The neutrophil to high-density lipoprotein cholesterol ratio is associated with disease severity in hypertriglyceridemic acute pancreatitis Authors: Xiaogang Hu, Jia Xu, Juntao Yang, Menggang Liu Published: 2025 ABSTRACT AbstractBackgroundThe neutrophil to high-density lipoprotein cholesterol ratio (NHR) predicts disease severity in biliary pancreatitis, yet its prognostic role in hypertriglyceridemic acute pancreatitis (HTG-AP) remains unclear.MethodsThis study enrolled 421 eligible patients diagnosed with HTG-AP. Baseline comparisons were made using the Mann-Whitney U Test and Chi-Square Test. Logistic regression analysis was employed to identify risk factors for HTG-AP severity. The predictive value of NHR was assessed using the Integrated Discrimination Improvement (IDI) and the Net Reclassification Index (NRI). Restricted cubic splines (RCS) were used to evaluate nonlinear associations. Subgroup analyses further explored the correlation between NHR and HTG-AP severity.ResultsPatients with moderate/severe HTG-AP exhibited significantly higher NHR levels than those with mild disease (p < 0.001). Elevated NHR correlated with prolonged hospitalization (p = 0.016), higher rates of acute peripancreatic fluid collection (APFC) (p < 0.001) and acute necrotic collection (ANC) (p = 0.008). NHR was an independent risk factor for moderate/severe HTG-AP (p < 0.001) and improved the predictive accuracy of the risk assessment model. NHR led to only a small improvement in AUC within the model, which indicated some limitations in predicting HTG-AP severity. RCS analysis revealed a linear relationship between NHR and HTG-AP severity (p for non-linearity > 0.05).ConclusionNHR is a simple, early biomarker for predicting HTG-AP severity, offering potential utility in guiding timely clinical interventions.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12876-025-04466-x.Keywords: Hypertriglyceridemic acute pancreatitis, Acute pancreatitis, Neutrophil to high-density lipoprotein cholesterol ratio, Disease severity Introduction Acute pancreatitis (AP), a prevalent gastrointestinal disease requiring hospitalization, arises from various etiologies, such as gallstones, alcohol abuse, and hypertriglyceridemia (HTG). Notably, the incidence of hypertriglyceridemic acute pancreatitis (HTG-AP) is rising worldwide [1, 2]. Importantly, HTG-AP presents with greater severity compared to other etiologies, underscoring the urgency to elucidate its pathophysiology and identify prognostic biomarkers [3]. AP severity is classified into mild (MAP), moderately severe (MSAP), and severe (SAP) categories based on the Revised Atlanta Classification [4]. While MAP typically follows a self-limiting course, MSAP and SAP are linked to poorer outcomes, with SAP mortality rates exceeding 20% despite advances in critical care [5]. Early prediction of disease progression is pivotal for optimizing therapeutic strategies and improving clinical results. Emerging biomarkers integrating inflammatory and metabolic pathways have gained attention for risk prediction of AP. Among these, the neutrophil-to-high-density lipoprotein cholesterol ratio (NHR) has demonstrated prognostic value in cardiovascular diseases and metabolic disorders, reflecting its dual role in inflammation and lipid dysregulation [6–9]. Yang et al. demonstrated an independent association between NHR and the severity of acute biliary pancreatitis (ABP) [10]. However, the relationship between NHR and the severity of HTG-AP remains unexplored, despite the distinct lipid-driven pathophysiology of HTG-AP. Therefore, we hypothesized that NHR, which encapsulates both the inflammatory and lipid components central to HTG-AP pathophysiology, could serve as a novel biomarker for predicting severity in this specific patient population. This study aims to investigate whether NHR serves as a biomarker for predicting HTG-AP severity, potentially filling knowledge gaps in early clinical decision-making. Patients and methods Patients This study enrolled patients diagnosed with HTG-AP at