• irritable bowel syndrome
• infection
• mental health
• prognosis

Chaudhary and Truelove were amongst the first to recognise that a proportion of patients with irritable bowel syndrome (IBS) report an apparent association between the onset of symptoms and a dysenteric illness.¹ The concept of “post-dysenteric IBS” (PD-IBS) has now been widely accepted with claims suggesting that it accounts for anything up to 25% of the totality of the condition. However, careful questioning of these patients sometimes, but not always, suggests that they may have had a “forme fruste” of the disorder before their infection, raising the possibility that there may be two forms of the illness, one being an exacerbation of a pre-existing disorder and another where the condition appears to arise de novo. Ultimately, issues surrounding the natural history of PD-IBS will only be resolved by detailed prospective and retrospective studies, such as the one reported by Neal and colleagues² in this issue of Gut[see page 410].

Given that dysentery leads to irritable bowel symptoms, what mechanisms might be responsible? Do the dysenteric organisms cause the problem directly, and if so, are some more noxious than others? Alternatively, does the resulting diarrhoea lead to a form of non-specific sensitisation?

A retrospective analysis of the prevalence of gastrointestinal symptoms following microbiologically confirmed gastroenteritis suggested that there was no difference in the risk of IBS between bacterial species.³ However, it is noteworthy that up to half of the stool cultures obtained in some studies of PD-IBS may in fact be negative. This suggests that viruses or possibly other pathogens such as parasites may also be capable of provoking IBS in some individuals. There has been relatively little work addressing the possible role of viruses in IBS although it is tempting to speculate that such agents might be implicated in some cases, especially as symptoms of IBS are very common in chronic fatigue syndrome for which a viral aetiology has been proposed.

In addition to the role of different pathogens, it is possible that exposure to other factors which may be associated with the infection could be critical to the subsequent development of IBS. For instance, it has been shown that diarrhoea induced by polyethylene glycol can lead to rectal sensitisation,⁴ suggesting that diarrhoea per se is important. It is also well known that the use of antibiotics often results in diarrhoea, and there is evidence that patients suffering from dysentery treated with antibiotics are more likely to develop IBS compared with those who do not receive this form of therapy. Furthermore, this effect has been observed with antibiotics used for reasons other than dysentery.⁵ A further explanation for the deleterious effect of antibiotics in IBS might be alteration of the bacterial flora in the gut. If this is the case, one might predict that measures aimed at favourably modifying the gut flora might offer therapeutic promise, and reports of the beneficial effect of probiotics would seem to support this view.

Gastroenteritis can lead to disruption of the gastrointestinal mucosa resulting in excessive antigenic exposure and loss of the brush border. However, it is well known that the resultant disaccharidase deficiency is usually transient, and therefore unlikely to be contributory to the pathogenesis of PD-IBS. A similar mechanism may lead to bile salt malabsorption, and a group of patients with chronic gastrointestinal symptoms and documented bile salt malabsorption have been reported as relating the onset of their problems to an episode of gastroenteritis. Bile salt chelating therapy has been used in this particular situation with apparent success, although it is generally an unrewarding form of treatment for IBS in general. Food intolerance has also been reported as being important in some patients with IBS, but there is little evidence that an immune mediated process due to increased antigen exposure is involved. However, abnormal bacterial fermentation, a process capable of elaborating toxic short chain fatty acids, has been suggested as a possible cause of food intolerance, and may follow both gastroenteritis and antibiotic usage.

Psychosocial factors are known to be important in IBS. Several studies, including that of Neal and colleagues,² which specifically relate to PD-IBS have shown that the risk of developing persistent symptoms following dysentery is related to the presence of psychopathology. Although gastroenteritis may lead to physiological changes that predispose to IBS, there is evidence that an adverse psychosocial milieu is necessary for the condition to fully develop.⁶ This is perhaps not surprising as it is now well recognised that stress can affect the immune, and hence the inflammatory response.⁷ Similarly, stress may increase intestinal permeability, an observation that may be particularly relevant as increased gut permeability has been demonstrated in some patients with PD-IBS.⁸

It would seem reasonable to assume that whatever the triggering factor, an inherited predisposition for IBS might be necessary. This is suggested by the observation that IBS tends to cluster within families, although this could also be explained by environmental factors and indeed, similarities in health related behaviour have been observed between close relatives of those with IBS. Nevertheless, twin studies have shown an increased prevalence of IBS in mono compared with dizygotic twins,^9,10 which might support a genetic background, but a study involving mono and dizygotic twins separated at birth would be required in order to reach a firm conclusion. Laboratory evidence also provides some support for the concept that inheritance is an important factor in the development of IBS. Studies on cytokines, which are known to be involved in the modulation of intestinal inflammation, have shown that mice lacking the interleukin 10 gene develop a spontaneous form of chronic enterocolitis, and that patients with ulcerative colitis are more likely to have genotypes associated with a lower production of interleukin 10. Similarly, a significantly reduced prevalence of the “high producer” gene for interleukin 10 has been reported in a group of unselected patients with IBS.¹¹

It is almost 40 years since Chaudhary and Truelove wrote their classic paper identifying the PD-IBS subgroup. We now know that female sex, younger age, prolonged duration of the initial illness, and psychological comorbidity appear to be important risk factors, and that sufferers usually have the diarrhoea predominant form of the condition. However, there is still much to learn, and emerging technologies will undoubtedly aid this process.