To Örebro University

OBJECTIVES: To determine the risk of inflammatory bowel disease (IBD) in patients with celiac disease (CeD) (and vice versa) compared to general-population comparators.

METHODS: Using Swedish histopathology and healthcare register data, we identified 48,551 patients with CeD and 83,529 with IBD diagnosed in 1969-2016. Each patient was compared to age- and sex-matched general-population comparators (CeD: n=240,136; IBD: n=408,195). Cox regression estimated hazard ratios (HRs) for IBD in CeD patients and vice versa. Our main analyses were limited to events beyond the first year of follow-up to reduce potential surveillance bias.

RESULTS: During follow-up, 784 (1.6%) CeD patients were diagnosed with IBD compared to 1015 (0.4%) matched comparators. In CeD patients the HR for IBD was 3.91 (95%CI 3.56-4.31), with largely similar HRs for Crohn's disease (4.36; 3.72-5.11) and ulcerative colitis (3.40; 3.00-3.85). During follow-up, 644 (0.8%) IBD patients and 597 (0.1%) comparators were diagnosed with CeD. The HR for CeD in IBD patients was 5.49 (95%CI 4.90-6.16), with the highest risk estimates seen in ulcerative colitis (HR=6.99; 6.07-8.05), the HR for Crohn's disease was 3.31 (2.69-4.06).In patients with CeD and IBD the diagnostic interval was usually <1 year; however, HRs of 3-4 were seen even after 10 years of follow-up. During 20 years of follow-up, 2.5% of CeD patients developed incident IBD and 1.3% of IBD patients developed CeD.

CONCLUSIONS: The bidirectional association between CeD diagnosis and IBD warrants attention in the initial assessment and follow-up of these conditions. Their co-occurrence, independent of temporal sequence, suggests shared etiology.