Chongqing Liangjiang New Area People’s Hospital from January 2021 to June 2024. Inclusion criteria were: (1) diagnosis of AP based on the Revised Atlanta classification system [4]; (2) serum triglyceride (TG) levels ≥ 11.30 mmol/L, or TG levels between 5.65 and 11.30 mmol/L with chylous serum; (3) exclusion of alternative AP etiologies (e.g., biliary, alcoholic, post Endoscopic Retrograde Cholangiopancreatography‌); (4) age ≧ 18 years. Exclusion criteria were: (1) admission >24 h post-abdominal pain onset; (2) hospitalization stays < 48 h; (3) incomplete clinical data; (4) history of chronic/recurrent pancreatitis; (5) malignancies; (6) pregnancy; (7) severe non-pancreatitis-related comorbidities or chronic inflammatory diseases, such as chronic obstructive pulmonary disease (COPD) or heart failure; (8) Lipid-lowering therapy within 6 months prior to admission. By comprehensively analyzing the patients’ medical records and applying the Revised Atlanta classification, we determined the severity of HTG-AP patients as MAP, MSAP, and SAP. All patients began fasting immediately after admission and received fluid resuscitation therapy. A combination of insulin and the molecular weight heparin calcium was administered to reduce triglycerides, and gabexate mesilate was used to inhibit protease activity. The treatment strategy for all patients was standardized. The effectiveness of fluid resuscitation was assessed by monitoring changes in urine output and heart rate, and these changes reflected improvements in circulatory volume and renal perfusion. Patients with hemodynamic instability or multiple organ dysfunction syndrome were transferred to the intensive care unit (ICU) for treatment. Data collection Demographic, clinical, and laboratory data were extracted from electronic medical records. Blood samples collected within 24 h of admission were analyzed for: complete blood count, lipid profile, C-reactive protein (CRP), pancreatic enzymes (amylase, lipase), hepatic and renal function parameters. Disease severity was categorized as MAP, MSAP, or SAP per the Revised Atlanta Classification [4]. Following the methodology of Yang et al., MSAP and SAP were combined as “Non-MAP” for analysis [10]. The NHR was calculated as: NHR = absolute neutrophil count (×109/L)/HDL-C (mmol/L). Data preprocessing and assessment The data in this study were mixed-type data with only 3.33% missing values, and we used the ‘missForest’ package to perform imputation of the missing values. The optimal NHR threshold for severity prediction was determined via receiver operating characteristic (ROC) curve analysis, maximizing the Youden index (J = Sensitivity + Specificity − 1). Univariable, multivariable, stepwise logistic and Lasso regression analyses were employed to identify potential risk factors for Non-MAP. Continuous variables were dichotomized using clinically validated cutoffs (Supplementary Table 1) before regression analyses. Variables with a p-value of < 0.05 in the univariable analysis were included in the multivariable model. The area under the ROC curve (AUC), Brier score, Delong test, net reclassification improvement (NRI), and comprehensive discrimination improvement (IDI) were used to evaluate the performance of the model. Furthermore, we constructed the calibration curve for the model and plotted the curves for sensitivity, specificity, positive predictive value, and negative predictive value across the full range of thresholds. Non-linear associations between NHR and HTG-AP severity were examined using Restricted Cubic Splines (RCS) with three knots. Subgroup analyses tested interactions between NHR and demographic/clinical variables (age, gender, diabetes, etc.). Statistical analysis Data normality was assessed using the Kolmogorov-Smirnov test. Non-normally distributed continuous variables were reported as median (interquartile range [IQR]) and compared with the Mann-Whitney U Test. Categorical variables were expressed as frequencies (%) and analyzed using Chi-Square or Fisher’s Exact Test. The primary R