Abstract
Background
The mortality following cardiac surgery remains high. The neutrophil-to-high-density lipoprotein cholesterol ratio (NHR) is a marker that reflects both inflammation and metabolic status, and it has shown promise in predicting outcomes across various diseases. However, the association of NHR with the outcomes in cardiac sur gery patients has not been fully validated.
Methods
This retrospective cohort study analyzed data from the MIMIC-Ⅳ database, including 2784 patients who underwent cardiac surgery. Patients were categorized into three groups (Q1, Q2, Q3) based on the NHR value. The primary outcome was 90-day all-cause mortality. The secondary outcomes included 180-day and 360-day all-cause mortality. Kaplan-Meier survival analysis, Cox proportional hazards regression, and restricted cubic spline (RCS) analysis were employed to assess the relationship between NHR and all-cause mortality.
Results
A total of 2784 patients (73.10% male) were enrolled. Higher NHR index levels were associated with an increased risk of 90-day,180-day, and 360-day all-cause mortality as shown by Kaplan–Meier curves. Cox proportional hazards analysis showed that the elevated NHR index was significantly related to all-cause death. Additionally, restricted cubic spline (RCS) analysis confirmed a linear positive relationship between NHR and all-cause mortality.
Conclusion
NHR is significantly associated with all-cause mortality in patients undergoing cardiac surgery. As a simple and cost-effective measure, NHR can support clinicians in the early identification of high-risk patients and guide personalized postoperative management strategies. To confirm its clinical utility and improve postoperative risk assessment and patient care, further large-scale, multicenter retrospective cohort studies are needed.
Keywords: Neutrophil, High-density lipoprotein cholesterol, Cardiac surgery, All-cause mortality, Prognosis
Introduction
Coronary artery bypass graft (CABG) and valve surgery are the most common types of cardiac surgery performed in adult patients [1, 2]. Despite significant advancement in surgical techniques and perioperative management, postoperative complications following cardiac surgery—such as hemodynamic instability, myocardial infarction, bleeding, pericardial effusion/tamponade, renal insufficiency, gastrointestinal and neurological complications, atrial fibrillation, difficulty in weaning from mechanical ventilation, and sternal wound infection/mediastinitis—remain considerable [3, 4]. In addition, the mortality rate of cardiac surgery is relatively high, with a mortality rate of 2.0% for CABG and 3.2% for aortic valve replacement [5]. Therefore, identifying risk factors for postoperative mortality in patients undergoing cardiac surgery is important for improving patient outcomes.
Previous studies have shown that neutrophil plays an important role in cardiovascular inflammatory responses [6]. High-density lipoprotein cholesterol (HDL-C), an important indicator of cardiovascular disease, has been found to have a correlation with cardiovascular events in recent research [7]. The neutrophil-to-high-density lipoprotein cholesterol ratio (NHR) is a composite indicator that integrates a pro-inflammatory marker (neutrophil) with a metabolic marker (HDL-C). It is calculated using the following formula: NHR = Neutrophil Count (×10⁹/L) / HDL-C (mg/dL) [8]. Both neutrophil and HDL-C are easily measured and can be standardized, making NHR a simple and practical parameter for assessing cardiovascular metabolic inflammation.
Previous studies have shown that NHR is a promising indicator for various conditions, including metabolic syndrome, ischemic stroke, myocardial infarction, Parkinson’s disease, and cardiovascular mortality [9]. However, research on the association between NHR and postoperative mortality in patients undergoing cardiac surgery remains limited. This study aimed to investigate the relationship between NHR and all-cause mortality following cardiac surgery.
Methods
Ethics approvement
The establishment of the Medical Information Mart for Intensive Care IV (MIMIC-Ⅳ) database was approved by the institutional review boards of Beth Israel Deaconess Medical Center and the Massachusetts Institute of Technology (Cambridge, MA, USA). Informed consent was obtained for the original data collection. As this study used de-identified, publicly available data, additional ethical approval and informed consent were not required.
Source of data
This retrospective study utilized data from the MIMIC-Ⅳ (version 3.1) database, a publicly available resource containing detailed clinical information on over 190,000 patients and 450,000 hospital admissions at Beth Israel Deaconess Medical Center between 2008 and 2019. The database includes comprehensive records on demographics, laboratory results, comorbidities, medications, vital signs, surgical procedures, diagnoses, and follow-up outcomes. One author (Zehan Guo) completed the required National Institutes of Health training and passed the Collaborative Institutional Training Initiative (CITI) certification to access the data. All patient records are fully anonymized.
Study design and population
We included adult patients (age ≥ 18 years) who underwent cardiac surgery, including CABG, valve surgery, or both CABG and valve surgery. Exclusion criteria include: (1) missing neutrophil data; (2) missing HDL-C measurement at admission; (3) extreme or biologically implausible values for neutrophil or HDL-C.
