|Home | About | Journals | Submit | Contact Us | Français|
We read with interest the study by van der Wouden et al (Gut 2007;56:444–5). The authors describe their experience of investigating patients with irritable bowel syndrome (IBS fulfilling the Rome II criteria) for coeliac disease. They tested 152 patients with IBS using IgA endomysial antibodies and total IgA level or duodenal biopsy; 36 patients were subsequently biopsied but there were no cases of coeliac disease diagnosed (prevalence 0%: 95% confidence interval, 0% to 4%). They concluded that the prevalence of coeliac disease in patients with IBS is low and that screening may be ineffective.
We feel cautious about accepting these results as definitive proof of no relation between these two common conditions. Their study has some methodological limitations. First, the referral pattern appears to be significantly different from that seen in the United Kingdom, with approximately 55 patients being assessed for IBS symptoms per year by three physicians. Does this suggest that IBS is not a condition commonly referred to secondary care in the Netherlands? Perhaps primary care physicians have already investigated patients for coeliac disease before referral? Second, assuming the accepted population prevalence for coeliac disease is 1%, then with a sample size of 152 (11 patients were not serologically tested or biopsied) there is only a 21.4% power to detect a threefold difference (assuming a significance level of 0.05). Thus this study is underpowered. Third, endomysial antibody (EMA) negative coeliac disease is well recognised. The use of EMA in isolation may lower the detection sensitivity and with only 22% of their patients (36/163) having a duodenal biopsy, it is possible that some cases of coeliac disease could have been missed. Fourth, importantly, there are several studies examining this issue (table 11),), not all of which were cited by the authors in their report. These international studies suggest that the prevalence of coeliac disease in cohorts of patients with IBS is higher than in the general population.
Furthermore, the association of coeliac disease and IBS symptoms is biologically plausible with many mechanisms being reported—for example, autonomic dysfunction, intussusception, exocrine pancreatic disease, small intestinal ulceration, and associated microscopic colitis. Our own group found an increased prevalence of coeliac disease in patients referred with surgical abdominal pain, notably in those with unexplained or non‐specific abdominal pain.9 The association between IBS and coeliac disease appears to operate in both directions, as patients with coeliac disease (on a gluten‐free diet) may describe IBS symptoms.10 Similarly, abdominal pain, diarrhoea, or constipation were associated with an increased risk of coeliac disease in a large multicentre study.11 All of these symptoms can overlap with IBS.
From a clinical perspective, is investigating IBS patients for coeliac disease a valid approach? It is recognised that the majority of patients with coeliac disease have significant delays in diagnosis and once established on the gluten‐free diet they derive symptomatic benefit. Case finding for coeliac disease in patients with IBS symptoms is both cost‐effective and beneficial in terms of quality of life years gained, even at prevalence values of 1.1–2%.12 We accept that further multicentre studies in both primary and secondary care are required to resolve this debate. In such studies assessment of quality of life and symptom resolution (when individuals are established on a gluten‐free diet) is imperative. Despite the existing evidence, investigating for coeliac disease in patients fulfilling the Rome II criteria is still not widely accepted (although recommended in UK guidelines). Why should this be? Is this a clash of ideological dogma when considering possible mechanisms for IBS?
Conflict of interest: None declared.