Review article: the gastrointestinal tract: neuroendocrine regulation of satiety and food intake

Authors


  • Conflicts of interest:
    The authors have declared no conflicts of interest.

  • This article appeared as part of a supplement sponsored by Nycomed bv.

Dr J. Maljaars, Division of Gastroenterology–Hepatology, Department of Internal Medicine, University Hospital Maastricht, PO Box 5800, 6202 AZ Maastricht, The Netherlands.
E-mail: pwj.maljaars@intmed.unimaas.nl

Summary

Background  The gastrointestinal tract elicits numerous signals regulating food intake and satiety, and recently many studies have been performed to elucidate the mechanisms regulating these signals.

Aim  To describe the effects of the gastrointestinal tract on satiety, satiation and food intake.

Methods  A PubMed search was performed to identify and select the relevant literature using search terms including ‘gastric satiety, intestine + satiety, satiation, cholecystokinin, ghrelin, peptide YY, glucagon-like peptide-1 and ileal brake’.

Results  Satiation, satiety and food intake result, among other factors, from signals originating in the stomach caused by distension and signals from the small intestine. These intestinal signals result from nutrient sensing in the gut and activate neural and humoral pathways. Activation of the distal part of the gut, the so called ileal brake, leads to reduction in hunger and food intake, and models of chronic ileal brake activation lead to massive weight loss.

Conclusion  Gastrointestinal signals are crucial for the regulation of food intake, satiety and satiation. The ileal brake deserves special attention, as both ileal intubation studies and surgical studies demonstrate that activation of the ileal brake reduces food intake. In the surgical models, weight loss occurs without adaptation to the anorectic effects of ileal brake activation.

Introduction

The gastrointestinal tract elicits numerous signals regulating food intake and satiety.1 Both in the stomach and in the small intestine, the presence of nutrients, through distension of the stomach or by chemical interactions between nutrients and the luminal wall, induces the release of satiety signals.2–4 These signals contribute to the inhibition of food intake. Distension of the stomach by a volume load is able to reduce food intake significantly.2, 3 Exposing the small intestine to nutrients leads to release of gut peptides and neurotransmitters that induce a reduction in hunger levels and food intake.4–7

During the process of meal ingestion, some nutrients will already have entered the small intestine and gastric and intestinal satiety signals interact to limit meal size,5, 8 to increase satiation and to increase satiety between meals.

In this review, we will discus the satiety signals from the stomach and the small intestine (with a special focus on the ileum), how both interact with satiation, satiety and food intake and the underlying neuro-endocrine regulation of these signals.

We use ‘satiation’ to refer to processes that act to reduce or terminate eating within a meal and the (acute) feeling state associated with this. ‘Satiety’ refers to processes that act to inhibit early postprandially and/or reduce intake at later meals, and the associated feeling state (hunger suppression).

Methods

Relevant studies were identified by a PubMed search using search terms including ‘gastric satiety, intestine satiety, CCK, cholecystokinin, ghrelin, PYY, peptide YY, glucagon-like peptide-1 (GLP-1), glucagon-like peptide 1, ileal brake’. Original articles and reviews were selected, and references in these papers were also examined for relevant papers. Where possible, emphasis was given to human studies, but when data were lacking, animal studies were used.

