Post-infectious Irritable Bowel Syndrome

Robin Spiller; Eugene Campbell

Curr Opin Gastroenterol. 2006;22(1):13-17. 

Abstract and Introduction

Purpose of Review: Irritable bowel syndrome patients form a heterogeneous group with a variable contribution of central and peripheral components. The peripheral component is prominent in irritable bowel syndrome developing after infection (post-infectious irritable bowel syndrome) and this has proved a profitable area of research.
Recent Findings: Recent studies have overthrown the dogma that irritable bowel syndrome is characterized by no abnormality of structure by demonstrating low-grade lymphocytic infiltration in the gut mucosa, increased permeability and increases in other inflammatory components including enterochromaffin and mast cells. Furthermore, increased inflammatory cytokines in both mucosa and blood have been demonstrated in irritable bowel syndrome. While steroid treatment has proved ineffective, preliminary studies with probiotics exerting an anti-inflammatory effect have shown benefit.
Summary: The study of post-infectious irritable bowel syndrome has revealed the importance of low-grade inflammation in causing irritable bowel syndrome symptoms. It has suggested novel approaches to irritable bowel syndrome including studies of serotonin and histamine metabolism which may be relevant to other subtypes of of the disease.

Although the idea of irritable bowel syndrome (IBS) developing after infection is not new, being first clearly described in 1962,[1] scientific study of mechanisms is relatively recent. The demonstration of mucosal abnormalities, overthrowing years of dogma that IBS is characterized by no abnormality of structure, has stimulated others to re-examine the IBS gut. Post-infective IBS (PI-IBS) develops in 3-30% of individuals with bacterial gastroenteritis. Known risk factors include female sex, severity of initial illness, bacterial toxigenicity and adverse psychological factors, including neuroticism, hypochondriasis, anxiety and depression, as reviewed in 2003.[2] The purpose of the current review is to update the literature since 2003, during which time there has been an explosion of interest and many productive new approaches with implications for novel treatments.

Clinical Features

The largest study so far of this condition examined questionnaires from 840 individuals who had stool culture-positive Campylobacter jejuni gastroenteritis. After excluding 46 with pre-existing IBS, 103 (13.8%) fulfilled the Rome I criteria for new IBS. Using the Rome II subtyping 63% were diarrhoea-predominant, 24% were alternating and 13% constipation-predominant. Ninety-six per cent reported loose motions and 60% bloating but, interestingly, one-third also reported straining. Similar predominantly diarrhoeal symptomatology was also reported in 18% of 169 students 6 months after returning from studying in Mexico, of whom six (10%) had developed PI-IBS.[3] Those who developed PI-IBS had significantly more episodes of diarrhoea [mean, 2.3 (range, 1-5) versus 1.1 (1-4) in those who did not]. Parry et al.[4] performed a prospective case-control study and found an incidence of PI-IBS of 16.7% with painless diarrhoea in a further 5.6%. The relative risk of developing functional gastrointestinal diseases in the 6 months following bacterial gastroenteritis compared with controls was 11.1 (95% confidence interval, 4.4-27.9). Interestingly, in this study functional dyspepsia was not increased; however, Mearin et al.[5*] investigating a Salmonella enteritidis outbreak in Spain reported a relative risk of developing new dyspepsia of 5.2 (2.7-9.8) only slightly less than that for IBS of 7.8 (3.1-19.7). The increased risk of IBS following Campylobacter and Salmonella infection appears non-specific, being also seen in studies on shigellosis,[6,7**] as well as a dual infection with Campylobacter and Escherichia coli 0157.[8*] These studies confirmed that severity of initial infection is the most important predictor of developing new IBS.[6,7**]

Importance of Psychiatric Features

Previous studies have indicated the importance of pre-existing psychological abnormalities and this was confirmed by Dunlop et al.,[9] who showed that depression and increased enterochromaffin cell counts were both independent risk factors for developing PI-IBS, conferring relative risks of 3.2 and 3.8, respectively, for each standard deviation increase. However, the importance of psychiatric disease appears less with PI-IBS since only 26% had a lifetime history of anxiety or depression compared with 54% of non-PI-IBS individuals.[10]

