Dig Dis Sci. 2020 Jul 2. doi: 10.1007/s10620-020-06439-4. Online ahead of print.

Anand Kumar ^1 2, Tatiana Teslova ², Erin Taub ³, Joshua D Miller ³, Dana J Lukin ^4 5

Author information

• ¹Division of Gastroenterology, Jill Roberts Center for Inflammatory Bowel Disease, Weill Cornell Medical College, 1315 York Avenue, Mezzanine SM1A15, New York, NY, 10021, USA.
• ²Division of Gastroenterology, Department of Medicine, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY, USA.
• ³Division of Endocrinology and Metabolism, Department of Medicine, Renaissance School of Medicine at Stony Brook, Stony Brook, NY, USA.
• ⁴Division of Gastroenterology, Jill Roberts Center for Inflammatory Bowel Disease, Weill Cornell Medical College, 1315 York Avenue, Mezzanine SM1A15, New York, NY, 10021, USA. djl9010@med.cornell.edu.
• ⁵Division of Gastroenterology, Department of Medicine, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY, USA. djl9010@med.cornell.edu.

Abstract

Background: Diabetes mellitus (DM) and inflammatory bowel diseases (IBD) are chronic systemic illnesses associated with chronic inflammation, dysbiosis, impaired immune function, and infection risk. The impact of DM in modifying disease activity in patients with IBD remains largely unknown.

Aim: To investigate the impact of DM on IBD-related disease outcomes, mortality, and infections in patients with IBD.

Methods: We performed a longitudinal cohort analysis. Using a large institutional database, patients with concurrent IBD and DM (IBD-DM), and IBD without DM (IBD cohort), were identified and followed longitudinally to evaluate for primary (IBD-related) and secondary (mortality and infections) outcomes. Cox proportional hazards models were used to determine the independent effect of DM on each outcome, adjusting for confounding effects of covariates.

Results: A total of 901 and 1584 patients were included in the IBD-DM and DM cohorts. Compared with IBD, IBD-DM had significantly higher risk of IBD-related hospitalization [adjusted hazard ratio (HR) 1.97, 95% confidence interval (1.71-2.28)], disease flare [HR 2.05 (1.75-2.39)], and complication [HR 1.54 (1.29-1.85)]. No significant difference was observed in the incidence of IBD-related surgery. All-cause mortality, sepsis, Clostridioides difficile infection (CDI), pneumonia, urinary tract infection, and skin infection were also more frequent in the IBD-DM than the IBD cohort (all p ≤ 0.05). Subgroup analysis of Crohn's disease (CD) and ulcerative colitis patients showed similar associations, except with an additional risk of surgery and no association with CDI in the CD-DM cohort.

Conclusion: Comorbid diabetes in patients with IBD is a predictor of poor disease-related and infectious outcomes.