Risk of development of acute pancreatitis with pre-existing diabetes: a meta-analysis.
ABSTRACT It is well established that acute pancreatitis (AP) often causes diabetes mellitus. However, whether pre-existing diabetes is associated with the development of AP remains unknown. To clarify the association of pre-existing diabetes and the development of AP, we carried out a meta-analysis of observational studies.
A computerized literature search was performed in MEDLINE (from 1 January 1966) and EMBASE (from 1 January 1974), through 31 January 2012. We also searched the reference lists of relevant articles. Summary relative risks with their corresponding 95% confidence intervals (CIs) were calculated using a random-effects model. Between-study heterogeneity was assessed using Cochran's Q statistic and the I 2.
A total of seven articles (10 523 incident cases of AP) were included in this meta-analysis. Analysis of seven studies indicated that, compared with nondiabetic individuals, diabetic individuals had a 92% increased risk of development of AP (95% CI 1.50-2.47). There was significant evidence of heterogeneity among these studies (P heterogeneity<0.001, I 2=93.0%). These increased risks were independent of alcohol use, gallstones, and hyperlipidemia.
Although the current evidence supports a positive link between pre-existing diabetes and an increased risk of development of AP, additional studies, with a perfect design, are required before definitive conclusions can be drawn.
- [Show abstract] [Hide abstract]
ABSTRACT: To determine the relative risk of pancreatitis in diabetes, and to establish whether diabetes-related as well as recognized risk factors contribute. We studied 1426 participants [mean (sd) age 62.1 (13.3) years, 49.6% male, 90.9% Type 2 diabetes, median (interquartile range) diabetes duration 4.0 (1.0-10.0) years] from the community-based Fremantle Diabetes Study Phase I and 5663 matched residents without diabetes from the same geographical area. Pancreatitis hospitalizations between 1982 and 2010 were ascertained using validated data linkage. For Fremantle Diabetes Study Phase I participants, chart review provided data on the likely causes of pancreatitis. A total of 21 Fremantle Diabetes Study Phase I participants (1.5%) were hospitalized for pancreatitis before study entry vs 29 (0.5%) of contemporaneous residents without diabetes. During a mean (sd) of 12.1 (5.4) years of follow-up from entry, 22 (1.6%) Fremantle Diabetes Study Phase I participants were hospitalized for a first-ever episode of pancreatitis on 37 occasions (1.31/1000 person-years) compared with 58 (1.0%) residents without diabetes on 81 occasions during a mean (sd) 13.6 (4.8) years (0.75/1000 person-years). The age- and sex-adjusted hazard ratio (95% CI) for first-ever pancreatitis hospitalization in Fremantle Diabetes Study Phase I participants was 1.73 (1.06-2.83; P=0.029). Chart review of 17 of the 22 Fremantle Diabetes Study Phase I participants (77%) with incident pancreatitis and available case notes revealed that four (24%) presented without objective evidence of pancreatitis, seven (41%) presented with cholelithiasis, three (18%) with excessive alcohol consumption, two (12%) as a complication of elective endoscopic retrograde cholangiopancreatography, and one (6%) with hypertriglyceridaemia. Consistent with previously published data, the risk of pancreatitis was higher in community-dwelling Fremantle Diabetes Study Phase I participants but conventional precipitants accounted for confirmed cases. These data question whether diabetes-specific risk factors cause or contribute to pancreatitis. This article is protected by copyright. All rights reserved.Diabetic Medicine 03/2014; · 3.24 Impact Factor