Data extraction
This study extracted the information of patients meeting the inclusion criteria from the database by using Structured Query Language (SQL) in PostgreSQL. The extracted data included the following variables: Demographic Information: Age, gender, height, and weight; Comorbidities: Hypertension, pulmonary tuberculosis, pneumonia, stroke, chronic kidney disease, type 1 diabetes mellitus (T1DM), type 2 diabetes mellitus (T2DM), hyperlipidemia, chronic bronchitis, heart failure, myocardial infarction, and chronic obstructive pulmonary disease (COPD); Vital Signs: heart rate, respiratory rate, temperature, arterial systolic blood pressure, and arterial diastolic blood pressure; Disease scores: the Sequential Organ Failure Assessment (SOFA) and Acute Physiology Score III (APS-III); Data on drug administration within 48 h of admission to the ICU: Total dose of Vasopressor drugs (Vptotalval). Laboratory Indicators After ICU Admission: White blood cell count, platelet count, red blood cell count, hemoglobin, anion gap, lactate, creatinine, blood urea nitrogen, glucose, potassium, PT-INR, and fibrinogen; Neutrophil and High-Density Lipoprotein Cholesterol After Hospital Admission.
Management of missing data
Patients with more than 20% missing data of variables were excluded from the analysis. For variables with less than 20% missing, multiple imputation was used to handle missing values.
Outcomes
The primary outcome was 90-day all-cause mortality. Secondary outcomes included 180-day and 360-day all-cause mortality.
Statistical analysis
Continuous variables were assessed for normality. Non-normally distributed data were analyzed using the Wilcoxon rank-sum test and reported as medians with interquartile ranges (IQR). Normally distributed data were analyzed using analysis of variance and reported as mean with standard deviation(SD). Categorical variables were compared using the chi-square test or Fisher’s exact test, with results expressed as frequencies and percentages. Survival outcomes were analyzed using Kaplan–Meier curves. Multivariate Cox proportional hazard regression was used to assess the association between the neutrophil-to-HDL-C ratio (NHR) and all-cause mortality, with the lowest NHR tertile serving as the reference group. Restricted cubic spline (RCS) analysis was performed to explore potential nonlinear dose-response relationships between NHR and mortality. Predefined subgroup analyses were conducted based on sex (male, female), age (≤ 75 or > 75 years), creatinine level (≤ 1.02 mg/dL or > 1.02 mg/dL), and the presence or absence of comorbidities, including hypertension, acute kidney injury (AKI), stroke, hyperlipidemia, myocardial infarction, heart failure, type 1 diabetes (T1DM), and type 2 diabetes (T2DM). All statistical analyses were performed using DecisionLinnc 1.0 software (DecisionLinnc Corte Team, 2023), a platform integrating multiple programming environments through a visual interface for comprehensive data processing and analysis. A two-sided p-value of less than 0.05 was considered statistically significant.
Results
A total of 2,784 patients met the inclusion criteria and were included in the final analysis. Based on the distribution of the neutrophil-to-HDL-C ratio (NHR), all participants were categorized into three groups (Fig. 1).
Fig. 1.
The flow chart of participant selection in this study
Baseline characteristics of the study participants
This study analyzed data from 10,011 patients in the MIMIC-Ⅳ database, of whom 2784 met the inclusion criteria. Participants were divided into three groups based on NHR percentiles (Q1, Q2, Q3). The baseline characteristics of each group are detailed in Table 1. The median age of the cohort was 68 years (range: 61–76 years), with male participants accounting for 73.10% (2035 individuals). Notably, the highest proportion of male participants was observed in Group Q3, at 85.78%. The common comorbidities were T2DM (36.35%), heart failure (33.51%), and myocardial infarction (17.85%), all of which had the highest prevalence in Group Q3. Additionally, patients in Group Q3 had the highest levels of postoperative hematocrit, white blood cell count, hemoglobin, platelet, anion gap, functional fibrinogen, creatinine, blood urea nitrogen, APS-III score, and duration of mechanical ventilation.
Table 1.
