Elevated peripheral blood CGRP correlates with acute pancreatitis severity and enables early prediction of severe acute pancreatitis via a multimarker decision tree model

Abstract Background Acute pancreatitis (AP) is a highly heterogeneous acute abdominal disorder, with approximately 20% of patients progressing to severe acute pancreatitis (SAP). However, there remains a lack of rapid and accurate early predictive markers for SAP. This study aimed to explore the correlation between peripheral blood calcitonin gene-related peptide (CGRP) levels and AP severity, and to develop an early prediction model for SAP to support clinical risk stratification and targeted interventions. Methods A total of 158 adult AP patients admitted within 24 h of symptom onset to The Affiliated Hospital of Putian University (2021–2024) were enrolled. Patients were categorized into the SAP group (n = 59) and non-SAP group (n = 99, including mild AP [MAP] and moderately severe AP [MSAP]). Peripheral blood CGRP concentrations were measured via ELISA. Clinical data were collected, and univariate logistic regression (P < 0.05) and multivariate logistic regression were employed to screen predictive factors. A decision tree model was constructed and validated using R language. Results The CGRP level in the SAP group (97.0 pg/mL) was significantly higher than that in the non-SAP group (75.0 pg/mL, P < 0.001). SAP patients also exhibited elevated white blood cell (WBC) count (13.36 ± 4.72 × 10⁹/L), C-reactive protein (CRP, 139.32 mg/L), procalcitonin (PCT, 7.67 ng/mL), interleukin-6 (IL-6, 85.00 pg/mL), and reduced albumin (36.0 g/L) (all P < 0.05). Univariate logistic regression identified five core predictors: CGRP (OR = 1.025, 95% CI 1.007–1.043, P = 0.005), WBC count (OR = 1.143, 95% CI 1.007–1.297, P = 0.039), CRP (OR = 1.005, 95% CI 1.000–1.010, P = 0.046), PCT (OR = 1.028, 95% CI 1.004–1.052, P = 0.020), and IL-6 (OR = 1.013, 95% CI 1.001–1.025, P = 0.029). Multivariate logistic regression suggests CGRP as an independent risk factor for SAP (OR = 1.025, 95% CI 1.009–1.041, P = 0.002). The decision tree model integrating these factors performed excellently (AUC = 0.878, overall accuracy = 92.4%, specificity = 96.0%), with CGRP ranking highest in variable importance. Conclusion Peripheral blood CGRP levels differ significantly among AP patients with varying disease severity and may serve as an independent predictive factor for SAP. The combined multimarker decision tree model enables effective early prediction of SAP, providing valuable references for AP risk stratification and optimization of clinical interventions.