Functional dyspepsia (FD) is a chronic symptom complex characterized by epigastric pain or burning, postprandial fullness, or early satiation.¹ The symptoms are thought to arise from the gastroduodenal region without any organic, metabolic, or systemic disease that is likely to explain the symptoms.² FD is a common but heterogeneous condition. Gastro-duodenal motor dysfunction, including delayed gastric emptying or hypo-motility, dysfunction of gastric accommodation, and altered intestinal hypersensitivity have been the focus of attention until recently.^3–7 The prevalence of motor dysfunction is about 20–30% in patients with FD including occasionally rapid gastric emptying.^3,4 While slow gastric emptying has been associated with FD symptoms in some studies, in particular post-prandial fullness and severe vomiting in FD,⁵ large scale, multi-center studies have failed to demonstrate any link between gastric emptying and specific symptoms.^6,7

The duodenum serves to regulate acid secretion from the stomach and the nutrient absorption in the small intestine via local signaling pathways, and connects with the central nerve system, via neuronal and endocrine mediators.⁸ Duodenal acidification may provoke epigastric pain and aggravate dyspeptic symptoms through the induction of hypersensitivity to gastric distension in healthy subjects.⁹ Intra-duodenal lipid, not glucose, may be responsible for symptom generation in FD through the induction of a duodenogastric reflex resulting in gastric relaxation, inhibition of antral motility, and increased sensitivity of the proximal stomach to distension, or an abnormal response to cholecystokinin in patients with FD.¹⁰

In FD, there are traditionally no structural or biochemical abnormalities found that explain chronic symptoms, however, recent research is providing evidence of structural upper gut alterations in certain subsets. In particular, subtle inflammatory and molecular changes in the duodenum may play an important role in the pathophysiology of FD. In this review, we discuss the proposed link between chronic low-grade inflammation and the pathogenesis of FD, potential treatment implications, and future perspectives.

Low-grade inflammation in patients with functional gastrointestinal (GI) disorders may provoke localized impairment in motor-sensory abnormalities, the GI neural axis, and neuroimmuno-dysregulation. Duodenal micro-inflammation is observed in FD, and has been associated with dyspeptic symptoms, such as postprandial symptoms and epigastric pain, in adults and pediatric populations.^11–13 The landmark study was conducted using a population based endoscopic study approach allowing for the inclusion of a representative community control group without FD.¹¹

Eosinophils and mast cells may have a potential key role, through a variety of immune responses.¹² Recent studies on low-grade inflammation in patients with FD are expressed using quantitative assessments (Fig. 1). A meta-analysis of 37 studies evaluating peripheral and mucosal immune cytokines and inflammatory cells was performed.¹⁴ Gastric mast cells were increased in patients with FD compared to controls, and in a subgroup analysis of 3 studies evaluating FD patients without Helicobacter pylori infection, mast cell counts remained elevated in FD compared to controls.^15–19 However, a number of studies failed to find an increase including a population-based case control endoscopic study.¹² A significant increase of eosinophils was also noted in the stomach of FD patients compared to controls.^{11,15,18–22} Duodenal intra-epithelial lymphocytes and neutrophils were not different among individuals between FD and controls,^12,22–25 and inflammatory cytokines in the stomach of FD patients, such as IL-1β, IL-6, IL-8, and IL-10, were not different but with study inconsistency.^17,18,20,23 The enterochromaffin cells (ECs) in the stomach were similar between 2 groups and serotonin contents, serotonin contents, TPH-1 mRNA, a rate limiting enzyme of 5-hydroxytryptamine synthesis in ECs, SER mRNA expression also were not different among individuals with FD and controls.²⁰ In another study, the number of endocrine cells was significantly lower in FD patients versus controls, whereas there was no significant difference in 5-hydroxytryptamine content.²⁶

In an adult population based endoscopic study, eosinophils were specifically increased in the duodenum of FD community subjects, but mast cells were also significantly increased in FD subjects with overlapping IBS and FD.¹² In a meta-analysis, increased duodenal eosinophils infiltration was noted in patients with FD compared to controls, despite significant heterogeneity and possible publication bias.¹⁴ Two studies showed increased eosinophils in postprandial distress syndrome (PDS),^11,23 not in epigastric pain syndrome (EPS), however, a subgroup meta-analysis demonstrated higher duodenal eosinophil counts in both EPS and PDS. Of the 10 studies that evaluated mast cell infiltration, 5 studies reported increased duodenal mast cells and the pooled results showed significantly higher mast cell counts in the duodenum.¹⁴ It is unclear if the increase in duodenal mast cells is restricted to those with FD and IBS overlap (one third of FD cases),¹² as increased mast cells have also been observed in the terminal ileum and jejunum in IBS.^27,28