Baseline characteristics of patients grouped according to NHR index tertiles
| Variable | Overall N = 2784 |
NHR | p-value | ||
|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | |||
| N = 928 | N = 928 | N = 928 | |||
| Age (years) | 68 (61–76) | 68 (61–77) | 69 (60–76) | 67 (60–74) | < 0.001 |
| Weight (kg) | 85 (74–98) | 79 (69–91) | 86 (74–99) | 90 (79–104) | < 0.001 |
| Height (m) | 1.73 (1.65–1.78) | 1.70 (1.63–1.78) | 1.73 (1.65–1.78) | 1.73 (1.68–1.79) | < 0.001 |
| SOFA | 5 (3–7) | 5 (3–7) | 5 (3–7) | 5 (3–7) | 0.014 |
| APS-III | 35 (27–46) | 33 (26–45) | 34 (27–45) | 36 (28–48) | 0.006 |
| Vptotalval (mg) | 60 (14–72) | 62 (24–72) | 61 (14–72) | 60 (12–72) | 0.020 |
| Hematocrit (%) | 28.70 (25.10–32.60) | 28.10 (24.90–32.10) | 28.90 (25.00–32.40) | 29.40 (25.70–33.10) | < 0.001 |
| Hemoglobin (g/dL) | 9.60 (8.30–11.00) | 9.40 (8.20–10.80) | 9.70 (8.30–11.00) | 9.80 (8.50–11.20) | 0.010 |
| Platelet (10^9/L) | 147 (118–184) | 141 (112–178) | 145 (118–181) | 156 (125–198) | < 0.001 |
| RBC (10^9/L) | 3.19 (2.78–3.63) | 3.08 (2.75–3.57) | 3.22 (2.76–3.61) | 3.29 (2.86–3.71) | < 0.001 |
| WBC (10^9/L) | 11.95 (8.80–15.50) | 10.80 (7.80–13.90) | 12.15 (8.95–15.60) | 13.00 (9.90–16.65) | < 0.001 |
| Anion gap (mmol/L) | 11 (9–13) | 11 (9–13) | 11 (9–13) | 11 (9–14) | 0.141 |
| Glucose (mg/dL) | 119 (104–137) | 117 (103–133) | 120 (105–137) | 121 (106–140) | < 0.001 |
| Potassium (mmol/L) | 4.30 (4.00–4.60) | 4.20 (3.90–4.60) | 4.30 (4.00–4.60) | 4.40 (4.00–4.70) | < 0.001 |
| Lactate (mmol/L) | 2.00 (1.40–2.60) | 1.90 (1.35–2.65) | 2.00 (1.40–2.70) | 2.00 (1.50–2.60) | 0.202 |
| Fibrinogen functional (mg/dL) | 210 (172–262) | 200 (169–238) | 209 (170–261) | 228 (181–284) | < 0.001 |
| INRPT | 1.40 (1.30–1.60) | 1.40 (1.20–1.50) | 1.40 (1.30–1.60) | 1.40 (1.30–1.60) | 0.002 |
| Creatinine (mg/dL) | 0.90 (0.70–1.10) | 0.80 (0.70–1.00) | 0.90 (0.70–1.10) | 1.00 (0.80–1.30) | < 0.001 |
| Urea nitrogen (mg/dL) | 16 (13–22) | 15 (12–21) | 16 (13–22) | 18 (14–24) | < 0.001 |
| HR (bmp) | 80 (75–87) | 80 (74–86) | 80 (75–87) | 80 (75–88) | 0.133 |
| ABPD (mmHg) | 58 (51–65) | 59 (52–66) | 58 (51–65) | 57 (51–65) | 0.091 |
| ABPS (mmHg) | 112 (101–123) | 113 (102–124) | 112 (101–122) | 111 (100–122) | 0.098 |
| RR (insp/min) | 15 (13–18) | 15 (13–18) | 15 (12–18) | 15 (13–18) | 0.485 |
| Temperature (℉) | 98.00 (97.60–98.50) | 98.00 (97.60–98.50) | 98.00 (97.60–98.50) | 98.00 (97.60–98.40) | 0.887 |
| Neutrophils (%) | 75.50 (68.40–81.30) | 70.00 (62.77–77.20) | 76.25 (69.63–81.90) | 78.80 (73.70–83.40) | < 0.001 |
| Cholesterol (mg/dL) | 45.42 (38.00–55.00) | 59.61 (53.00–66.50) | 46.31 (42.00–50.00) | 35.03 (32.00–39.00) | < 0.001 |
| NHR | 1.64 (1.32–2.00) | 1.21 (1.07–1.32) | 1.64 (1.54–1.75) | 2.17 (2.00–2.44) | < 0.001 |
| BMI (Kg/m2) | 28.96 (25.76–32.82) | 27.85 (24.82–31.58) | 29.07 (25.75–32.75) | 30.13 (26.87–34.21) | < 0.001 |
| Ventilation (hour) | 27.22 (17.00–55.25) | 27.00 (16.04–53.93) | 26.39 (17.47–52.00) | 28.65 (18.00–61.18) | 0.046 |
| Male n (%) | 2,035 (73.10%) | 534 (57.54%) | 705 (75.97%) | 796 (85.78%) | < 0.001 |
| HTN, n (%) | 1,500 (53.88%) | 503 (54.20%) | 530 (57.11%) | 467 (50.32%) | < 0.013 |
| AKI, n (%) | 554 (19.90%) | 131 (14.12%) | 199 (21.44%) | 224 (24.14%) | < 0.001 |
| PTB, n (%) | 27 (0.97%) | 10 (1.08%) | 7 (0.75%) | 10 (1.08%) | 0.714 |
| PNA, n (%) | 164 (5.89%) | 37 (3.99%) | 62 (6.68%) | 65 (7.00%) | 0.010 |
| CVA, n (%) | 217 (7.79%) | 75 (8.08%) | 75 (8.08%) | 67 (7.22%) | 0.726 |
| CKD, n (%) | 536 (19.25%) | 137 (14.76%) | 176 (18.97%) | 223 (24.03%) | < 0.001 |
| T2DM, n (%) | 1,012 (36.35%) | 248 (26.72%) | 334 (35.99%) | 430 (46.34%) | < 0.001 |
| T1DM, n (%) | 62 (2.23%) | 21 (2.26%) | 22 (2.37%) | 19 (2.05%) | 0.891 |
| HLD, n (%) | 1,935 (69.50%) | 630 (67.89%) | 656 (70.69%) | 649 (69.94%) | 0.398 |
| CB, n (%) | 135 (4.85%) | 33 (3.56%) | 44 (4.74%) | 58 (6.25%) | 0.026 |
