrolled patient provided written informed consent for voluntary participation in the experiment and to de-identified use of his medical records and images, and the Cantonal Ethics Committee of Zurich approved this study (BASEC-Nr. Req-2017-0616).

A standardized mixed-meal test (MMT) (200 mL Resource 2.0+Fibre; 400 kcal; carbohydrate 43%, fat 39%, and protein 18%) was administered at 0930 h, after an overnight fast at a similar delivery rate of 100 kcal/min (400 kcal in 4 min) by three different routes on consecutive days: 1) oral, 2) intragastric, and 3) intraduodenal (Supplementary Fig. 1). Each administration route was repeated once at the end of the first 3-day cycle. Results were calculated as the average of the two separate measurements per route. Blood was drawn at baseline and 15, 30, 60, 90, and 120 min after the end of the MMT for GLP-1, GIP, peptide YY (PYY), glucose, and insulin. At similar time points, systolic blood pressure and heart rate were recorded, and a validated visual analogue scale questionnaire on “hunger” and “fullness” was administered. On three separate days, early gastric emptying was assessed by an abdominal X-ray at 30 min after administering 100 mL of 1:1 diluted Gastrografin in 2 min by the three different routes described above, mimicking the rate and routes of previous meal administrations. Blood glucose levels were measured with the hexokinase method, and insulin was measured with a chemiluminescence enzyme immunoassay. Five milliliters of blood drawn into a chilled K2-EDTA tube was placed on ice immediately and centrifuged for 10 min at 3,000 rpm/min at 4°C, and the supernatant was stored at −80°C. Total GLP-1, total GIP, and total PYY were measured with Merck-Millipore human ELISA kits (EZGLP1T-36K, EZHGIP-54K, and EZHPYYT66K). Area under the curve (AUC) was calculated using the trapezoidal model. No formal statistics were performed due to low numbers (two measurements per route in one patient).

Postprandial perception of hunger and fullness ranged widely from “very hungry/not full” (intragastric) to “intermediately hungry/intermediately full” (intraduodenal) to “not hungry/highly full” (oral) (Fig. 1A and B). A similarly distinct distribution was observed in 30-min insulin responses, with a fivefold difference among the three routes: remnant gastric ∼60 pmol/L, duodenal ∼250 pmol/L, and oral ∼1,615 pmol/L (Fig. 1C). Glycemia peaked at 15–30 min ∼8 mmol/L after oral administration, whereas it remained steady after the other routes, with the duodenal route resulting in a small but constant increase compared with the gastrostomy (5.5 vs. 5 mmol/L) (Fig. 1D). The AUC for GLP-1 and PYY was comparable between the oral (26,523 pmol/L/min and 62,871 pg/mL/min, respectively) and duodenal route (28,801 pmol/L/min and 73,276 pg/mL/min), which were both much higher than those observed after intragastric feeding (11,529 pmol/L/min and 26,393 pg/mL/min) (Fig. 1E and F). The early GIP response was highly increased only after the oral route, with a peak at 15 min at 700 pg/mol, whereas the two other routes induced a peak ∼300 pg/mol at 60 min (Fig. 1G).

Supplementary Figure 2 illustrates the position of the contrast medium 30 min after administration via the three different routes. All contrast medium was in the small bowel after oral and duodenal administration, in contrast to the gastrostomy route, which showed some of the contrast still in the remnant stomach.

In this case study, we demonstrate that appetite, glycemia, entero-insular hormones, and gastric transit time after RYGB depend on the route by which nutrients enter the GI tract. Consistent with others, we observed the greatest reductions in appetite, along with highest postprandial glucose and insulin levels, after oral meal intake, whereas intragastric administration led to only minor changes in these variables.

However, observed differences in the physiological response to oral versus intragastric nutrient intake may be predominantly due to gastric-remnant emptying rate, which determines the rate of nutrient exposure of intestinal enteroendocrine cells, and thus, entero-insular hormonal responses. We specifically addressed this confounder by administering the same MMT at a similar delivery rate directly into the duodenum. To our surprise, and contrary to our hypothesis, the hormonal and appetite responses could not entirely be explained by the hindgut theory. We found that the duodenal MMT led to similar AUC values as the oral route for GLP-1 and PYY only, even though there were marked differences in the time course of the hormonal responses; an earlier rise and earlier decrease in GLP-1 levels was observed after oral administration. In contrast, for glucose, insulin, and GIP responses, the duodenal route was more similar to the intragastric route. Interestingly, when comparing responses after the three routes of administration, insulin levels did not seem to be tightly associated with elevations in incretin hormones, suggesting that incretins, especially GLP-1, did not exert a major influence on hyperinsulinemia or glycemic control in this study. Moreover, postduodenal appetite levels were intermediate between oral and intragastric routes.

Beside the role of mere food rerouting in provoking altered GI hormonal responses, other factors might contribute to changes in postbariatric glucose metabolism: changes in stimulation of mechanoreceptors (3), altered intestinal nutrient sensing (1), accelerated food transit (13,14), and upregulation of the sodium–glucose cotransporter 1 in the alimentary limb after RYGB (15). Oral MMT might lead to a short, but intense, activation of satiation signals by stretching the gastric pouch, which may amplify the effect of intestinal nutrient sensing on appetite (3).

The main limitations of this study include the low sample size and limited comparability due to the patient’s distal RYGB, as well as the lack of a nonsurgical control.

In conclusion, this study is the first to compare metabolic responses to MMTs administered by three different GI routes after RYGB. Duodenal nutrient sensing contributes to the regulation of both food intake and glucose homeostasis (16), and intraduodenal meal administration seems to have a pivotal role in exploring the metabolic consequences of GI rerouting after bariatric/metabolic surgery (17). Our findings indicate that the remnant gastric emptying rate impedes the metabolic response to MMTs administered by gastrostomy after RYGB. Hence, intraduodenal meal administration should be used in future studies to circumvent this bias.

Acknowledgments. The authors are grateful to Pelagia Kakka (architecture student, Democritus University of Thrace, Xanthi, Greece) for creating the schematic illustration. The authors are also grateful to Udo Ungethuem (lead laboratory technician, Department of Surgery and Transplantation, University Hospital Zurich) for his important help in performing the immunoassays.

Funding. This study was entirely funded by the assistant-professorship research grant awarded by the University of Zurich to M.B.

Duality of Interest. R.E.S. is employed by DSM Nutritional Products. D.E.C. is a principal investigator on both the Comparison of Surgery vs. Medicine for Indian Diabetes (COSMID) trial, which is funded by Johnson & Johnson, and the Alliance of Randomized Trials of Medicine vs. Metabolic Surgery in Type 2 Diabetes (ARMMS-T2D) trial, which was previously sponsored by Johnson & Johnson as well as Covidien but which is now entirely funded by the National Institutes of Health. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. D.G. designed the study; acquired, analyzed, and interpreted data; and drafted the manuscript. R.E.S. designed the study and acquired, analyzed, and interpreted data. H.H. acquired data. D.E.C. interpreted data. M.B. designed the study, analyzed and interpreted data, and drafted the manuscript. All authors critically revised and gave final approval of the manuscript. M.B. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.