Eosinophils are detected normally at low levels in the GI tract from the stomach to the small and large intestine. Unlike intraepithelial lymphocytes and mast cells, eosinophils are not normally present in Peyer’s patches or intraepithelial locations. Mature mast cells are ready for optimal interaction with the local environment and comprise 1–5% of mononuclear cells in the lamina propria and the submucosa of the gut.²⁹ A reference range for significant increased eosinophils and mast cell counts is still lacking because of the standardization on the methodology used to count these cells, differences in patients and control selection, inter-individual variability, geographic variation, and the relatively small numbers for individual studies.

The eosinophil has pleomorphic effects: (1) eosinophils release cytotoxic granules, eosinophil peroxidase, major basic protein, eosinophil cationic protein, and eosinophil-derived neurotoxin; (2) eosinophils release a variety of cytokines and neuro-mediators; (3) eosinophils release lipid mediators, such as leukotrienes or platelet activating factor; and (4) eosinophils induce the expression of MHC class II and co-stimulatory (eg, B7) molecules by presenting antigen to T-cells leading to immune activation.³⁰

Mast cells induce bone marrow-derived eosinophils to migrate into the mucosa and in turn, eosinophils can activate mast cells via cytokines or mediators.³⁰ Eosinophil-derived major basic proteins can induce vagal M2 receptor dysfunction.³¹ Also, eosinophils activate mast cells which release leukotrienes, a potent stimulator of smooth muscle.³⁰ Mast cell activation provokes the spontaneous or stimulated release of mediators, most commonly tryptase and histamine and less often carboxypeptidase A2, heparin, hexosaminidase, chromogranin A, leukotriene E4, and prostaglandin D2.²⁹ Recent studies suggest that proton pump inhibitors (PPIs) directly inhibit IL-4 stimulated eotaxin-3 expression, an eosinophil chemo-attractant, and block STAT6 binding to the promoter in patients with eosinophilic diseases.^32,33 These findings suggest that PPIs may have a therapeutic value in a subset of patients with FD through a direct anti-inflammatory action, and in randomized controlled trials PPIs are superior to placebo in FD particularly PDS.³⁴

H. pylori infection clearly provokes chronic mucosal inflammation in the stomach and duodenum, which in turn, might lead to gastroduodenal motor and sensory dysfunction, and also has been linked with dyspeptic symptoms. In subjects with H. pylori infection, the increase in eosinophils was not significantly correlated with the degree of gastric or duodenal mucosal inflammation or the accompaniment of allergy, but the eosinophil count in the esophagus and gastric mucosa was significantly higher compared to a H. pylori negative group.³⁵ A recent meta-analysis showed that mast cells and eosinophils were increased in patients with FD compared to controls, however, in a subgroup analysis of 3 studies evaluating its association with H. pylori infection, gastric mast cells were still increased, but gastric eosinophils were not increased in FD subjects without H. pylori infection.^18,19,22

H. pylori is a common and worldwide infection, with a prevalence ranging from 18.9% to 87.7%,³⁶ which is regionally diverse. Furthermore, the presence of H. pylori is not associated with any specific symptom profile.³⁷ A recent meta-analysis reported a significant improvement in dyspeptic symptoms in the H. pylori treatment group (40.6%) versus controls (34.0%). The number needed to treat for H. pylori eradication to provide dyspeptic symptom improvement was 15, a modest effect.³⁸ Although the magnitude of the relevance of H. pylori infection in FD has been a contentious issue, it does seem that its eradication provide symptom relief in as small subset.

Among a subset of FD patients, dyspeptic symptoms may arise after a bout of gastroenteritis, so called post-infectious FD. Salmonella, Escheria coli, Giardia lamblia, Campylobacter jejuni, and norovirus may lead to the development of new onset chronic dyspeptic symptoms after an acute bout of gastroenteritis that resolves.³⁹ A meta-analysis revealed that the prevalence of post-infectious FD was around 10% in the adult population, and the summary odd ratio for the development of post-infectious FD was 2.54 (95% CI, 1.76–3.65) at more than 6 months after acute gastroenteritis.⁴⁰ Post-infectious FD is presumed to induce a persistent inflammatory state due to the occurrence of inefficient down-regulation of the immune response to the initial infectious agent. In patients with post-infectious-FD, duodenal biopsy revealed persisting focal CD8+ T-cell aggregates, and decreased numbers of CD4+ T cells.⁴¹ Another study from Japan provide additional evidence of immune dysfunction by showing microscopic duodenitis with increased eosinophils and CD68+ and chemokine receptor 2 (CCR2)+ macrophages.⁴² In a study comparing 37 FD patients with a history of acute gastroenteritis 6 to 12 months ago and non-specific FD patients, the number of ECs and mast cell were increased in the gastric antrum.⁴³ Electron microscopy revealed activating mast cells and ECs at a distance of < 5 μM of nerve fibers were significantly greater in post-infectious FD in the antrum versus nonspecific FD or controls.⁴³