| HF, n (%) | 933 (33.51%) | 312 (33.62%) | 300 (32.33%) | 321 (34.59%) | 0.585 |
| MI, n (%) | 497 (17.85%) | 120 (12.93%) | 160 (17.24%) | 217 (23.38%) | < 0.001 |
| 90-day mortality, n (%) | 38 (1.36%) | 6 (0.65%) | 12 (1.29%) | 20 (2.16%) | 0.019 |
| 180-day mortality, n (%) | 46 (1.65%) | 7 (0.75%) | 17 (1.83%) | 22 (2.37%) | 0.021 |
| 360-day mortality, n (%) | 59 (2.12%) | 10 (1.08%) | 21 (2.26%) | 28 (3.02%) | 0.014 |
NHR, neutrophil-to-high-density lipoprotein cholesterol ratio; Vptotalval, The total dose of vasopressors within 48 h; SOFA, sequential organ failure assessment; APS-III, acute physiology score III; RBC, red blood cell count; WBC, white blood cell count; INRPT, prothrombin time international normalized ratio; HR, heart rate; ABPD, arterial blood pressure diastolic; ABPS, arterial blood pressure systolic; RR, respiratory rate; BMI, Body mass index; HTN, hypertension; AKI, acute kidney injury; PTB, pulmonary tuberculosis; PNA, pneumonia; CVA, stroke; CKD, chronic kidney disease; T2DM, type 2 diabetes; T1DM, type 1 diabetes; HLD, hyperlipidemia; CB, chronic bronchitis; HF, Heart failure; MI, Myocardial infarction
*Statistically significant: a value greater than 0.05 is interpreted as a meaningful difference
There were significant differences in clinical outcomes across NHR tertiles. In the highest tertile, 90-day, 180-day, and 360-day all-cause mortality rates were 2.16%, 2.37%, and 3.02%, respectively. Cox regression analysis demonstrated a positive association between NHR and 90-day all-cause mortality. Compared with Q1, the unadjusted model showed [HR, 3.352 (95% CI 1.346–8.348); p = 0.009], the partially adjusted model yielded [HR, 4.404 (95% CI 1.715–11.302); p = 0.002], and the fully adjusted model showed [HR, 3.104 (95% CI 1.186–8.124); p = 0.021]. Similar associations were observed for 180-day and 360-day all-cause mortality in multivariable Cox analyses. These findings indicate that higher NHR is consistently associated with increased mortality at 90, 180, and 360 days, as summarized in Table 2.
Table 2.
Cox regression models for 90-day, 180-day, and 360-day all-cause mortality
| Categories | Model 1 | Model 2 | Model 3 | ||||||
|---|---|---|---|---|---|---|---|---|---|
| HR (95% CI) | P value | P for trend | HR (95% CI) | P value | P for trend | HR (95% CI) | P value | P for trend | |
| 90-day mortality | |||||||||
| Quartile | 0.007 | 0.001 | 0.016 | ||||||
| Q1 (N = 928) | Ref | Ref | Ref | ||||||
| Q2 (N = 928) | 2.005 (0.753–5.342) | 0.164 | 2.297 (0.855–6.172) | 0.099 | 1.837 (0.679–4.971) | 0.231 | |||
| Q3 (N = 928) | 3.352 (1.346–8.348) | 0.009 | 4.404 (1.715–11.302) | 0.002 | 3.104 (1.186–8.124) | 0.021 | |||
| 180-day mortality | |||||||||
| Quartile | 0.008 | 0.001 | 0.027 | ||||||
| Q1 (N = 928) | Ref | Ref | Ref | ||||||
| Q2 (N = 928) | 2.438 (1.001–5.878) | 0.047 | 2.743 (1.129–6.667) | 0.026 | 2.138 (0.875–5.224) | 0.096 | |||
| Q3 (N = 928) | 3.167 (1.353–7.413) | 0.008 | 4.043 (1.682–9.718) | 0.002 | 2.760 (1.129–6.748) | 0.026 | |||
| 360-day mortality | |||||||||
| Quartile | 0.004 | 0.001 | 0.017 | ||||||
| Q1 (N = 928) | Ref | Ref | Ref | ||||||
| Q2 (N = 928) | 2.112 (0.994–4.484) | 0.052 | 2.317 (1.082–4.958) | 0.03 | 1.880 (0.875–4.041) | 0.106 | |||
| Q3 (N = 928) | 2.828 (1.374–5.822) | 0.005 | 3.377 (1.601–7.124) | 0.001 | 2.517 (1.179–5.374) | 0.017 | |||
Model 1: unadjusted
Model 2: adjusted for age, sex
Model 3: adjusted for age, sex, AKI, HF, MI, Vptotalval
*Statistically significant: a value greater than 0.05 is interpreted as a meaningful difference; HR, Hazard ratio; CI, Confidence interval
Q1 (0.6–1.44); Q2 (1.44–1.86); Q3 (1.86–3.10)
Survival analysis
Kaplan-Meier survival analysis revealed significant differences in all-cause mortality survival rates at 90 days, 180 days, and 360 days across the tertiles of NHR. Compared with patients in the lower tertiles of NHR, those in Group Q3 had the lowest survival rates at all time points (Fig. 2).
