Search in the Abstract Database

Abstracts Search 2017

* = Presenting author

P720 Use of proton pump inhibitors associated with a markedly increased risk of microscopic colitis

Bonderup O.*1, Lauge Nielsen G.2, Dall M.3, Pottegård A.3, Hallas J.3

1University Research Clinic for Innovative Patient Pathways, Aarhus University, Denmark, Diagnostic Centre, Section of Gastroenterology, Regional Hospital Silkeborg, Silkeborg, Denmark 2Department of Clinical Medicine, The Faculty of Medicine, Aalborg University, Denmark, Medical Department, Hospital Himmerland, Aalborg, Denmark 3Clinical Pharmacology and Pharmacy, University of Southern Denmark, and Clinical Pharmacology, Odense University Hospital, Odense, Denmark


Microscopic colitis (MC) is a chronic inflammatory bowel disease with unknown etiology. Proton pump inhibitor (PPI) use has been consistently linked to an increased risk of MC. The specific role of the different classes of PPI is unknown. Based on the nationwide Danish registries we explored the association between MC and use of PPI and in addition the effect of different classes of PPIs and dose-dependency. Secondly we explored the effect of concomitant use of PPI and NSAID.


In a ten-year period (January 2004 to December 2013) we identified 10652 patients with a first-time recorded diagnosis of MC including 6254 (59%) with collagenous colitis (CC) and 4398 (41%) with lymphocytic colitis (LC). MC cases were identified via the Danish Pathology Registry. For each MC case, we sampled 10 population controls. Information on PPI use was obtained from the Danish Prescription Register. The analysis conformed to a case-control design, estimating the odds ratio (OR) for the association between PPI use and the risk MC using conditional logistic regression, while adjusting for potential confounders.


We found strong associations between current PPI use and both CC (adjusted OR [aOR] 8.75; 95% CI: 8.12–9.43) and LC (aOR 5.03; 95% CI: 4.61–5.49). This association was seen across all individual PPIs, however, the strongest association was seen with current use of lansoprazole, for both CC (aOR 20.10; 95% CI: 18.05–22.37) and LC (aOR 7.70; 95% CI: 6.79–8.73). When considering timing, ORs were highest for current use of PPI, while the association quickly diminished when considering recent or past exposure. When analyzing the prescribed strength of PPI, we found no apparent dose-response pattern. The use of NSAID was associated with a modest increased risk of CC (aOR 1.70; 95% CI: 1.55–1.87). However, the combination of NSAID and PPI and was associated with a higher risk for CC (aOR 9.72; 95% CI: 8.65–10.92) compared to the use of one of the drugs alone.


In a large comprehensive case-control study based on nationwide Danish registries we found an increased risk of MC associated with the use of PPI and especially an increased risk with the use of lanzoprazole. The combination of NSAID and PPI and was associated with a high risk for CC.