Patients with irritable bowel syndrome or constipation have an increased risk for ischaemic colitis

Abstract

To estimate the relative risk for ischaemic colitis in patients with and without irritable bowel syndrome or constipation, and to evaluate the role of irritable bowel syndrome and constipation as confounders in the relationship between commonly used gastrointestinal medications and ischaemic colitis. Patient cohorts were identified with the use of longitudinal MarketScan research databases from 1 January 1999 to 31 December 2002. Patients in each study cohort were matched 1:1 with comparable control patients using a propensity score. A Cox proportional hazards models were used to estimate relative risk for ischaemic colitis. The relative risk for ischaemic colitis was 3.17 and 2.78 times higher for patients with irritable bowel syndrome and constipation, respectively, than for those without these disorders. Patients who were taking an antispasmodic, a proton pump inhibitor, or an H2-antagonist were at increased risk for ischaemic colitis [relative risk with 95% CI 2.73 (1.41-5.39), 2.00 (1.05-3.79), 2.75 (1.22-6.17) respectively]; however, when these results were adjusted for irritable bowel syndrome or constipation, the relative risks were attenuated and no longer statistically significant. Patients with irritable bowel syndrome or constipation demonstrated a two- to threefold increased risk for ischaemic colitis. Moreover, irritable bowel syndrome and constipation strongly confounded the relationship between gastrointestinal drug use and the risk for ischaemic colitis, suggesting that etiologic studies of ischaemic colitis risk must account for the presence of irritable bowel syndrome or constipation.