Increased viscero-perception is an important etiologic factor in FD and acid may play a role. Although meta-analysis revealed that PPIs and histamine H₂ receptor antagonists are efficacious therapies for treating FD, low-dose PPIs have similar efficacy as standard-dose PPIs.³⁴ Furthermore, gastric acid secretion is within the normal range in these patients.⁴⁴ Duodenal acidification induces gastric relaxation by exerting an inhibitory effect on the stomach slowing gastric emptying.^45,46 Intraduodenal acid, but not saline, increased upper GI symptoms and an experimental hyper-acidic state in the proximal duodenum decreased duodenal contractile velocity in patients with FD, but not in healthy volunteers.^47,48

Low-grade inflammation in FD, most notably eosinophils and mast cells, is likely to alter neural structure and function.⁴⁹ In IBS, mast cells in close proximity to nerves in the colonic mucosa significantly correlated with the frequency and severity of abdominal pain or discomfort.⁵⁰ The co-labeling technique of mast cells and nerves in fixed human specimens revealed altered cell density and reactivity, however, application of advanced imaging techniques based on the use of dyes that detect changes in membrane potential or intracellular calcium have enabled the detection of nerve activation in submucosal or myenteric plexus layers of human tissues.⁵¹ In patients with IBS, increased mast cells in the colonic mucosa were localized close to the nerve fibers, with a marked increase in the firing rate of visceral sensory nerves and enhanced Ca2⁺ mobilization in the vast majority of capsacin-sensitive dorsal root ganglion neurons following exposure to IBS mucosal supernatants.⁴⁹ Therefore, mast cells in the mucosa of patients with IBS can release mediators that excite afferent neurons and may cause visceral hypersensitivity.

There are a few studies on the potential role of the eosinophil-mast cell axis in altering enteric nervous system function in FD patients. In pediatric patients with FD who also had atopy, increased mast cells were closely associated with mucosal nerve fibers and released tryptase co-localized with proteinase-activated receptors on mucosal nerve fibers.⁵² Cirillo et al⁵³ demonstrated impaired neuronal function in the submucous plexus of FD patients, as shown by decreased calcium responses to depolarization and electrical stimulation. Further, the neuronal impairment significantly correlated with eosinophil and mast cell infiltration in the submucosal layer of FD patients. They hypothesized that, through mediator release and recruitment of mast cells and eosinophils, neuronal muscle cell functioning is altered. There was a significant negative correlation between nerve activity and immune cell numbers.^52,53

Gut-trophic T cells that express high levels of α4-integrin (CD49d) and β7-integrin and CCR9 tend to migrate to the lamina propria of the small intestine.⁵⁴ Circulating CD4⁺CCR9⁺ T cells are increased in the peripheral blood of patients with small bowel celiac disease and Crohn’s disease.⁵⁵ CD4+ T lymphocytes co-expressing the gut homing marker CD49d, β7-integrin, and CCR9 were increased in H. pylori-negative FD patients, which demonstrated an association of systemic cellular immune activation with symptom manifestation in FD.⁵⁶

Psychological factors, as manifested by anxiety or depression, may contribute to the immune activation or be in some cases driven by it.⁵⁷ Corticotrophin-releasing hormone (CRH) is released by stress by activation of the hypothalamus, and also produced by peripheral inflammatory cells.⁵⁸ In patients with IBS, anxiety was related to increased infiltration of rectal mast cells, and anxiety, depression, and somatization yielded statistically significant relationships with gastric mast cell density in pediatric FD patients.^59,60 Neuro-immune interactions have been implicated in IBS,⁶¹ which can be precipitated by physical or psychological stress.^62,63 Acute stress evoked by immobilization leads to colonic responses via CRH.⁶⁴ CRH can induce intestinal mast cell degranulation directly leading to increased vascular permeability.⁵⁸

Taken together, the data suggest that low-grade inflammation may activate the enteric nervous system and interact with the central nervous system. Higher pre-existing psychological problems also may result in autonomic nervous system imbalance, further impairing sensory-motor dysfunction.