Gastric satiety signals

Instillation of a volume load in the stomach induces distension of the stomach wall. The satiating effect of distension of the stomach has repeatedly been demonstrated: Feinle et al.9 used an inflated barostat balloon to demonstrate that a gastric distension increases fullness and Rigaud et al. and Nieben et al. found that the presence of an inflated gastric balloon in the stomach, as a treatment for obesity, reduced food intake and hunger.10, 11 Geliebter et al.12 showed that the threshold volume for a gastric balloon to reduce food intake is around 400 mL. These results suggest that a volume load by itself is sufficient to induce a gastric satiety signal and reduce food intake and hunger.9–12 The role of nutrient composition of a gastric volume load in the reduction of food intake and hunger has been studied extensively in rats: Phillips et al.2, 3 infused either saline or different nutrient solutions into the rat stomach and found that equivalent volumes produce equivalent reductions in food intake, without any effect of the nutritive value of the gastric load. However, when this load was allowed to exit the stomach, the nutrient infusions inhibited food intake more potently than the saline infusions.2 In humans, Goetze et al.13 found that postprandial hunger and satiety were correlated to postprandial gastric volumes, without significant effects of the nutrient composition of the meal. From these studies, one may conclude that gastric satiety is volume-dependent. Although the stomach is able to sense some aspects of the nutrient content of the stomach (i.e. to regulate gastrin release), this does not seem to play a role in gastric satiety.14 Up to 40% of a meal may have emptied from the stomach into the intestine before the end of meal ingestion occurs8 and interaction occurs between gastric and postgastric signals in physiological conditions. Many small intestinal satiety signals induce a delay in gastric emptying, and it is often thought that this inhibition is relevant for prolonging gastric distension.15

The role of the vagus nerve in gastric satiety has been studied in rats. Rats with pyloric cuffs underwent either sham vagotomy or total subdiaphragmatic vagotomy. After surgery, rats in both groups were fed 0-, 5- or 10-mL meals, and the effect of these premeals on food intake was assessed. In the sham-operated group, the 5-mL meal reduced food intake by 60% compared with the 0-mL meal, and by 80% in the 10 mL group. In the vagotomy-group, both 5 mL and 10 mL meals reduced food intake by 20% in either group, demonstrating a central role for the vagal nerve for gastric volumetric satiety.3 From this and other studies, it follows that gastric satiety is predominantly mediated by neural pathways.16–18

Intestinal satiety signals

Exposure of the small intestine to nutrients increases satiety and limits food intake.4, 6, 19 This increase in satiety occurs within seconds after the start of the intestinal nutrient infusion, suggesting that at least some of the intestinal satiety signals arise directly from the gut, rather than from postabsorptive nutrient processing.14 French et al.4 demonstrated that infusion of certain long chain fat emulsions into the duodenum reduced food intake significantly compared with a saline infusion and this has also been demonstrated for carbohydrates and protein.20 In the study by French et al.,4 a 180-kcal infusion of intralipid reduced food intake by over 200 kcal, demonstrating that intestinal nutrient infusion is capable of releasing potent satiety signals.

Gastric emptying of already a small part of a meal will activate duodenal nutritive signals, and from there, gastric volumetric and intestinal nutritive signals cooperate. In rats, gastric infusions of saline reduced food intake, whereas an equal infusion in the duodenum had no effect on food intake.3 An equivalent volume of 2.5–10% glucose did reduce food intake when infused into the duodenum, and this reduction in food intake could be enhanced by concomitant infusion of saline into the stomach, showing that a combination of gastric and duodenal signals produced an additive inhibition of food intake.3 In humans, Oesch et al. combined a 300-mL gastric distension prior to a meal with a duodenal infusion of either fat or saline. The distension combined with the saline infusion did not affect hunger, maybe due to a sub-threshold volume load,21 but when this distension was combined with a duodenal fat infusion, a decrease in hunger was observed. Food intake is significantly reduced by fat infusion, independent of the presence or absence of a distension.21 A gastric distension up to 600 mL was necessary to reduce hunger without fat infusion, while the fat infusion lowered the distension threshold profoundly.

Similarly, Feinle et al.9, 22 have shown that gastric distension-induced satiety scores were increased by a concomitant duodenal nutrient infusions, and subjects stated that adding a 20% lipid or maltodextrose infusion to the distension made the satiety sensation more meal-like.22 These studies are examples of how the stomach and the small intestine interact to limit food intake and reduce hunger feelings.

While gastric satiety is of mechanical origin (volume-dependent), intestinal satiety is typically nutrient-dependent: the satiety signals are released upon interaction between the gut wall and nutrients.3 Exposure of the small intestine to nutrients increases vagal afferent firing, and it is suggested that this is the case in intestine-mediated satiety.23–25 Intestinal nutrient exposure induces release of 5HT, a neurotransmitter, and Ondansetron, a 5HT3 receptor antagonist, has been shown to attenuate the inhibitory effects of intraduodenal Intralipid on food intake in rats.7 Direct vagal activation by intestinal nutrients may therefore be mediated by 5HT3.