Fig. 2.
Kaplan–Meier survival analysis curves for all-cause mortality. Kaplan–Meier curves of hospital 90-day mortality, 180-day mortality, and 360-day mortality stratified by NHR index, NHR, neutrophil-to-high-density lipoprotein cholesterol ratio
Restricted cubic spline
Restricted cubic spline(RCS) analysis demonstrated a significant linear relationship (p for nonlinearity > 0.05) between NHR and all-cause mortality at each time point (90, 180, and 360 days). As NHR values increased, the all-cause mortality rate showed an upward trend (Fig. 3).
Fig. 3.
RCS of NHR index with all-cause mortality. RCS of NHR index with 90-day mortality, 180-day mortality, and 360-day mortality. NHR, neutrophil-to-high-density lipoprotein cholesterol ratio; RCS, restricted cubic splines
Stratified analyses
We conducted subgroup analyses to assess the potential modifying effects of age (≤ 75 years or > 75 years), gender (male, female), BMI (≤ 25 or > 25), hypertension (HTN), acute kidney injury (AKI), chronic kidney disease (CKD), type 2 diabetes (T2DM), hyperlipidemia (HLD), myocardial infarction (MI), and heart failure (HF) on the association between NHR and all-cause mortality at 90, 180, and 360 days (Fig. 4). The results showed that NHR was significantly associated with 90-day, 180-day, and 360-day mortality in patients without HTN, AKI, HLD, T2DM, and MI (p < 0.05). In contrast, no significant associations were observed in patients with these conditions. Among patients aged ≤ 75 years, NHR was significantly associated with 360-day mortality. In females, NHR was significantly associated with 90-day, 180-day, and 360-day mortality. In patients without CKD, NHR was significantly associated with 360-day mortality, whereas no significant associations were observed at 90 or 180 days. Conversely, among patients with CKD, a significant association was observed at 90 days, but not at 180 or 360 days. For AKI, the interaction effect reached borderline statistical significance at 180 days (P for interaction = 0.049). Hyperlipidemia (HLD) showed significant heterogeneity, with a p-value for interaction < 0.05.
Fig. 4.
Forest plots of hazard ratios for the 90-day mortality, 180-day mortality, and 360-day mortality in different subgroups. HR, hazard ratio; CI, confidence interval; HTN/AKI/CKD/T2MD/HLD/MI/HF: 0 = no, 1 = yes; Age: 0 ≤ 75, 1 > 75; Gender: 0 = Female, 1 = Male; BMI: 0 ≤ 25, 1 > 25
Discussion
This is the first study to confirm that the relationship of neutrophil-to-high-density lipoprotein cholesterol ratio (NHR) with all-cause mortality at 90, 180, and 360 days in patients undergoing cardiac surgery. Compared with the low NHR group (Q1), patients in the high NHR group (Q3 > 2.17) had approximately a 2.8–3.3-fold higher risk of death. Kaplan–Meier survival analysis demonstrated that elevated NHR was significantly associated with reduced survival, and restricted cubic spline analysis further confirmed a significant linear relationship between NHR and postoperative mortality.
Previous research on NHR has focused on non-surgical populations, particularly in the context of cardiovascular and cerebrovascular diseases. Numerous studies have shown that NHR, as an indicator of both inflammatory and metabolic status, can effectively predict the risk of cardiovascular events and all-cause mortality [10, 11]. A community-based population study found that higher NHR levels were significantly associated with the increased incidence of cardiovascular diseases and higher mortality, suggesting that NHR—an inexpensive and easily obtainable prognostic marker—has important clinical value [12]. Additionally, studies have shown that NHR is an independent risk factor for acute ischemic stroke and is positively correlated with disease severity, further expanding its potential applications in neurological disorders [13]. The present study confirms the role of NHR in predicting postoperative mortality risk in patients undergoing cardiac surgery.
