Backgrounds: With the change in dietary patterns, the number of patients suffering from hypertriglyceridemia-induced acute pancreatitis (HTG-AP) has been increasing year by year. HTG-AP patients have a higher likelihood of progressing to severe conditions and exhibit a markedly increased risk of recurrence compared to other etiologies. At present, effective prediction models for the recurrence of HTG-AP need to be established and improved urgently, which is of great significance in improving the prognosis of patients.
Methods: A total of 429 HTG-AP patients were included in this study. Variables were selected by LASSO regression to construct a Logistic regression prediction model, and the performance of the model was verified by ROC curve, calibration curve and decision curve analysis (DCA).
Results: Elevated Thyroid-stimulating hormone (TSH) and triglyceride (TG) levels, Female and Age >33 years were identified as independent risk factors (P < 0.05). A further dietary retrospective study found that the intake of fat and carbohydrate in the recurrent HTG-AP (RAP) group was significantly higher than that in the primary HTG-AP (OAP) group (P < 0.01), while the salt intake was much lower.
Conclusion: Elevated TSH was likely to be an independent risk factor for HTG-AP recurrence. In addition, the new nomogram model based on TSH level shows high accuracy in predicting HTG-AP recurrence, which is expected to provide certain help for clinical prevention and treatments.
Keywords: Hypertriglyceridemia; Prediction model; Recurrent acute pancreatitis; TSH.
Copyright © 2025 The Authors. Published by Elsevier B.V. All rights reserved.