26 - Predictors of Diabetes Mellitus Following Admission for Acute Pancreatitis: Analysis From a Prospective Observational Cohort

Tuesday, October 25, 2022

9:40 AM – 9:50 AM ET

Location: Hall C2

Presenting Author(s)

Aneesa Chowdhury, MD

University of Southern California
Los Angeles, CA

PDF of Slides View Slides

Aneesa Chowdhury, MD¹, Niwen Kong, MD¹, Jin Sun Kim, MD¹, Brent Hiramoto, MD², Selena Zhou, MD¹, Ira A. Shulman, MD¹, Teya Nastaskin, MD¹, James L. Buxbaum, MD¹
¹University of Southern California, Los Angeles, CA; ²Brigham and Women's Hospital, Boston, MA

Introduction: Diabetes mellitus (DM) has become increasingly recognized as a potential sequela of acute pancreatitis (AP). However, specific risk factors for the development of DM after an episode of AP are incompletely defined. We aim to characterize the clinical and biochemical factors associated with the subsequent development of DM.

Methods: We identified patients hospitalized for acute pancreatitis between January 2015 and March 2021 as part of an ongoing prospective observational cohort study. We included patients who had an episode of acute pancreatitis without pre-existing DM to evaluate for our primary outcome defined as the subsequent development of DM (using A1c >6.5) following index admission for acute pancreatitis. At least 9 months of follow up were available following hospitalization. Information on demographics, medical history, biochemical data, severity of the pancreatitis episode (Revised Atlanta Classification), and imaging were obtained for analysis. Logistic regression was used for comparative analysis.

Results: A total of 723 unique patients who were hospitalized for acute pancreatitis without pre-existing DM were analyzed. Within the cohort, 360 (50%) were female with a mean age of 43 (± 14) years old. A majority of this pancreatitis population was Hispanic (n=578, 80%), consistent with the overall population demographics of our hospital (63.4% of all admitted patients in our hospital were Hispanic between 2019-2020). Most (n=610, 84.4%) of the cases of pancreatitis were mild.

The development of DM was identified in 33 (4.6%) patients within the cohort. Strong predictors for subsequent DM included episodes of moderate to severe pancreatitis [OR 2.94 (1.34-6.42)], history of smoking [2.17 (1.02-4.63)] and cirrhosis [3.23 (1.02-4.63)]. Etiology of pancreatitis was not shown to be a significant predictor in multivariate analysis.

Discussion: Moderate to severe pancreatitis, smoking, and cirrhosis were found to predict the subsequent development of DM after AP. This suggests that a combination of local injury, toxin exposure, and systemic disease may contribute to the endocrine dysfunction acute pancreatitis patients may develop.

	Total Cohort (n=723); n (%)	Subsequent DM Cohort (n = 33); n (%)	Univariate Analysis Odds Ratio (95% CI)	Multivariate Analysis Odds Ratio (95% CI)
Female, n (%)	360 (50.0)	10 (30.3)	0.42 (0.20-0.90)
Age > 65 years old	47 (6.5)	3 (9.1)	1.47 (0.43-5.00)
Mean, years (SD)	43 (14)	43 (14)
Hispanic Ethnicity	578 (80.0)	27 (81.8)	1.14 (0.46-2.80)
Severity of Pancreatitis
Mild	610 (84.4)	22 (66.7)	0.35 (0.16-0.74)
Mod/Severe	113 (15.6)	11 (33.3)	2.88 (1.36-6.12)	2.94 (1.34-6.42)
Local Complications	48 (12.5)	6 (18.2)	2.32 (0.81-6.67)
Etiology of Pancreatitis
Alcohol	175 (24.2)	12 (36.4)	2.29 (1.01-5.22)	1.39 (0.56-3.45)
Gallstone	386 (53.4)	12 (36.4)	1
Other	162 (22.4)	9 (27.2)	1.83 (0.75-4.43)
Smoking	186 (26.7)	15 (45.5)	2.53 (1.24-5.17)	2.17 (1.02-4.63)
Cirrhosis	41 (5.7)	6 (18.2)	4.16 (1.61-10.73)	3.23 (1.02-4.63)
CHF	18 (2.5)	1 (3.0)	1.24 (0.16-9.59)
CKD3	27 (3.7)	1 (3.0)	0.80 (0.10-6.07)

Table: Title: Table 1. Predictors of the development of subsequent diabetes following index admission for acute pancreatitis

Footnote: *Variables included in the multivariate model: mod/severe pancreatitis, etiology of pancreatitis, cirrhosis, smoking

Disclosures:

Aneesa Chowdhury indicated no relevant financial relationships.

Niwen Kong indicated no relevant financial relationships.

Jin Sun Kim indicated no relevant financial relationships.

Brent Hiramoto indicated no relevant financial relationships.

Selena Zhou indicated no relevant financial relationships.

Ira Shulman indicated no relevant financial relationships.

Teya Nastaskin indicated no relevant financial relationships.

James Buxbaum: Boston Scientific – Consultant.

Aneesa Chowdhury, MD¹, Niwen Kong, MD¹, Jin Sun Kim, MD¹, Brent Hiramoto, MD², Selena Zhou, MD¹, Ira A. Shulman, MD¹, Teya Nastaskin, MD¹, James L. Buxbaum, MD¹, 26, Predictors of Diabetes Mellitus Following Admission for Acute Pancreatitis: Analysis From a Prospective Observational Cohort, ACG 2022 Annual Scientific Meeting Abstracts. Charlotte, NC: American College of Gastroenterology.