Postoperative inflammatory response is common following cardiac surgery and is associated with all-cause mortality [14]. Cardiac surgery procedures involving cardiopulmonary bypass (CPB) can trigger systemic inflammatory response syndrome (SIRS) through mechanisms such as tissue injury, hypothermia, neuroendocrine activation, and medication use. This inflammatory cascade can lead to hypermetabolism, insulin resistance, the release of large quantities of pro-inflammatory factors and chemokines, which are associated with organ dysfunction, coagulation abnormalities, and adverse outcomes [15, 16]. Typically, postoperative inflammation presents as systemic leukocytosis, characterized by a mild increase of neutrophils [17]. HDL-C transports peripheral cholesterol back to the liver for metabolism through reverse cholesterol transport(RCT). HDL-C has anti-inflammatory, antioxidant, and endothelial-protective effects [18]. However, in acute and chronic inflammatory states, HDL becomes dysfunctional. Dysfunctional HDL adopts pro-inflammatory properties, losing its ability to inhibit oxidized low-density lipoprotein (oxLDL) accumulation or suppress the monocyte chemoattractant activity induced by low-density lipoprotein (LDL). The buildup of oxLDL promotes leukocyte activation, secretion of pro-inflammatory cytokines, expression of leukocyte adhesion molecules, cell degranulation, reactive oxygen species (ROS) release, and endothelial dysfunction [19]. Following cardiac surgery, inflammatory responses and metabolic disturbances are risk factors for complications [20]. An imbalance, such as an elevated neutrophil-to-HDL-C ratio (NHR), may worsen postoperative myocardial repair and increase the risk of organ damage. The association between elevated NHR and higher postoperative mortality in cardiac surgery may therefore be mediated by combined inflammatory and metabolic pathways. Previous literature has highlighted the clinical relevance of NHR, relating it to coronary artery stenosis severity, ischemic stroke, and thrombus burden [21–23]. Furthermore, studies have shown that NHR is independently associated with the risk of major adverse cardiovascular events (MACE) [24]. Several studies have demonstrated that perioperative cardiac biomarkers such as NT-proBNP and cardiac troponins reflect distinct but complementary pathophysiological pathways in cardiovascular risk prediction. NT-proBNP levels and Troponin T levels constitute complementary pathophysiological dimensions in the perioperative period. Preoperative NT-proBNP reflects myocardial wall stress and ventricular function, and a prospective cohort study has demonstrated its ability to independently predict early adverse outcomes after cardiac surgery [25, 26]. Meanwhile, studies have shown that an increase in postoperative troponin T levels can serve as an important predictor of mortality after cardiac surgery [27]. Patients with higher NHR had significantly higher levels of NT-proBNP and troponin T, suggesting that systemic inflammation and direct myocardial damage may have complementary effects in prognosis [28, 29]. NHR captures the interaction between systemic inflammation and lipid metabolism and may therefore provide prognostic information that is complementary to myocardial injury biomarkers. Future studies should evaluate multimarker strategies integrating NHR with NT-proBNP and troponin T to improve postoperative risk stratification.
In the subgroup analysis, NHR was significantly associated with 90-day, 180-day, and 360-day mortality after cardiac surgery in patients without diabetes or hyperlipidemia. In the third tertile, the mean HDL-C value was 35.03 mg/dL, which is well below the normal range of 1.16–1.55 mmol/L (approximately 45–60 mg/dL). Such a low HDL-C level is closely related to metabolic abnormalities, including diabetes (T2DM accounted for 46.34%), obesity (BMI 30.13 kg/m²), and hyperlipidemia. Previous studies have shown that, in obesity, macrophages undergo significant changes, with increased numbers and a more pro-inflammatory phenotype. Research into the pathogenesis of obesity, insulin resistance, and diabetes has suggested that inflammation plays a central role in initiating insulin resistance, impairing insulin secretion, and disrupting energy homeostasis [30]. Serum HDL-C levels are strongly and negatively correlated with both the presence and severity of hyperlipidemia, and they are also negatively correlated with neutrophil counts. As a result, neutrophil levels are significantly higher in patients with hyperlipidemia compared with those without [31]. In insulin resistance, HDL-C levels tend to be even lower [32]. Reduced synthesis and impaired function of HDL-C, combined with inflammation-driven neutrophil recruitment, may create a persistent “inflammation–metabolism” vicious cycle. Under inflammatory conditions, HDL is prone to dysfunction, losing its anti-inflammatory and antioxidant properties, with serum amyloid-mediated alterations considered a key mechanism. HDL cholesterol efflux capacity, a functional measure of HDL, is independently and inversely associated with atherosclerotic events, suggesting that “functional HDL,” rather than HDL-C levels alone, better reflects risk. This understanding provides biological support for the prognostic value of NHR, which uses HDL as the denominator [33, 34]. In our study, the significant association between NHR and postoperative mortality (P < 0.05) in patients without hyperlipidemia is of particular interest. Hypercholesterolemia is a well-established major risk factor for atherosclerosis [35], and for cardiovascular disease patients with hyperlipidemia, risk assessment and treatment are now well-defined [36]. However, much less is known about the prognosis of patients without hyperlipidemia. In this group, the inflammatory signal reflected by NHR may have independent clinical value and could serve as a useful indicator for risk stratification and prognosis assessment. This finding also suggests that inflammation management in this population should focus on pathways unrelated to lipid metabolism. Compared with the hyperlipidemia subgroup, these results enhance our understanding of the mechanism by which the interaction between inflammation and metabolism determines the prognostic effect of NHR, and they provide an important basis for the individualized application of NHR according to lipid profile. On the other hand, lipid-lowering therapy (especially statins) not only significantly reduces LDL-C but also lowers high-sensitivity C-reactive protein (hs-CRP), indicating anti-inflammatory (immunomodulatory) effects [37]. Therefore, in patients with hyperlipidemia or those already receiving lipid-lowering treatment, the anti-inflammatory action of the medication may dilute the inflammatory burden reflected by NHR, thereby weakening its apparent association with mortality risk. This is consistent with the attenuated effect observed in our subgroup interaction analyses. It is possible that patients without hyperlipidemia also have other uncontrolled confounding factors, such as stricter health management or a lower baseline disease burden, which might influence both NHR and outcomes. These factors could potentially lead to an overestimation of the association. Therefore, multivariate adjustment, sensitivity analysis, and validation in external cohorts are necessary to reduce bias from the randomness of a single analysis.