packages employed in this analysis include missForest, moonBook, reportROC, gtsummary, stepReg, glmnet, ggvenn, rcssci, jstable, rms, predictABEL, and runway. All analyses and visualizations were performed using R software (version 4.4.1, Vienna, Austria), with a two-tailed (P < 0.05) deemed statistically significant. Results Baseline characteristics A total of 421 HTG-AP patients were included (Fig. 1), with 254 classified as mild (MAP) and 167 as moderate/severe (Non-MAP). The median age was 38 years (range: 22–66 years), with no significant age difference between groups. Patients with Non-MAP had longer hospital stays (median [IQR]: 10 [8–13] vs. 6 [5–7] days; P < 0.001) and higher ICU admission rates (27% vs. 0%; P < 0.001) compared to those with MAP. A total of two patients died, both in the Non-MAP group. Non-MAP patients exhibited elevated NHR levels (16.1 [10.4–23.9] vs. 12.2 [8.3–17.0]; P < 0.001), as well as elevated white blood cell counts (WBC), neutrophil counts, monocyte counts, C-reactive protein (CRP) levels, aspartate aminotransferase (AST) levels, total cholesterol (TC) levels, TG levels, low-density lipoprotein cholesterol (LDL-C) levels, amylase levels, and lipase levels, alongside reduced lymphocyte counts, albumin levels, and serum calcium levels (all P < 0.05, Table 1). However, high-density lipoprotein cholesterol (HDL-C) levels were not significantly different between MAP and Non-MAP groups (P = 0.086), which directly supported the value of using NHR instead of the individual components. Study flowchart of patient enrollment Table 1Patient demographic characteristics and clinical dataVariablesTotal (n = 421)MAP (n = 254) Non-MAP (n = 167)P valueDemographics Age (years)39.0 (34.0–48.0)39.0 (34.0–48.0)39.0 (34.5–47.0)0.869 Sex (male,%)346 (82.2%)210 (82.7%)136 (81.4%)0.845 Smoking (n, %)207 (49.2%)122 (48.0%)85 (50.9%)0.634 Drinking (n, %)79 (18.8%)48 (18.9%)31 (18.6%)1.000Comorbidities (n, %) Hypertension64 (15.2%)33 (13.0%)31 (18.6%)0.156 Diabetes124 (29.5%)69 (27.2%)55 (32.9%)0.246 Fatty liver225 (53.4%)133 (52.4%)92 (55.1%)0.653 Other34 (8.1%)21 (8.3%)13 (7.9%)1.000Hospital stay (days)7 (5.0–10.0)6 (5–7)10 (8–13)< 0.001ICU admission (n, %)27 (6.4%)27 (6.4%)0 (0.0%)< 0.001WBC (×109/L)13.2 (10.4–16.0)12.6 (9.9–15.4)14.3 (11.5–17.4)< 0.001Neu (×109/L)10.8 (7.9–13.3)10.0 (7.5–12.6)11.9 (9.2–14.9)< 0.001Lym (×109/L)1.7 (1.2–2.2)1.7 (1.3–2.4)1.5 (1.0–2.1)0.002Mono (×109/L)0.6 (0.4–0.8)0.5 (0.4–0.7)0.6 (0.4–0.9)0.002RDW (%)12.8 (12.3–13.2)12.8 (12.3–13.2)12.8 (12.3–13.2)0.939PLT (×109/L)226.0 (179.0–272.0)223.5 (179.0–266.0)228.0 (180.5–282.5)0.403CRP (mg/L)13.6 (3.7–66.4)8.4 (3.0–37.2)35.2 (5.6–120.5)< 0.001ALT (U/L)28.0 (19.0–42.0)26.5 (18.0–40.0)30.0 (20.0–44.0)0.085AST (U/L)24.0 (19.0–32.0)22.5 (18.0–29.0)26.0 (20.0–34.0)0.001TC (mmol/L)7.2 (5.6–9.2)6.8 (5.6–8.7)8.0 (5.8–9.8)0.003TG (mmol/L)16.9 (11.2–23.0)15.8 (10.5–22.6)18.3 (12.3–23.3)0.026HDL-C (mmol/L)0.8 (0.6–1.0)0.8 (0.6–1.0)0.7 (0.6–1.0)0.070LDL-C (mmol/L)1.1 (0.8–1.8)1.2 (0.9–2.0)1.0 (0.8–1.6)0.002Cr (µmol/L)58.8 (47.0–70.9)59.2 (47.7–71.0)58.0 (47.0–69.0)0.312BUN (mmol/L)4.9 (4.0–5.9)4.8 (4.0–5.7)4.9 (3.9–6.4)0.182Ca (mmol/L)2.3 (2.2–2.4)2.3 (2.2–2.4)2.2 (2.1–2.4)0.001ALB (g/L)43.8 (40.7–46.4)44.3 (41.3–46.7)43.1 (39.6–45.9)0.003AMY (U/L)168.0 (76.0–389.0)141.0 (72.0–284.0)231.0 (93.5–603.0)< 0.001LIP (U/L)280.0 (99.0–616.0)229.0 (80.0–451.0)393.0 (125.5–920.0)< 0.001NHR13.2 (8.7–19.7)12.2 (8.3–17.0)16.1 (10.4–23.9)< 0.001MAP Mild acute pancreatitis, non-MAP Moderate and severe acute biliary pancreatitis, WBC White blood cell, Neu Neutrophil, Lym Lymphocyte, Mono Monocyte, RDW Red cell distribution width, PLT Platelet, CRP C-reactive protein, ALT Alanine aminotransferase, AST Aspartate aminotransferase, TC Total cholesterol, TG Triglyceride, HDL-C High-density lipoprotein cholesterol, LDL-C Low-density lipoprotein cholesterol, Cr Creatini, AMY Amylase, LIP Lipase, BUN Blood urea nitrogen, Ca Calcium, ALB Albumin, NHR Neutrophil high-density lipoprotein cholesterol ratio Patient demographic characteristics