Triglyceride-Related Indices and Risk of Acute Pancreatitis: A Prospective Study in UK Biobank.

Abstract

Triglyceride-related indices, including triglyceride-glucose (TyG), TyG related to body mass index (TyG-BMI), TyG related to waist circumference (TyG-WC), TyG related to waist-to-height ratio (TyG-WHtR), atherogenic index of plasma (AIP), cardiometabolic index (CMI), visceral adiposity index (VAI), and lipid accumulation product (LAP), are simple surrogate indicators of insulin resistance. Nevertheless, few studies have explored the association between triglyceride-related indices and acute pancreatitis (AP). A total of 387,275 participants from the UK Biobank were included. Cox regression models provided hazard ratios (HRs) with 95% confidence intervals (CIs) for the association between triglyceride-related indices and the risk of AP. Restricted cubic splines (RCS) and receiver operating characteristic (ROC) were employed to evaluate and visualize the relationship. These triglyceride-related indices were significantly associated with an increased risk of AP. Compared with the lowest quartile, the HRs (95% CIs) of AP in the highest quartile were 1.52 (1.33, 1.74), 1.54 (1.34, 1.77), 2.32 (2.02, 2.65), 2.83 (2.43, 3.30), 2.86 (2.46, 3.32), 2.16 (1.88, 2.49), 1.95 (1.71, 2.22), and 2.49 (2.15, 2.87), respectively, for the TyG, AIP, TyG-BMI, TyG-WC, TyG-WHtR, CMI, VAI, and LAP index. RCS analyses showed that the above relationships except for AIP were nonlinear (P for nonlinear < 0.05). ROC curves showed that TyG-WHtR was the best predictor for AP. Elevated triglyceride-related indices were positively related to AP onset. Among them, the TyG-WHtR index displayed the strongest predictive power for the onset of AP, which was expected to become a more effective metric for identifying populations at early risk of AP.