Dietary factors are believed to potentially influence the risk of pancreatitis. Here, we systematically investigated the causal relationships between dietary habits and pancreatitis by using two-sample Mendelian randomization (MR). Large-scale genome-wide association study (GWAS) summary statistics for dietary habits were obtained from the UK Biobank. GWAS data for acute pancreatitis (AP), chronic pancreatitis (CP), alcohol-induced AP (AAP) and alcohol-induced CP (ACP) were from the FinnGen consortium. We performed univariable and multivariable MR analyses to evaluate the causal association between dietary habits and pancreatitis. Genetically driven alcohol drinking was associated with increased odds of AP, CP, AAP and ACP (all with p < 0.05). Genetic predisposition to higher dried fruit intake was associated with reduced risk of AP (OR = 0.280, p = 1.909 × 10⁻⁵) and CP (OR = 0.361, p = 0.009), while genetic predisposition to fresh fruit intake was associated with reduced risk of AP (OR = 0.448, p = 0.034) and ACP (OR = 0.262, p = 0.045). Genetically predicted higher consumption of pork (OR = 5.618, p = 0.022) or processed meat (OR = 2.771, p = 0.007) had a significant causal association with AP, and genetically predicted higher processed meat intake increased the risk of CP (OR = 2.463, p = 0.043). Our MR study showed that fruit intake may be protective against pancreatitis, whereas dietary intake of processed meat has potential adverse impacts. These findings may inform prevention strategies and interventions directed toward dietary habits and pancreatitis.

Chronic pancreatitis (CP) is a fibroinflammatory disorder of the pancreas. Our understanding of CP pathogenesis is partly limited by the incomplete characterization of pancreatic cell types. Here, we performed single-cell RNA sequencing on 3825 cells from the pancreas of one control mouse and mice with caerulein-induced CP. An analysis of the single-cell transcriptomes revealed 16 unique clusters and cell type-specific gene expression patterns in the mouse pancreas. Sub-clustering of the pancreatic mesenchymal cells from the control mouse revealed four clusters of cells with specific gene expression profiles (combinatorial expressions of Smoc2, Cxcl14, Tnfaip6, and Fn1). We observed that immune cells in the pancreas of the CP mice were abundant and diverse in cellular type. Compared to the control, 547 upregulated genes (including Mmp7, Ttr, Rgs5, Adh1, and Cldn2) and 257 downregulated genes were identified in ductal cells from the CP group. The elevated expression levels of MMP7 and TTR were further verified in the pancreatic ducts of CP patients. This study provides a preliminary description of the single-cell transcriptome profiles of mouse pancreata and accurately demonstrates the characteristics of pancreatic ductal cells in CP. The findings provide insight into novel disease-specific biomarkers and potential therapeutic targets of CP.

Background Chronic pancreatitis (CP) is a progressive fibroinflammatory disorder with persistent immune activation and limited therapeutic options. While physical activity (PA) benefits many chronic diseases, it is often presumed neutral or potentially harmful in CP. Objective To assess whether PA protects against CP and defines the underlying mechanisms. Design We analysed the association between PA and CP risk in the UK Biobank cohort (>500 000 participants) and validated findings in an independent clinical cohort. In mice, experimental CP was induced and the effects of exercise interventions on pancreatic injury, fibrosis and immune responses were evaluated via histopathology, immunohistochemistry, flow cytometry, bulk and single-cell RNA-sequencing and proteomics. Results In the UK Biobank, regular PA was independently associated with a lower risk of CP. This association was consistent across alcohol intake strata and disease subtypes. Consistently, physically active patients with CP exhibited milder clinical manifestations. In mice, exercise interventions, including both preconditioning and postdisease initiation, attenuated pancreatic injury, fibrosis and ferroptosis, with resistance exercise providing greater protection. Mechanistically, skeletal muscle-derived extracellular vesicles (EVs) induced by PA accumulated within inflamed pancreata and dampened mitochondrial DNA-driven innate immune activation while promoting inflammation-resolving states, at least in part through modulation of myeloid stimulator of interferon genes (STING) signalling. Importantly, inhibition of EV release partially attenuates these protective effects. Proteomic profiling identified PRDX6 as a muscle-derived vesicular factor that inhibits ferroptosis and, by binding to the zinc-thumb motif of cyclic GMP-AMP synthase, contributes to suppression of STING activation and inflammatory damage. Conclusion PA restrains CP progression by reprogramming pancreatic immune responses and ferroptosis pathways.