and clinical data MAP Mild acute pancreatitis, non-MAP Moderate and severe acute biliary pancreatitis, WBC White blood cell, Neu Neutrophil, Lym Lymphocyte, Mono Monocyte, RDW Red cell distribution width, PLT Platelet, CRP C-reactive protein, ALT Alanine aminotransferase, AST Aspartate aminotransferase, TC Total cholesterol, TG Triglyceride, HDL-C High-density lipoprotein cholesterol, LDL-C Low-density lipoprotein cholesterol, Cr Creatini, AMY Amylase, LIP Lipase, BUN Blood urea nitrogen, Ca Calcium, ALB Albumin, NHR Neutrophil high-density lipoprotein cholesterol ratio Patients with high NHR exhibit more severe conditions ROC Curve analysis determined an optimal NHR cutoff value of 16.66, dividing patients into two groups based on this threshold (Fig. 2). High NHR (≥ 16.66) was associated with a higher proportion of patients with Non-MAP (55.0% vs. 31.3%; P < 0.001), a higher rate acute peripancreatic fluid collection (APFC) (47.0% vs. 29.0%; P < 0.001) and acute necrotic collection (ANC) (14.1% vs. 5.9%; P = 0.008). Patients with high NHR also showed prolonged hospitalization (median [IQR]: 8 [6,–10] vs. 7 [5,–9] days; P = 0.016), elevated WBC counts(15.9 [13.6–18.6] vs. 11.8 [9.2–13.9] ×109/L; P < 0.001) and neutrophil counts, alongside reduced lymphocytes counts and HDL-C levels (Table 2). Although there was no significant difference in mortality between the high NHR group and the low NHR group, there were only two deaths in both groups. This small number of deaths was insufficient to reflect the relationship between NHR and mortality. ROC curve of NHR prediction of disease severity in patients with HTG-AP. The area under the curve (AUC) is 0.620 (P < 0.01), the cutoff value is 16.660 with a sensitivity of 0.736 and a specificity of 0.491. The AUC of NHR is fair but not excellent, indicating that its main value is as an independent factor of non-MAP and an additive to a larger predictive model for the servity of HTG-AP Table 2Comparison of characteristics between high- and low-NHR groupsVariablesNHR < 16.66 (n = 272)NHR > 16.66 (n = 167)P valueDemographics Age (years)40.0 (35.0–48.0)38.0 (33.0–47.0)0.101 Sex (male,%)230 (84.6%)116 (77.9%)0.113 Smoking (n, %)131 (48.2%)76 (51.0%)0.648 Drinking (n, %)60 (22.1%)19 (12.8%)0.027Comorbidities (n, %) Hypertension43 (15.9%)21 (14.1%)0.744 Diabetes82 (30.1%)42 (28.2%)0.757 Fatty liver140 (51.5%)85 (57.0%)0.32 Other19 (7.0%)15 (10.0%)0.356Hospital stay (days)7 (5–9)8 (6–10)0.016ICU admission (n, %)18 (6.6%)9 (6.0%)0.981Non-MAP85 (31.3%)82 (55.0%)< 0.001Complications, n (%) APFC79 (29.0%)70 (47.0%)< 0.001 ANC16 (5.9%)21 (14.1%)0.008 PP5 (1.9%)1 (0.7%)0.430 WON3 (1.1%)4 (2.7%)0.250WBC (×109/L)11.8 (9.2–13.9)15.9 (13.6–18.6)< 0.001Neu (×109/L)9.2 (6.9–11.5)13.4 (11.8–16.0)< 0.001Lym (×109/L)1.7 (1.3–2.4)1.5 (1.0–2.0)0.010Mono (×109/L)0.5 (0.4–0.7)0.7 (0.5–0.9)< 0.001RDW (%)12.8 (12.3–13.2)12.8 (12.4–13.3)0.423PLT (×109/L)220.0 (176.0–269.0)236.0 (192.0–2275.0)0.078CRP (mg/L)12.5 (3.3–71.3)13.6 (3.7–66.4)0.438ALT (U/L)29.0 (19.0–45.5)26.0 (19.0–36.0)0.059AST (U/L)24.0 (19.0–33.5)23.0 (19.0–29.0)0.094TC (mmol/L)7.5 (5.8–9.9)6.8 (5.1–8.2)0.001TG (mmol/L)17.0 (10.9–23.3)16.6 (11.4–21.8)0.855HDL-C (mmol/L)0.9 (0.7–1.1)0.6 (0.5–0.7)< 0.001LDL-C (mmol/L)1.3 (0.9–1.9)0.9 (0.7–1.4)< 0.001Cr (µmol/L)59.4 (48.1–72.0)57.5 (46.9–69.0)0.086BUN (mmol/L)4.8 (4.0–5.8)4.9 (4.0–6.3)0.595Ca (mmol/L)2.3 (2.2–2.4)2.3 (2.1–2.4)0.053ALB (g/L)43.9 (40.8–46.7)43.7 (40.7–45.9)0.222AMY (U/L)153.0 (74.5–344.0)194.0 (89.0–435.0)0.138LIP (U/L)269.5 (86.5–589.0)299.0 (115.0–704.0)0.169MAP Mild acute pancreatitis, non-MAP Moderate and severe acute biliary pancreatitis, APFC Acute Peripancreatic fluid collection, ANC Acute necrotic collection, PP Pancreatic pseudocyst, WON Walled-off pancreatic necrosis, WBC White blood cell, Neu Neutrophil, Lym Lymphocyte, Mono Monocyte, RDW Red cell distribution width, PLT Platelet, CRP C-reactive protein, ALT Alanine aminotransferase, AST Aspartate aminotransferase, TC Total cholesterol, TG Triglyceride, HDL-C High-density lipoprotein