Risk of Diabetes Mellitus after First-Attack Acute Pancreatitis: A National Population-Based Study.

Abstract

Population-based data on the risk of diabetes mellitus onset after acute pancreatitis (AP) are lacking. We assessed the incidence of diabetes in AP survivors compared with matched controls. The study cohort, drawn from Taiwan National Health Insurance claims data, included 2,966 first-attack AP patients and 11,864 non-AP general controls individually matched on age and sex, with an AP/non-AP ratio of 1:4. Incidence rate was estimated under Poisson assumption. Relative risks of diabetes were indicated by hazard ratios (HRs) estimated from Cox proportional hazard regression models with a partitioning of time at 3 months to account for proportionality. In the first partition of time (<3 months), the incidences of diabetes were 60.8 and 8.0 per 1,000 person-years in AP and control groups, respectively; representing a covariate-adjusted HR of 5.90 (95% confidence interval (CI) 3.37-10.34). In the second partition (≥3 months), the incidences of diabetes were 22.5 and 6.7 per 1,000 person-years in AP and control groups, respectively (adjusted HR 2.54, 95% CI 2.13-3.04). In the second partition, the risk of diabetes was greater in men than in women (HR 3.21 vs. 1.58, P=0.0004). When the analyses were stratified by severity of AP, the results for mild AP were similar to those for all AP. The risk of diabetes increases by twofold after AP; therefore, a long-term screening is necessary to evaluate diabetes after an attack regardless of severity. Further research should be conducted to develop cost-effective follow-up strategies, and to elucidate the underlying mechanisms of the relationship between diabetes and AP.