Role of Serotonin

Previous studies had shown that increases in both inflammatory cells and serotonin-containing enterochromaffin cells were a near-universal feature of Campylobacter jejuni gastroenteritis. However whether this correlated with symptoms was unproven until Dunlop et al.[9] showed that those developing PI-IBS had a significant increase in both enterochromaffin cells and lymphocytes in rectal biopsies compared to age and sex-matched infected individuals who had recovered without developing PI-IBS. The same authors also demonstrated that PI-IBS patients had increased postprandial peripheral plasma levels of serotonin[11**] when compared with both healthy controls and also patients with constipated IBS, who in contrast had a significantly depressed serotonin release. The same study demonstrated the expected decreased turnover of serotonin in the rectal mucosa of those with constipated IBS, as shown by a depressed 5-hydroxyindoleacetic acid/5-hydroxytryptamine ratio but also a paradoxical decrease in this ratio in PI-IBS patients.[11**] Since 5-hydroxyindoleacetic acid is derived from serotonin released from the enterochromaffin cell, in PI-IBS in which 5-hydroxytryptamine release is increased the 5-hydroxyindoleacetic acid/5-hydroxytryptamine ratio should be increased. One explanation for the observed depression is an impairment of serotonin reuptake. This idea was supported by a study demonstrating depressed serotonin transporter (SERT) both by immunofluorescence and mRNA analysis in both diarrhoea and constipated IBS.[12**] Depressed SERT may be acquired, possibly secondary to inflammation since it is depressed in ulcerative colitis. However, there is also a genetic basis to this with a promoter polymorphism which leads to lower expression of SERT. Those homozygous for the short form (ss) were over-represented in a diarrhoea-predominant IBS cohort reported by Yeo et al..[13*] However, the data on the SERT polymorphisms have been contradictory, with others finding an excess of the ss homozygotes in constipated IBS.[14] Recent data presented in abstract form suggest that this discrepancy may relate to the fact that the SERT promoter polymorphism may not result in altered SERT expression in the intestine since the mucosal SERT appears to have a different promoter site.[15]

Role of Inflammatory Cytokines

Two studies have recently reported increased expression of interleukin-1β in PI-IBS.[7**,16] While Gwee and colleagues[16] used rectal biopsies, Wang and colleagues[7**] performed colonoscopy on all their patients and were able to provide valuable data about the regional changes in mucosa following Shigella infection. They demonstrated increased interleukin-1β mRNA expression in both the rectosigmoid region and ileum, but not in IBS without an infectious origin.[7**] The same study is also interesting for its examination of the role of mast cells.

Role of Mast Cells

Mast cells were increased in the terminal ileal mucosa in both PI-IBS and non-PI-IBS, while mast cells in the rectum were within normal limits,[7**] as others have found.[10] Other authors have also recently reported increased numbers of mast cells in unselected IBS.[17**] These authors also demonstrated close association between mast cells and mucosal nerves. The numbers in close proximity to the nerves (<5 µm) were significantly correlated with severity and frequency of abdominal pain. Similar findings were reported by Wang et al.,[18] who found that in IBS patients there was a greater density of nerve fibres around the mast cells in both PI-IBS and non-PI-IBS when compared with controls. The study from Bologna[17**] is also notable because for the first time it used biopsies to assess mucosal release of mediators. They showed markedly enhanced release of histamine and mast cell tryptase.[17**] This is important because the biopsy supernatants have subsequently been used in a bioassay consisting of isolated gut segments from which the firing of afferent nerves is recorded.[18] This showed that such supernatants stimulated afferent firing and this may go some way towards explaining the link between PI-IBS and visceral hypersensitivity, which was noted early on by Gwee et al..[19] Others have also noted increased numbers of activated mast cells in close proximity to enteric nerves in both the caecum and rectum of IBS patients compared with controls.[20] Mast cells may also be important as mediators of increased gut permeability, which has been reported in PI-IBS.[8*,21] The Walkerton Health Study studied an outbreak of acute gastroenteritis in 2300 local residents, caused by contamination of the water supply by E. coli and C. jejuni. Of the 105 new cases of IBS which followed, a lactulose/manitol ratio of >0.02 was seen in 35% compared with just 13% for non-IBS controls. Increased permeability implies a breakdown of the normal barrier and access of bacterial products to the lamina propria, and may be a mechanism for perpetuating chronic inflammation.

Evidence of Chronic Inflammation in Irritable Bowel Syndrome

Whereas most studies of blood markers of inflammation have used IBS patients as negative controls, more recently it has been realized that IBS patients may show increased levels of inflammatory mediators. The most direct way to test this is to examine inflammatory mediators in the mucosa. A novel mucosal patch technique, in which pieces of filter paper attached to balloons are compressed against the rectal mucosa,[22*] has recently been reported. Filter-paper concentrations of myeloperoxidase and other neutrophil mediators such as human neutrophil lipokalin (HNL) were assessed in both active ulcerative colitis and compared with IBS patients and controls. As predicted, active ulcerative colitis showed markedly increased levels of inflammatory mediators but interestingly levels in IBS were comparable to those seen in inactive ulcerative colitis but significantly higher than in healthy controls. Another study used renal tubular proteinuria as a marker of inflammatory changes in 21 patients with IBS. Seven IBS patients had increased urinary α1-macroglobulin, a feature only seen in diarrhoea-predominant IBS patients.[23] The origin of this increased inflammation is uncertain, but may represent failure to downregulate inflammatory mediators induced by a bout of gastroenteritis that an average individual can expect to occur once a year. A recent genetic study found that the interleukin-10 polymorphism associated with high production of interleukin-10 (homozygous -1082*G) was significantly less common in IBS; 21 versus 32% in controls.[24] The theme of chronic inflammation and IBS-like symptoms was explored by Minderhoud et al.,[25] who demonstrated IBS-like symptoms in one-third of ulcerative colitis patients and 42% of Crohn's patients in remission. This confirms earlier work and supports the idea that chronic inflammation may leave permanent deficits.