The intestinal barrier is a complicated and effective defense system that limits luminal antigen access while maintaining nutrient and electrolyte absorption. Intestinal permeability is mainly determined by epithelial physicochemical barrier function aided by mucosal immunity and neural activity.⁶⁵ An increase in intestinal permeability has been found in inflammatory bowel disease and IBS, especially diarrhea-predominant IBS or post-infectious IBS (PI-IBS)^66,67 and more recently also in FD.²³

In patients with PI-IBS, intestinal permeability was increased after an outbreak of bacterial gastroenteritis by documentation of an increased lactulose/mannitol ratio in the urine of IBS patients compared to healthy controls.⁶⁷ Increased gut permeability was also noted in non-PI-IBS.⁶⁸ CDH-1, which codes for E-cadherin, a tight junction protein, is linked to PI-IBS, hence the increased permeability may underlie the pathogenesis of PI-IBS.⁶⁹ The apical junctional complex keeps enterocytes tightly sealed and regulates paracellular permeability. The tight junction is constituted by surface-membrane proteins, including occludin, claudins, junctional adhesion molecules and tricellulin, and intracellular proteins, zonula occludins (ZO)-1, 2, 3, which anchor to the actin cytoskeleton.⁷⁰ Adherence junctions are mainly made up of e-cadherin, catenin, and actin filaments.^71,72 In the colonic mucosa of IBS patients, increased permeability has been linked to the down-regulation of ZO-1 and proteasome-mediated occludin degradation^73,74 and mast cell activation was associated with down-regulation and redistribution of ZO-1 and occludin in the jejunum.⁶⁵

Vanheel et al²³ reported impaired mucosal integrity and abnormal gene expression of occludin, β-catenin and desmosomal proteins in FD patients compared to controls. Furthermore, there was a significant association between the expression of cell-to-cell adhesion proteins, increased permeability and the low-grade duodenal inflammation. It needs to be determined if the reduced barrier function precedes the activation of local immunity already observed in FD patients and whether this condition drives symptom development.

Complex dietary components can be converted by the microbiota to produce various metabolites. The fermentation of complex carbohydrates mostly results in the production of short chain fatty acids, such as acetate, propionate, and butyrate.⁷⁵ These metabolites are sensed by the enterocytes and also diffuse across the epithelium, where they can provoke immune activation and be connected with the enteric nervous system.⁷⁶ In patients with IBS, brushings of the duodenal mucosa in patients with IBS showed significantly lower levels of Bifidobacterium catenulatum.⁷⁷

Stress reduces water absorption and increases jejunal secretion in healthy subjects through the parasympathetic nerves and mast cell activation, and the corticotrophin-releasing factor enhances transcellular uptake in the human colon through the CRF-R1 and CRF-R2 receptors on colonic mast cells.⁷⁸ In an IBS rat model, chronic early life stress induced hypercortisolemia, and enhanced intestinal permeability and stress decreased fecal microbial diversity, which was characterized by an increased abundance of gram-positive cocci and reduction of fiber degrading, butyrate-producing, and mucus-resistant microbes.⁷⁹ This dysbiosis may affect the susceptibility of people to visceral hypersensitivity through immune activation. Studies on microbiota are rare in dyspepsia patients compared with IBS. Several studies on small intestinal dysbiosis related with PPIs and its impact on the gut have been published.⁸⁰ PPI use is associated with decreased bacterial richness and deviation of oral-associated bacteria in the gut microbiota.⁸⁰

Zhong et al⁸¹ assessed stool and duodenal mucosal microbiota in patients with FD and controls (Fig. 2). In FD patients, there was an inverse relationship between the relative abundance of Streptococcus and the anaerobic genera, and clear distinction from controls. In an analysis of the gastric fluid microbiota in patients with FD, Bacteroides > Proteobacteria abundance and the absence of Acidobacteria at the phylum was found.⁸²

More microbiota studies will be important. Given wide human microbiome heterogeneity and that microbial diversity depends on diet and other environmental exposures, a well standardized methodology and longitudinal studies as well as intervention studies will be needed to determine if duodenal dysbiosis is relevant to the pathogenesis of FD.