cholesterol, LDL-C Low-density lipoprotein cholesterol, Cr Creatini, AMY Amylase, LIP Lipase, BUN Blood urea nitrogen, Ca Calcium, ALB Albumin, NHR Neutrophil high-density lipoprotein cholesterol ratio Comparison of characteristics between high- and low-NHR groups MAP Mild acute pancreatitis, non-MAP Moderate and severe acute biliary pancreatitis, APFC Acute Peripancreatic fluid collection, ANC Acute necrotic collection, PP Pancreatic pseudocyst, WON Walled-off pancreatic necrosis, WBC White blood cell, Neu Neutrophil, Lym Lymphocyte, Mono Monocyte, RDW Red cell distribution width, PLT Platelet, CRP C-reactive protein, ALT Alanine aminotransferase, AST Aspartate aminotransferase, TC Total cholesterol, TG Triglyceride, HDL-C High-density lipoprotein cholesterol, LDL-C Low-density lipoprotein cholesterol, Cr Creatini, AMY Amylase, LIP Lipase, BUN Blood urea nitrogen, Ca Calcium, ALB Albumin, NHR Neutrophil high-density lipoprotein cholesterol ratio High NHR is an independent risk factor for HTG-AP severity Univariable analysis identified eight possible predictors of Non-MAP (WBC counts, neutrophils counts, lymphocyte counts, CRP levels, serum calcium levels, albumin levels, TG levels, and NHR levels; Supplementary Table 2). Subsequent multivariable and stepwise regression analyses demonstrated that NHR levels, CRP levels, TG levels, lymphocyte counts, and hypocalcemia independently predicted disease severity (Table 3). The stepwise regression process and a forest plot of influencing factors are shown in Fig. 3. Table 3Risk factors for Non-MAPCharacteristicMultivariable analysisStep analysisOR95% CIp-valueOR95% CIp-valueWBC (×109/L) ≤ 10—— > 101.290.61, 2.780.510Neu (×109/L) ≤ 6.3—— > 6.31.050.41, 2.730.913Lym (×109/L) ≤ 3.2———— > 3.20.270.09, 0.660.0080.280.10, 0.680.009CRP (mg/L) ≤ 6.00———— > 6.001.611.02, 2.570.0411.741.11, 2.740.016Ca (mmol/L) 2.18———— > 2.180.450.27, 0.730.0010.390.25, 0.61< 0.001ALB (g/L) ≤ 40.00—— > 40.000.660.37, 1.170.150TG (mmol/L)1.041.01, 1.060.0051.031.01, 1.060.005NHR < 16.66———— ≥ 16.662.191.38, 3.48< 0.0012.341.53, 3.62< 0.001All variables were included in the univariate logistic analysis. Variables that did not have a significant effect on outcome (p > 0.05) in the univariate analysis were not included in the multivariate logistic analysisWBC White blood cell, Neu Neutrophil, Lym Lymphocyte, CRP C-reactive protein, Ca Calcium, TG Triglyceride, NHR Neutrophil high-density lipoprotein cholesterol ratio Risk factors for Non-MAP All variables were included in the univariate logistic analysis. Variables that did not have a significant effect on outcome (p > 0.05) in the univariate analysis were not included in the multivariate logistic analysis WBC White blood cell, Neu Neutrophil, Lym Lymphocyte, CRP C-reactive protein, Ca Calcium, TG Triglyceride, NHR Neutrophil high-density lipoprotein cholesterol ratio Screening process of stepwise logistic regression analysis and forest plot of influencing factors. A: Screening process of stepwise logistic regression analysis. B: Forest plot of influencing factors. Variable:0=low, 1=high. sls: Significance Level to Stay. A predictor must have p-value < sls to remain in the model. sle: Significance Level to Enter. A predictor must have p-value < sle to enter the model NHR improves predictive model performance The final set of variables for model construction was selected by combining all factors identified as significant (P < 0.05) in the univariable, multivariable, stepwise regression analyses, and Lasso regression analyses(Fig. 4). This resulted in five factors: lymphocytes counts, CRP levels, serum calcium levels, TG, levels and NHR levels. Two predictive models were constructed. Model 1 (without NHR) included CRP levels, TG levels, calcium levels, lymphocyte counts (AUC: 0.696; 95% CI: 0.644–0.749). Model 2 (with NHR) showed significantly improved discrimination (AUC: 0.736; 95% CI: 0.685–0.786). NRI (value: 0.408, P < 0.001) and IDI (value: 0.48, P < 0.001) confirmed the incremental value of NHR (Fig. 5A). DeLong’s Test results