Animal Models of Post-infective Irritable Bowel Syndrome

The Trichinella spiralis model is a widely used mode of PI-IBS and in the last 2 years there have been several important studies throwing light on the mechanisms. The most recent from McMaster University showed that interleukin-4 and interleukin-13 were able to induce hypercontractility in human cultured smooth muscle cells, an effect mediated via the STAT6 pathway,[26*] confirming what they had previously shown in mice.[27] The second study using the T. spiralis model showed that exposure of muscle cells to interleukin-4 or interleukin-13 increased transforming growth factor (TGF)-1β, cyclooxygenase 2 (COX2) protein and prostaglandin E2. COX2 inhibitors attenuated this hypercontractility and the authors conclude that maintenance of hypercontractility in the long term depends on increased expression of TGF-β1 and upregulation of COX2 and prostaglandin E2.[28*] The final common pathway appears to be prostaglandin E2, which increases the contractile response of muscle cells to carbachol. The same group from McMaster also showed that Lactobacillus paracasei was able to specifically reduce TGF-β production and the associated COX2 and prostaglandin E2 levels in muscles. This effect was seen during treatment with both the probiotic culture and spent culture medium, which contained only soluble mediators.[29*] These animal data give support to the idea that probiotics might be effective in PI-IBS.

Anti-inflammatory Effect of Probiotics

Numerous studies have shown an anti-inflammatory effect but the study by McCarthy et al.[30] is of particular interest since it showed a benefit in the interleukin-10-knockout mouse model of colitis using two probiotic bacteria, Lactobillus salivaris and Bifidobacterium infantis, which have since been used in IBS patients (see below). The same group also demonstrated that bacteria did not need to be living to be effective. Nor did they need to be administered orally but could be effective when given subcutaneously,[31] showing that the anti-inflammatory effect is systemic with a generalized decrease in proinflammatory cytokines.

Anti-inflammatory Treatments in Irritable Bowel Syndrome

With this evidence of low-grade lymphocytosis and inflammatory mediators the idea of treating flares in IBS like an exacerbation of asthma has certain attractions. The first direct trial of prednisolone in PI-IBS was, however, negative.[32] This randomized, double-blind, placebo-controlled trial randomized 29 patients with PI-IBS following C. jejuni enteritis to 3 weeks of oral prednisolone, 30 mg daily. Although mucosal lymphocyte counts fell, enterochromaffin cell numbers did not, nor did bowel symptoms improve significantly. Furthermore, prednisolone was poorly tolerated by IBS patients, indicating that this is unlikely to be a successful treatment. A more subtle and better-tolerated anti-inflammatory effect may be achieved by the use of probiotics. Since inflammation in the bowel appears in many instances to be a reaction to the commensal flora, modifying this by means of probiotics has proved an attractive option given its success in some inflammatory conditions such as ileal pouchitis. There have been several previous studies in IBS with mixed results, but the most recent one used the same two probiotics, L. salivaris and B. infantis, as McCarthy et al..[30] Only B. infantis was demonstrated to improve composite symptom scores as well as abdominal pain, bloating and distension.[33**] Like previous studies, this showed a small improvement in symptoms but, unlike previous studies, they also demonstrated a mechanism by measuring the release of cytokines by peripheral blood lymphocytes. They showed that at baseline IBS patients had a depressed interleukin-10/interleukin-12 ratio and that the active probiotics normalized this ratio. Since interleukin-12 is recognized to be a key proinflammatory cytokine, while interleukin-10 is anti-inflammatory, this imbalance may be important. The study certainly needs repeating, but this is the first study in which alterations in mucosal inflammatory mediators have been demonstrated to relate to improvement in symptoms (for review see)[34].


Studies of patients with PI-IBS over the last 2 years have been reviewed which demonstrate low-grade inflammation in the mucosa with increased inflammatory mediators, including serotonin and cytokines. Studies in other IBS groups have also shown increased mast cell numbers and in some cases increased release of mediators such as mast cell protease and histamine. This is plainly an active area in which significant advances with application to clinical practice can be expected in the foreseeable future.