Differential impact of recent heavy drinking on first and recurrent acute pancreatitis.

Abstract

Background: While alcohol is known to sensitize the pancreas to acute injury, the role of short-term episodic drinking in regular drinkers is unknown.

Methods: We conducted a case-crossover study to (1) determine the hazardous period of drinking prior to a first episode of acute pancreatitis (FAP) or recurrent acute pancreatitis (RAP) and (2) evaluate the dose-response association between short-term drinking and FAP/RAP. Patients hospitalized for FAP/RAP with an AUDIT-C score of ≥3 were enrolled. Recent and lifetime drinking history were collected through interviews. Drinking prior to the index pancreatitis attack was compared to that of an asymptomatic control period. Conditional logistic regression quantified the association of heavy drinking and FAP/RAP.

Results: Of 141 patients who completed a short-term drinking questionnaire, 77 had RAP, and 64 experienced FAP. We found that both FAP and RAP patients drank at moderate-to-heavy levels regularly, with modest day-to-day variation (intraclass correlation of drinks/day 67%-82%). Alcohol consumption increased 2 days preceding the onset of the index pancreatitis attack as compared to the week prior. Stratifying by prior AP history, heavy drinking in the hazard period was associated with RAP (OR = 3.79, 95% confidence interval [CI] 1.57-9.12). Each drink was associated with 1.22-fold (95%CI 1.10-1.35) increased odds of RAP. Short-term heavy drinking was not associated with a FAP (OR = 1.06, 95%CI 0.43-2.57).

Conclusion: In summary, we found that patients with a prior history of AP face a higher risk of RAP due to excess drinking. Drinking intensity did not increase prior to a FAP, which may have been triggered by other cofactors warranting further examination.