revealed a ΔAUC (difference in AUC) of 0.040 (95% CI: 0.011–0.068; P = 0.006), suggesting that the addition of NHR enhanced the model’s performance. Visualized data demonstrated that Model 2 had higher sensitivity, specificity, positive predictive results (PPV), and negative predictive value (NPV) across most threshold ranges (Fig. 5B). Furthermore, we computed the Brier scores for both models and found that both scores were below 0.25, with Model 2 scoring 0.196, lower than that of Model 1 (0.208). Additionally, we provided a visual representation of the calibration curve (Fig. 5C). Screening for Non-MAP Risk Factors. A: Based on the cross-validation results, the LASSO regression model identified 14 optimal features for inclusion. B: By taking the intersection of the risk factors obtained through univariable, multivariable, stepwise, and LASSO regression analyses, five independent risk factors were finally selected The influence of NHR on the diagnostic effect of the prediction model. Model 1(without NHR): Logit(P) = 2.991667-0.324.991667.324.991667.324×Lym + 0.007×CRP-1.702×Ca + 0.037×TG. Model 2 (with NHR): Logit(P) = 1.644 − 0.265×Lym + 0.008×CRP-1.549×Ca + 0.040×TG + 0.051×NHR. A: Analyses of IDI. B: The sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) of two models. C: The calibration curve Linear association between NHR and disease severity To further explore the association between NHR and disease severity, Restricted Cubic Spline (RCS) curves were generated (Fig. 6). Both unadjusted (Fig. 6A) and adjusted (Fig. 6B, C and D) RCS curves demonstrated a linear association between NHR levels and disease severity (P for non-linearity > 0.05), indicating that higher NHR levels were associated with increased the severity of HTG-AP. The restricted cubic spline (RCS) curve of the relationship between NHR and HTG-AP severity. A: Unadjusted; B: Adjusted by factors including CRP, TG, lymphocyte count and serum calcium; C: Adjusted by factors including WBC, neutrophils, lymphocytes, CRP, serum calcium, serum albumin and TG. D: Adjusted by factors including lymphocytes, monocyte, CRP, serum calcium, ALT, AST, serum albumin, amylase, TC, TG, LDL-C, hypertension, diabetes Subgroup analysis based on demographic variables Subgroup analyses stratified by demographic factors (age, gender, diabetes, etc.) consistently demonstrated the association between high NHR levels and disease severity (all P < 0.05), and no significant interaction between NHR levels and the subgroup variables was observed (P for interaction > 0.05, Fig. 7). Demographic subgroup analysis of the relationship between NHR and the severity of HTG-AP. Variable: 0 = No, 1 = Yes. OtherCom: other comorbidities Discussion Current prognostic tools for acute pancreatitis (AP), such as the Acute Physiology and Chronic Health Evaluation II (APACHE II) and the Bedside Index for Severity in Acute Pancreatitis (BISAP), rely on complex parameters that are often unavailable during early hospitalization [11]. Consequently, many researchers are focusing on developing simpler and more efficient prediction methods. NHR offers a simple, rapid biomarker measurable at admission. Yang et al. demonstrated its predictive value in ABP [10]. Our study extends this evidence to HTG-AP, a subtype with a distinct lipid-driven pathophysiology. We found that elevated NHR levels independently were associatied with moderate/severe HTG-AP and enhanced the accuracy of severity prediction models. This association remained robust across demographic subgroups and provided secondary evidence underscoring the generalizability of NHR. Neutrophil-driven inflammation is pivotal to AP progression [12]. While neutrophil infiltration is essential for pathogen clearance in infectious conditions, this infiltration can exacerbate tissue damage in non-infectious diseases. During the early stages of AP, when infection is absent, excessive neutrophil activation can worsen disease severity, whereas neutrophil depletion strategies has been shown to have protective effects [13]. Neutrophil