Exposure of the intestinal mucosa to nutrients also induces release of many gut peptides and after release, these peptides either enter the bloodstream to function as a hormone, or they activate neural pathways.14 Many of these gut peptides delay gastric emptying, thereby contributing to prolonged distension of the stomach, satiety and subsequently to meal termination. Regional differences exist in the release of gut peptides (Figure 1, Table 1): in the duodenum and jejunum, exposure of the gut wall to fat and protein results in the release of CCK, whereas in the distal small intestine, the presence of nutrients induces the release of GLP-1, oxyntomodulin and PYY.14

Figure 1.

 Principal site of secretion of gut peptides involved in regulation of eating behaviour.

Table 1.   Gut peptides involved in satiety
Peptide Location of releaseFood intake
  1. GRP, gastrin-releasing peptide; NMB, neuromedin B.

CholecystokininProximal intestine, I-cells
Glucagon-like peptide 1Distal intestine, L-cells
Peptide YYDistal intestine, L-cells
OxyntomodulinDistal intestine, L-cells
EnterostatinExocrine pancreas
Apolipoprotein A-IVIntestinal epithelium
Pancreatic peptidePancreatic F-cells
AmylinPancreatic β-cells
GRP, NMBGastric myenteric plexus
LeptinGastric chief and P cells
GhrelinGastric X/A like cells

Proximal small intestine: cholecystokinin (CCK)  Cholecystokinin has been widely studied as satiety signal.1, 26, 27 CCK is released from intestinal I-cells in the duodenal and jejunal mucosa in response to the intraluminal presence of digestion products of fat and protein in the small intestine. Various biological forms of CCK exist, classified according to the number of amino acids they contain.28, 29 In humans, the main circulating forms are CCK-8, -22, -33 and -58.30 Two types of CCK receptors have been identified: the alimentary CCK-1 receptor (CCK1R) mediates gall-bladder contraction, relaxation of the sphincter of Oddi, pancreatic growth and enzyme secretion, delay of gastric emptying and inhibition of gastric acid secretion.31 The CCK-2 receptor (CCK2R) is the predominant CCK receptor in the brain, although the CCK1R has also been demonstrated in the hindbrain and the hypothalamus.31

Gibbs et al.27 demonstrated in rats that administration of CCK-8 induced a dose-dependent reduction in both the size and duration of a meal. In humans, Kissileff et al.32 were the first to demonstrate that i.v. administration of CCK-8 reduces food intake and hunger in humans. Following a 216-kcal preload, i.v. infusion of CCK-8 reduced ad libitum food intake by 19% when compared with a saline infusion.32 These results have been confirmed by several investigators.33–36 Lieverse et al.37 demonstrated a decrease of 19% in food intake after i.v. infusion of CCK-33. However, the effect of CCK-infusion on food intake is short-lasting: when the interval between the infusion and the start of the test meal is over 30 min, no anorexic effect will be observed.14

The role of endogenous CCK in the regulation of food intake has been studied with the specific CCK1R antagonist loxiglumide. I.v. infusion of loxiglumide attenuated the inhibitory effects of an intraduodenal fat infusion on subsequent energy intake38 and reduced satiety in humans.39,40

After release, CCK activates CCK1R on gastric and duodenal vagal afferents. These vagal afferents transmit the signal to the nucleus of the solitary tract and area postrema in the hindbrain, from where it is relayed to the hypothalamus.31 Vagotomy has been shown to attenuate the anorexic effects of peripheral CCK-administration,41 demonstrating a crucial vagal pathway for CCK-mediated satiation.