In the stratified analysis by AKI and CKD status, NHR demonstrated a stronger and more stable predictive effect on intermediate-term outcomes in patients without AKI or before the onset of AKI. Once AKI occurred, however, intermediate-term mortality became predominantly driven by powerful determinants such as worsening renal function, fluid and electrolyte imbalances, and renal replacement therapy, thereby diluting the incremental information provided by the inflammation–lipoprotein composite phenotype [38]. This phenomenon aligns closely with observations in cardiac surgery populations, where cardiopulmonary bypass triggers a systemic inflammatory response and neutrophil-centered immune activation [39]. Meanwhile, in patients with CKD, the presence of chronic low-grade inflammation and HDL dysfunction over the long term creates a baseline characterized by high NHR values with limited variability, making it more prone to time-dependent and unstable associations.
This study also performed subgroup analyses of the association between NHR and postoperative all-cause mortality across a range of comorbid conditions. The interaction terms for subgroups such as hypertension (HTN), myocardial infarction (MI), and heart failure (HF) did not reach statistical significance (P > 0.05). Nevertheless, the association between NHR and mortality remained significant (P < 0.05) in patients without these comorbidities. One previous study showed that NHR is an independent risk factor for all-cause mortality in the elderly, with higher NHR levels associated with increased mortality risk [40]. This suggests that in patients with fewer underlying diseases and better overall health status, NHR may serve as a refined risk marker for identifying mortality risk linked to elevated inflammation. Such a marker could help address the limited discriminatory power of traditional scoring systems, such as SOFA and APS III, in identifying high-risk individuals within low-risk populations. Further research is needed to validate this conclusion.
However, this study has several limitations that should be considered when interpreting the results. First, its retrospective design and the use of data from a single medical center, specifically the MIMIC-IV database, limit the ability to establish causality. Although multiple confounding factors were adjusted for and subgroup analyses were performed, the possibility of residual confounding cannot be excluded. Second, the relatively low proportion of positive indicators in the sample suggests that the findings should be confirmed in larger cohort studies. Third, some potentially important perioperative variables were incompletely captured. Specifically, NT-proBNP and Troponin T were missing in a substantial proportion of patients. Including these biomarkers would have reduced the effective sample size and may have introduced selection bias. Therefore, we did not incorporate NT-proBNP or Troponin T into the primary models and could not reliably evaluate their correlations with NHR or their incremental prognostic value beyond the main analyses. Finally, although a significant association was observed between NHR and all-cause mortality at 90, 180, and 360 days, the biological mechanisms underlying this relationship remain unclear. This gap in mechanistic understanding may limit the generalizability and external applicability of the conclusions.
Conclusion
In conclusion, the neutrophil-to-HDL-C ratio (NHR) appears to be a promising and cost-effective biomarker for predicting both short-term and long-term all-cause mortality in adults undergoing cardiac surgery. By capturing the interplay between systemic inflammation and metabolic dysfunction, NHR provides prognostic information that complements conventional risk factors. Importantly, its predictive value remains significant even in patients without dyslipidemia, highlighting its potential applicability across a wide surgical population. Further validation through large-scale, prospective, multicenter studies is warranted to confirm these findings and to better define their role in clinical practice.
Acknowledgements
This work was supported by the [Shandong Medical Association Clinical Research Fund-Qilu Special Project] (No.YXH2022ZX02088) and the [Natural Science Foundation of Shandong Province] (No. ZR2021MH182).
Abbreviations
- NHR
Neutrophil-to-high-density lipoprotein cholesterol ratio
- Vptotalval
The total dose of vasopressors within 48 h
- SOFA
Sequential organ failure assessment
- APS-III
Acute physiology score III
- RBC
Red blood cell count
- WBC
White blood cell count
- INRPT
Prothrombin time international normalized ratio
- HR
Heart rate
- ABPD
Arterial blood pressure diastolic
- ABPS
Arterial blood pressure systolic
- RR
Respiratory rate
- BMI
Body mass index
- HTN
Hypertension
- AKI
Acute kidney injury
- PTB
Pulmonary tuberculosis
- PNA
Pneumonia
- CVA
Stroke
- CKD
Chronic kidney disease
- T2DM
Type 2 diabetes
- T1DM
Type 1 diabetes
- HLD
Hyperlipidemia
- CB
Chronic bronchitis
- HF
Heart failure
- MI
Myocardial infarction
Author contributions
GZH performed the analysis and drafted the manuscript. LTS and ZPP provided clinical advice and managed the significance of clinical metrics. ZK and LGF assisted in creating and revising the tables and figures. HLH and LXY performed the secondary checking of the data in tables and graphs. ZQ and LM supervised the study and were involved in revising the manuscript. All authors gave final approval of the version to be published.
