Abstract
BACKGROUND AND AIM: The Global Leadership Initiative on Malnutrition (GLIM) has proposed criteria for the diagnosis of malnutrition. No studies validated the GLIM criteria in acute pancreatitis (AP). The present study aimed to validate the predictive capacity of GLIM criteria for adverse outcomes in AP patients. PATIENTS AND METHODS: Clinical data of 269 patients with AP were analyzed retrospectively. The Nutritional Risk Screening 2002 (NRS2002) was chosen as the screening tool. Multivariate logistic regression analyses evaluated the adverse clinical outcomes in malnourished patients. RESULTS: Overall, 160 patients (59.5%) were at nutritional risk and 38 (14.1%) were malnourished. Reduced muscle mass/ low body mass index + inflammation combinations contributed most to malnutrition overall and in each subgroup. The malnourished group had lower hemoglobin, neutrophils, albumin, total cholesterol, and triglycerides than the well-nourished group. The malnourished group had higher hospitalization costs (CNY, 11319.34 vs 9258.22, p <0.001) and more local complications (34.2% vs 14.7%, p =0.009) than the well-nourished group. There was an interaction between malnutrition and overweight/obesity on local complications (p for interaction = 0.023). Multivariate logistic regression showed malnutrition was significantly associated with local complications (OR 12.2, 95% CI: 2.51-59.37), infectious complications (OR 9.95, 95% CI: 1.25-79.44) and composite adverse outcome (OR 4.78, 95% CI: 1.05-21.73) in the overweight/obesity subgroup. There was no association between malnutrition and the rate of various adverse outcomes in the non-overweight/obesity subgroup. Additionally, we observed an association between malnutrition and composite adverse outcome (OR 6.75, 95% CI: 1.49-30.68) in patients <70 years only in females. CONCLUSION: Malnourished AP patients were more likely to have adverse outcomes than well-nourished patients. Malnutrition was associated with various adverse outcomes only in the overweight/obesity subgroups.