counts have been found to predict organ failure in biliary pancreatitis [14]. However, neutrophil counts alone exhibit limited prognostic value [14], necessitating integration with other parameters. HDL-C, beyond its cardiovascular implications, is recognized for its anti-inflammatory properties [15]. Studies have shown that low HDL-C levels are associated with increased severity of AP [16, 17]. In AP patients, decreased serum HDL-C reflects impaired anti-inflammatory function, leading to an accumulation of free fatty acids (FFAs), thereby exacerbating pancreatic cell damage and disease severity [17]. Notably, while Yang et al. reported higher HDL-C levels in mild ABP [10], we observed no significant differences in HDL-C levels across HTG-AP severities, which was consistent with Lin et al.’s findings [18]. This discrepancy may arise from differences in lipid metabolism between patients with ABP and those with HTG-AP. Patients with hypertriglyceridemia have a higher lipid burden and are more likely to experience abnormalities in high-density lipoprotein (HDL) levels and function [19]. Due to the technical challenges of directly assessing HDL particle functionality, HDL-C is widely used as a surrogate marker in clinical practice. The bidirectional interaction between HDL levels and neutrophil counts—where HDL modulates neutrophil activity and activated neutrophils alter HDL composition—has spurred interest in NHR as a novel biomarker [20]. Initially validated in acute myocardial infarction, the prognostic utility of NHR has expanded to other inflammatory conditions. Yang et al. first demonstrated the predictive value of NHR in biliary pancreatitis (ABP), linking elevated NHR levels to lipid metabolism dysregulation and disease severity [10]. Building on this, our study confirms a similar association in HTG-AP, with high NHR patients exhibiting prolonged hospitalization and higher complication rates, highlighting the clinical relevance of NHR for HTG-AP risk stratification. Although NHR is a novel inflammation scoring index, the relationship between elevated NHR levels and various diseases has attracted increasing attention. There are significant differences in the degree of NHR elevation among different diseases, with the magnitude of NHR elevation being significantly lower in chronic diseases compared with acute diseases. For example, Xu et al. found that the cutoff value of NHR for predicting deep venous thrombosis in patients with spontaneous intracerebral hemorrhage was 6.55 [21]. Hu’s study suggested that an NHR level greater than 3.65 could increase all-cause mortality in individuals aged 60 years and above [22]. In Yang’s study, patients with acute biliary pancreatitis and an NHR level greater than 9.54 had a higher risk of developing critical illness. In this study, the cutoff value of NHR was 16.66, which is greater than its value in other diseases. This may be because HTG-AP is an acute disease with a more severe systemic inflammatory response. Our study has several limitations. First, the single-center retrospective design and relatively smallt sample size may restrict the generalizability of findings and reduce statistical power for subgroup analyses (e.g., SAP; only 26 SAP patients were included in this study). Second, dichotomizing patients into mild and non-mild (moderate to severe) categories precluded specific evaluation the association between NHR levels and SAP. Third, many confounding factors such as body mass index (BMI) and eating habits have been largely missing, leading to insufficient comprehensiveness of the indicators included in the analysis, which may affect the accuracy of the research results. Finally, the mechanistic interplay underlying the association between NHR levels and HTG-AP severity—including neutrophil extracellular trap formation or free fatty acid metabolism—remains unexplored and warrants experimental validation. Conclusion NHR levels were independently associated with moderate/severe HTG-AP and enhanced risk stratification models. By