Fourteen studies met the inclusion criteria (n=4,204 patients tested for celiac; n=2,278 patients met IBS diagnostic criteria; sample size ranged from 100 to 1,200). Seven studies were case series and seven studies were case control studies. Only four of the included studies reported recruiting consecutive patients.
The percentage of patients meeting IBS criteria who tested positive for celiac disease ranged from 0% to 18% for IgA-class AGAs(pooled prevalence 4%, 95% CI: 1.7, 7.2%, seven studies, n=1,104), 0% to 11.4% for tTGAs (pooled prevalence 1.63%, 95% CI: 0.7, 3.0%, 13 studies, n=2,021), and 0% to 11.4% for biopsy (pooled prevalence 4.1%, 95% CI: 1.9, 7.0%, seven studies, n=1,464).
In studies comparing IBS diagnosed patients with controls, there was a higher rate of celiac disease in patients diagnosed with IBS when measured using IgA-class AGAs (OR 3.4, 95% CI: 1.62, 7.13, six studies, n=1,002), EMAs or tTGAs (OR 2.94, 95% CI: 1.36, 6.35, seven studies, n=2,978) and biopsy (OR 4.34, 95% CI: 1.78, 10.58, five studies, n=2,750).
Results for subgroup analyses were reported. The authors reported that there was no evidence of funnel plot asymmetry.