Gastric and intestinal satiety signals interact to reduce hunger. Regarding CCK, an intestinal satiety signal, it has been shown that the satiating effect of i.v. CCK infusion is enhanced by concomitant gastric distension. I.v. CCK-infusion combined with a-300 mL gastric balloon distension resulted in a stronger reduction in food intake compared with either CCK or distension alone.36 Furthermore, the CCK-induced dorsal vagal complex activity was increased by concomitant gastric distension, indicating a stronger anorexic effect.42, 43

Proximal small intestine: ghrelin  Ghrelin, an acylated peptide hormone, was first discovered by Kojima et al.44 Two/thirds of circulating ghrelin is produced by X/A-like cells in the gastric oxyntic mucosa, and most of the remainder originates in X/A-like cells of the small intestine. Gastrectomy results in a decrease in plasma ghrelin concentrations of 75%.45 Ghrelin is also produced in other organs, such as the pancreas, thyroid, adrenal, pituitary, hypothalamus, kidney, heart, lung, etc., but in much lower amounts than in the stomach.46

In humans, energy intake was increased by 28%, and hunger increased significantly in response to the ghrelin infusion.47 Administration of ghrelin stimulates gastric emptying and small bowel transit.

Ghrelin plasma concentrations peak before a meal and decrease postprandially, in response to nutrient ingestion.48 The premeal increase in ghrelin-concentrations implicates a role for ghrelin in mealtime hunger and meal initiation, as a signal for energy depletion.48 However, a recent study by Drazen et al.49 showed that when rats were forced to eat larger meals, this resulted in an increase in the preprandial ghrelin peak. These findings demonstrate that conditioning, and not just energy depletion, plays a role in premeal increases in ghrelin concentrations. In line with other observations, during a 24h fast, ghrelin concentrations increased before and decreased after the time points of customary meals, pointing to a conditioned pattern.50 It is therefore uncertain whether these preprandial increases reflect the need for food by the body, resulting in increased hunger and ingestion, or that these increases are not related to premeal hunger, but reflect a preparatory response to facilitate the subsequent digestion, absorption and metabolism of the upcoming meal.51

At least two mechanisms have been proposed to explain the decrease in ghrelin concentrations. Firstly, as mentioned above, when fasting, ghrelin concentrations decreased in the absence of a meal, and this is probably mediated by cephalic neural mechanisms.52 Secondly, the postprandial decrease in ghrelin concentrations is related to the ingested caloric load,53 with fat being less potent than either carbohydrates or proteins.51, 54 However, the mechanism regulating this decrease in ghrelin concentrations is not fully understood. While the majority of circulating ghrelin is produced in the stomach,46 stomach content does not play a role in the regulation of ghrelin concentrations, as intestinal infusion of nutrients also results in a decrease in plasma ghrelin concentrations,54 while inhibition of intestinal exposure by a pylorus-occluding cuff in rats did not affect ghrelin levels.55 Whether small intestinal exposure to nutrients has a direct effect, as suggested by Caixas et al.,56 or influences ghrelin concentrations via insulin release57–59 remains to be determined.

Several gut peptides, such as PYY, GIP, GLP-1, Somatostatin, CCK and also the vagal nerve, influence plasma ghrelin concentrations, and may therefore have a role in the physiological regulation of ghrelin concentrations.52

The orexogenic effect of ghrelin requires an intact vagal nerve: patients who underwent vagotomy did not demonstrate an increase in energy intake after i.v. administration of ghrelin.60

Distal small intestine: GLP-1  Glucagon-like peptide-1 is released from enteroendocrine L-cells in the distal small intestine, in response to exposure of the gut wall to especially fat and carbohydrates.61 Secretion occurs not only when nutrients are present in the lumen of the distal small intestine, but also from the duodenum by an indirect neurohumoral pathway involving the autonomous nervous system, Gastrin Releasing Peptide and acetylcholine.62

Infusion of GLP-1 results in a dose-dependent reduction in food intake and elicits satiety in normal weight, obese and diabetic humans.63 Apart from the effects on food intake, GLP-1 increases glucose-dependent insulin-secretion, reduces glucagon-secretion and increases pancreatic β-cell growth.61 GLP-1 mimetics are therefore being developed for the treatment of diabetes. Diabetic patients treated with GLP-1 or the GLP-1 receptor (GLP1R) agonist exenatide experienced progressive weight loss in trials lasting up to two years.14, 64 A 14-week trial with a long-acting GLP-1 analogue resulted in a significant weight loss compared to placebo,65 and this might be because of a decrease in appetite.64 A recent meta-analysis showed that administration of GLP-1 resulted in a reduction of food intake by an average 11%.63

