Body mass index and the risk of acute pancreatitis by etiology: A prospective analysis of Korean National Screening Cohort

BACKGROUND: It is unclear whether obesity increases the incidence of acute pancreatitis (AP) in the general population. Further, no study has prospectively examined the associations of the risk of AP by etiology with measured body mass index (BMI) values. METHODS: A total of 512 928 Korean participants in routine health examinations during 2002-2003 were followed up until 2013 via linkage to national hospital discharge records to assess AP incidence. Multivariable-adjusted hazard ratios were calculated using BMI measurements. RESULTS: During 10.5 mean years of follow-up, 1656 persons developed AP (337 gallstone related and 1319 non-gallstone related). Nonlinear associations were found: U-curves for total and non-gallstone-related AP and a reverse L-curve for gallstone-related AP. Each 5 kg/m2 increment in BMI increased gallstone-related AP by 123% (95% confidence interval = 48-234%) and non-gallstone-related AP by 42% (9-84%) in the range ≥ 25 kg/m2 (Pheterogeneity  = 0.068). Obese persons had a doubled risk of gallstone-related AP compared with normal-weight persons. In the range < 25 kg/m2 , BMI had inverse association with non-gallstone-related AP but no association with gallstone-related AP (Pheterogeneity  < 0.001). In subgroup analyses, for non-gallstone-related AP, hazard ratios per each 5 kg/m2 BMI increment were 0.50 (men), 0.73 (women), 0.46 (alcohol drinkers), 0.69 (alcohol non-drinkers), 0.43 (ever smokers), and 0.73 (never smokers). CONCLUSIONS: Gallstone-related AP and non-gallstone-related AP have different nonlinear associations with BMI. Higher BMI increases the risk of both gallstone-related AP and non-gallstone-related AP but more strongly for gallstone-related AP. For non-gallstone-related AP, in the range < 25 kg/m2 , BMI has inverse associations that were stronger in men, current alcohol drinkers, and ever smokers than in their counterparts.