Many patients with FD report meal-related symptoms, such as fullness, bloating, epigastric pain or burning, so it is likely that food and dietary habits can induce and/or exacerbate dyspeptic symptoms. Intra-duodenal infusion of fat, in contrast to glucose, induced dyspeptic symptoms by impaired gastric accommodation in patients with FD.⁴⁸ However, the association between symptoms and motor abnormalities appeared to be relatively weak. The role of dietary factors related directly to food ingestion, including patterns of nutrient intake and potential intolerance to specific foods or macronutrients, might be more important.⁸³ A recent systematic review showed that dietary fat was associated with dyspepsia onset after a meal challenge,⁸⁴ and the proposed mechanism was by hypersensitivity to GI hormones such as cholecystokinin⁸⁵ or delayed gastric emptying.⁸³

Wheat, specifically gluten proteins, may trigger dyspeptic symptoms. Furthermore, a gluten-free diet induced symptomatic improvement in 2 studies,^86,87 but elimination of dietary wheat also substantially reduces the fermentable oligosaccharides, disaccha-rides, monosaccharides, and polyols (FODMAP) content.⁸⁷ Non-celiac gluten or wheat sensitivity (NCG/WS) is characterized by upper or lower GI symptoms following the ingestion of gluten- or wheat-containing foods, in the absence of coeliac disease or what allergy.⁸⁸ The underlying mechanism of NCG/WS is unknown, but the innate immune system has been implicated, and there is some possibility of overlap of the condition and FD. Uhde et al⁸⁹ reported that serum soluble CD14 and lipopolysaccharide-binding protein and fatty acid-binding protein 2, a marker of intestinal cell damage, were significantly elevated in wheat sensitivity, and these results suggest a state of immune activation in conjunction with a compromised intestinal barrier affecting in a subset of NCG/WS.

Duodenal eosinophilia have been documented in children with dyspepsia. In pediatric patients with FD, a randomized controlled trial with monolukast, an eosinophil stabilizing drug, showed that pain and peak eosinophilic infiltration of the duodenum were significantly improved.⁹⁰ Another study was conducted in pediatric dyspeptic patients with duodenal eosinophilia to evaluate the efficacy of a H1/H2 antagonist. About half of the patients showed a response to histamine H1/H2 blocker and when oral cromolyn was administered to the non-responders, it appeared effective at 89%.⁹¹ These studies have stimulated considerable interest in the possibility that eosinophils are important effector cells in FD and replication of these studies are waiting to be conducted in adults with FD. Cor-ticosteroids have been the mainstay of pharmacological treatment for eosinophilic disorders, however, they have not been tested yet in FD.

Probiotics have been reported to be beneficial in the treatment of IBS and may act by improving mucosal permeability and altering the intestinal microbiota.⁹² Probiotics might improve H. pylori-associated dyspepsia through an inhibitory effect on H. pylori,⁹³ however, the effects of probiotics on H. pylori-negative FD is unclear. Igarashi et al⁸² reported that the overall bacterial community structure in the gastric fluid of patients with FD was significantly different from that in healthy controls, and the shifted composition of the gastric fluid microbiota was “restored” by treatment with Lactobacillus gasseri OLL271.⁸² A randomized controlled trial with L. gasseri OLL2716 in 116 patients with H. pylori-uninfected FD revealed that the symptom resolution was achieved by 17% and 35.5% in the placebo and L. gasseri OLL2716 (P = 0.048) arms, respectively.⁹⁴

Emerging data which account for the effectiveness of rifaximin in FD was reported.⁹⁵ Adequate relief of global dyspeptic symptoms was 79% versus 47% in rifaximin and placebo, respectively, at 8 weeks (P = 0.008), as well as improvement in belching and postprandial fullness in sub-group analysis. Rifaximin might alter the intestinal gut microbiota, which has been demonstrated in IBS studies.⁹⁶

Dietary factors are known to precipitate or exacerbate FD symptoms. Because of the association with NCG/WS, nutritional management may be beneficial. A gluten-free diet in patients with dysmotility-like dyspepsia symptoms and a normal endoscopy followed for 18 months resulted in dyspeptic symptom relief in 92% (31/34).⁸⁷ FODMAP can evoke the luminal distension which is related to symptom generation and associate with visceral hypersensitivity in IBS.⁹⁷ There are several studies that show the efficacy of the low FODMAP diet in IBS.⁹⁸ There is overlap of symptom manifestations between FD and IBS, and therefore, a low FOD-MAP diet might have some role in the management of FD, but has not been studied yet.

There are several lines of evidence that duodenal low-grade inflammation may be involved in the etiopathogenesis of FD inducing mucosal immune activation, duodenal barrier dysfunction, and sensory-motor dysfunction (Fig. 3). An altered duodenal gut microbiota, food antigens or infection may precipitate duodenal micro-inflammation in a subset of FD patients. If this hypothesis is correct, there are important implications in terms of potential new therapeutic targets in FD.