Degen et al.66 demonstrated that the anorexic effect of GLP-1 could be increased by a 400-mL (225 kcal) protein drink, but not by 400 mL water administered 20 min prior to the test meal. As the authors did not measure gastric distension or gastric emptying, it is impossible to determine whether this reduced appetite is because of a nutrient-specific effect of gastric distension or because of duodenal exposure to nutrients. As gastric satiety signals are considered to be mechanical rather than nutritive, an additional duodenal signal is more likely to explain the additional anorectic effect. This study clearly demonstrated that GLP-1 interacts with gastric or duodenal satiety signals to increase satiety and reduce food intake.66

The mechanisms resulting in the anorexic effect of GLP-1 are not fully understood. A crucial role of the vagus nerve was demonstrated in rodents, where vagotomy abolished the anorexic effect of peripheral GLP-1 administration.67 Reduction of food intake by GLP-1 was mediated by the GLP1R, as a GLP1R knock-out mouse did not demonstrate a reduction in food intake.68 This receptor can be found in the gut, pancreas, brainstem, hypothalamus and vagal-afferent nerves.61

As long-term administration of GLP-1 reduces food intake and hunger and improves glucose control, increasing endogenous GLP-1-release by specific food products is therefore an interesting target.

Distal small intestine: PYY  Peptide YY is co-secreted with GLP-1 by enteroendocrine L-cells in the distal small intestine. It is released postprandially in proportion to the amount of ingested calories, with fat being the most potent macronutrient, followed by carbohydrates, followed by proteins.69 Direct exposure of the distal small intestine to nutrients will release PYY, but an indirect neurohumoral pathway from the proximal small intestine to the distal small intestine also exists.70, 71 PYY is released as PYY1-36, and rapidly cleaved by DPP4 into the active form of PYY, PYY3-36.69

Batterham et al.72 demonstrated that i.v. infusion of PYY3-36 significantly reduced energy intake by 33% over a 24-h period. Degen et al. have questioned the role of PYY as a physiological mediator of satiety. These investigators found that during infusion, PYY3-36 was able to reduce energy intake, but only at PYY-levels that were higher than they observed after a 1500-kcal meal, and therefore should be considered as supraphysiological.73 One study even failed to demonstrate any anorexic effect of PYY administration.74 In humans, a significant reduction in food intake is observed only at supraphysiological plasma PYY concentrations, and adverse events that may influence food intake such as nausea and abdominal discomfort are frequently observed.73, 75 For instance, in a recent study, five of nine participants failed to complete the PYY3-36 infusion protocol, because of severe malaise.75 These results have prompted some pharmaceutical companies to stop the clinical development of intranasal PYY3-36 because of insufficient efficacy.14

We previously conducted a study in which we compared the effects of intestinal fat infusion with i.v. PYY infusion. At PYY-concentrations that were similar after i.v. infusion and after intestinal fat, the effects on satiety after the intestinal fat infusion were much more pronounced.19

Batterham et al.72 have demonstrated a reduction in energy intake and resulted weight loss in mice. Although several investigators have not been able to reproduce and confirm these results,74, 76 the current consensus is that, in rodents, PYY is a potent anorexic agent and long-term administration leads to a significant weight loss.77, 78

The anorectic effects of PYY3-36 are thought to be mediated by the Y2-receptor in the hypothalamic arcuate nucleus, and this receptor may be directly accessible via the bloodstream.14 However, vagotomy abolished the anorectic effects of peripheral PYY administration, suggesting a role for the vagus nerve.67, 79

Ileal brake and satiety

Studies in animals have clearly demonstrated that the reduction in food intake is larger when the distal in stead of the proximal half of the small intestine is exposed to nutrients.80 In the ileum, exposure of the gut mucosa to especially fat and carbohydrates activates the ileal brake.40, 81, 82 The ileal brake is a negative feedback mechanism that potently inhibits gastric emptying and small intestinal transit (Figure 2).83 Furthermore, activation of the ileal brake results in a reduction in gastric acid-secretion, pancreatic enzyme secretion and bile acid secretion.84–86

Figure 2.