integrating neutrophilic inflammation and lipid dysregulation, NHR provides actionable insights for clinicians to prioritize high-risk patients and tailor aggressive therapeutic strategies, potentially reducing complications and healthcare burdens. Supplementary Information Supplementary Material 1 Supplementary Material 2 Acknowledgements Not applicable. Abbreviations Acute Pancreatitis Hypertriglyceridemia Hypertriglyceridemic Acute Pancreatitis Neutrophil to High-Density Lipoprotein Cholesterol Ratio Mild Acute Pancreatitis Moderately Severe Acute Pancreatitis Severe Acute Pancreatitis Acute Biliary Pancreatitis Receiver Operating Characteristic Area Under the ROC Curve Integrated Discrimination Improvement Net Reclassification Index Restricted Cubic Splines White Blood Cell C-Reactive Protein High-Density Lipoprotein Cholesterol Low-Density Lipoprotein Cholesterol Total Cholesterol Triglyceride Alanine Aminotransferase Aspartate Aminotransferase Creatinine Blood Urea Nitrogen Albumin Amylase Lipase Acute Peripancreatic Fluid Collection Acute Necrotic Collection Pancreatic Pseudocyst Walled-Off Pancreatic Necrosis Intensive Care Unit Acute Physiology and Chronic Health Evaluation II Bedside Index for Severity in Acute Pancreatitis Interquartile Range Confidence Interval Positive Predictive Value Negative Predictive Value Authors’ contributions - Menggang Liu: Contributed to study conception, data collection, and manuscript review.- Xiaogang Hu: Contributed to manuscript writing, data analysis, methodology development, and data visualization.- Jia Xu: Contributed to data collection.- Juntao Yang: Contributed to manuscript review.All authors have read and approved the final and revised versions of the manuscript. Funding No funding was received to support this study. Data availability The data presented in this study are available on reasonable request from the corresponding author, Menggang Liu (Department of Hepatobiliary and Pancreatic Surgery, The People’s Hospital of Chongqing Liang Jiang New Area, No. 2 Jinkai Avenue, Renxing Road, Liangjiang New District, Chongqing 401121, China; E-mail: Menggang_liu@hotmail.com). Due to restrictions related to patient privacy and ethical guidelines, the data cannot be made publicly available. Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of the People’s Hospital of Chongqing Liang Jiang New Area (Ethics approval document number 189). All procedures involving human participants and data in this study were conducted in compliance with the principles of the Declaration of Helsinki (https://www.wma.net/policies-post/wma-declaration-of-helsinki/). Informed consent was waived by the Ethics Committee of the People’s Hospital of Chongqing Liang Jiang New Area due to the retrospective nature of the study and the anonymization of all patient data, which ensured that individual patient identities remained confidential. Consent for publication Not applicable. This study is a retrospective analysis of anonymized patient data, with no individual patient identifiers or personal health information included that could lead to patient identification. Therefore, consent for publication is not required. Competing interests The authors declare no competing interests. Footnotes The original online version of this article was revised: Tables 1 and 2 and Figure 5 were updated to correct processing errors. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Change history 1/27/2026 A Correction to this paper has been published: 10.1186/s12876-026-04627-6 References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement The data presented in this study are available on reasonable request from the corresponding author, Menggang Liu (Department of Hepatobiliary and Pancreatic Surgery, The People’s Hospital of Chongqing Liang Jiang New Area, No. 2 Jinkai Avenue, Renxing Road, Liangjiang New District, Chongqing 401121, China; E-mail: Menggang_liu@hotmail.com). Due to restrictions related to patient privacy and ethical guidelines, the data cannot be made publicly available.