BACKGROUND:

Irritable bowel syndrome has a high prevalence. Consensus diagnostic criteria (ROME II) based on symptoms have been established to aid diagnosis. Although coeliac disease can be misdiagnosed as irritable bowel syndrome, no prospective study has been published in which patients with this disorder are investigated for coeliac disease. We aimed to assess the association of coeliac disease with irritable bowel syndrome in patients fulfilling ROME II criteria.

METHODS:

We undertook a case-control study at a university hospital. 300 consecutive new patients who fulfilled Rome II criteria for irritable bowel syndrome, and 300 healthy controls (age and sex matched) were investigated for coeliac disease by analysis of serum IgA antigliadin, IgG antigliadin, and endomysial antibodies (EMA). Patients and controls with positive antibody results were offered duodenal biopsy to confirm the possibility of coeliac disease.

FINDINGS:

66 patients with irritable bowel syndrome had positive antibody results, of whom 14 had coeliac disease (11 EMA positive, three EMA negative). Nine patients with positive antibody results were lost to follow-up or refused biopsy (only one EMA-positive patient refused biopsy), and 43 had normal duodenal mucosa. Two controls, both of whom were EMA positive, had coeliac disease. Compared with matched controls, irritable bowel syndrome was significantly associated with coeliac disease (p=0.004, odds ratio=7.0 [95% CI 1.7-28.0]).

INTERPRETATION:

Patients with irritable bowel syndrome referred to secondary care should be investigated routinely for coeliac disease. With only EMA, three of 14 cases would have been missed.