 Effects of ileal brake activation.

Welch et al.81, 87–89 were the first to demonstrate the potent anorexic effects of ileal brake activation in humans. Perfusing the ileum with 370-kcal corn oil emulsion resulted in a 575-kcal reduction in energy intake, demonstrating a net reduction of 205 kcal.88 In a subsequent study, they compared an ileal and a jejunal fat infusion to a control infusion, and found that the jejunal infusion induced the largest reduction in energy intake.87 It should be noticed that these infusions were at high supraphysiological rates. During the jejunal infusion, fat may also have activated ileal receptors, thereby explaining the increased potency of the jejunal infusion to reduce food intake.87 Our group recently conducted a study comparing the effects of duodenal and ileal infusion of fat on satiety levels, and found that the reduction in hunger levels was significantly larger after the ileal compared to the duodenal fat infusion.90

These studies have employed ileal intubation techniques to expose the ileum to increased amounts of nutrients, resulting in ileal brake activation. However, surgical techniques provide excellent examples of the effects of chronic ileal brake activation on food intake and satiety. The jejuno-ileal bypass (JIB) is a procedure in which the proximal part of the jejunum is anastomosed to the distal part of the ileum.91 This surgical procedure has been introduced as a treatment for massive obesity.92 Firstly, the mechanism resulting in weight loss was thought to be malabsorption. Condon et al. and others93–95 demonstrated that malabsorption could only account for 22% of weight loss; therefore, other mechanisms must be involved. JIB results in rapid transit of nutrients to the distal intestine, with increased exposure of the ileum to nutrients.96 Postoperatively, ad libitum food intake decreases.94, 95, 97 Bray et al. found that, compared to presurgical energy intake, energy intake was decreased by 80% at 1 month and with 45% at 12 months after JIB.95 Naslund et al.97 found a decrease of 13% 9 months after JIB. As JIB can lead to diarrhoea, psychological factors have also been suggested as an additional cause for this reduction in food intake,98 but more importantly, both GLP-1 and PYY levels are increased after JIB compared to preJIB levels.99 The chronically increased PYY and GLP-1 concentration may account for the reduction in food intake and weight loss.99

Which neural or hormonal mechanisms mediate the ileal brake is not exactly known. As the effects of an ileal infusion on pancreatic enzyme secretion could be abolished by infusion of a specific GLP1R-antagonist, GLP-1 is thought to play an important role in the ileal brake.100 Pironi et al. and Spiller et al. found that PYY correlated with changes in gastrointestinal motility, indicating a role for PYY in the ileal brake.82, 101 Therefore, PYY and GLP-1 are both considered mediatory peptides for ileal brake-activation.61, 102 Furthermore, Lin et al.71, 83, 103, 104 have demonstrated that different neural mechanisms are involved in mediating the different effects of ileal brake-activation.

The results from these surgical models demonstrate that no compensating behaviours on intake occur in response to the anorectic effects of chronic ileal brake activation. Therefore, we consider the ileal brake to be an excellent and promising target for the development of food products aimed a reducing food intake and hunger feelings.

Conclusion

The satiety signals that are derived from the stomach and the small intestine have been reviewed with special emphasis on their interaction in the reduction in hunger and food intake. Activation of the ileal brake significantly increases satiation and satiety, and reduces food intake. Surgical procedures that lead to increased exposure of the ileum to ingested nutrients confirm these results and indicate that adaptation to the anorectic effects of ileal brake activation does not occur, not even in the long-term follow-up.

The gastrointestinal tract is host to many signals that affect satiety and food intake. The ileal brake is considered an interesting target when designing products or procedures that are able to reduce food intake and produce weight loss. Not only from studies using ileal delivery of nutrients via oro-ileal, but also from studies on surgical models, it has become apparent that activation of the ileal brake leads to a reduction in food intake, weight loss, without long-term adaptation to these anorectic effects.

Ancillary