Funding
This work was supported by the Shandong Medical Association Clinical Research Fund-Qilu Special Project (YXH2022ZX02088) and the Natural Science Foundation of Shandong Province (ZR2021MH182).
Data availability
Publicly available datasets were analyzed in this study. These data can be found at [https://mimic.mit.edu/].
Declarations
Ethics approval and consent to participate
The data was extracted from Medical Information Mart for Intensive MIMIC-IV (version 3.1). The identification information was concealed, and the privacy of patients in MIMIC-IV was protected. Thus, there were no additional consent procedures from the institutional ethics committee.
Consent for publication
No applicable.
Competing interests
The authors declare no competing interests.
Data sharing
To obtain these datasets, interested parties should submit a formal request. This request should be addressed to the corresponding author of this study.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Meng Lv, Email: qylvmeng@163.com.
Qian Zhang, Email: 279807939@qq.com.
References
- 1.Unger P, Pibarot P, Tribouilloy C, et al. Multiple and mixed valvular heart diseases: pathophysiology, imaging, and management[J]. Circ Cardiovasc Imaging. 2018;11(8):e007862. [DOI] [PubMed] [Google Scholar]
- 2.Ruel M, Chikwe J. Coronary artery bypass grafting: past and future[J]. Circulation. 2024;150(14):1067–9. [DOI] [PubMed] [Google Scholar]
- 3.Otto CM, Nishimura RA, Bonow RO, Carabello BA, Erwin JP 3rd, Gentile F, Jneid H, Krieger EV, Mack M, McLeod C, O'Gara PT, Rigolin VH, Sundt TM 3rd, Thompson A, Toly C. 2020 ACC/AHA Guideline for the Management of Patients With Valvular Heart Disease: Executive Summary: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2021;143(5):35–71. [DOI] [PubMed]
- 4.Anthony A, Sendelbach S. Postoperative complications of coronary artery bypass grafting surgery[J]. Crit Care Nurs Clin N Am. 2007;19(4):403–15. [DOI] [PubMed] [Google Scholar]
- 5.Bridgewater B. Mortality data in adult cardiac surgery for named surgeons: retrospective examination of prospectively collected data on coronary artery surgery and aortic valve replacement[J]. BMJ. 2005;330(7490):506–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Silvestre-Roig C, Braster Q, Ortega-Gomez A, et al. Neutrophils as regulators of cardiovascular inflammation[J]. Nat Reviews Cardiol. 2020;17(6):327–40. [DOI] [PubMed] [Google Scholar]
- 7.Kulik A, Ruel M, Jneid H et al. Secondary prevention after coronary artery bypass graft surgery[J].
- 8.Shi K, Hou J, Zhang Q, et al. Neutrophil-to-high-density-lipoprotein-cholesterol ratio and mortality among patients with hepatocellular carcinoma[J]. Front Nutr. 2023;10:1127913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tian C, Wang X, Tao L, et al. Association of neutrophil to high density lipoprotein cholesterol ratio with aortic dissection and aneurysm risk: epidemiological insights from prospective cohort study based on UK biobank[J]. BMC Public Health. 2025;25(1):886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chen DL, Lin YK, Wang GJ, et al. Neutrophil to high-density lipoprotein cholesterol ratio as a potential inflammatory marker for predicting all-cause mortality in out-of-hospital cardiac arrest survivors[J]. Sci Rep. 2025;15(1):17181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jiang M, Sun J, Zou H, et al. Prognostic role of neutrophil to high-density lipoprotein cholesterol ratio for all-cause and cardiovascular mortality in the general population[J]. Front Cardiovasc Med. 2022;9:807339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chuang SM, Liu SC, Chien MN, et al. Neutrophil-to-high-density lipoprotein ratio (NHR) and neutrophil-to-lymphocyte ratio (NLR) as prognostic biomarkers for incident cardiovascular disease and all-cause mortality: a comparison study[J]. Am J Prev Cardiol. 2024;20:100869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yu L, Ma K, Hao J, et al. Neutrophil to high-density lipoprotein cholesterol ratio, a novel risk factor associated with acute ischemic stroke[J]. Medicine. 2023;102(26):e34173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Squiccimarro E, Lorusso R, Consiglio A, et al. Impact of inflammation after cardiac surgery on 30-day mortality and machine learning risk prediction[J]. J Cardiothorac Vasc Anesth. 2025;39(3):683–91. [DOI] [PubMed] [Google Scholar]
- 15.Jakob SM, Stanga Z. Perioperative metabolic changes in patients undergoing cardiac surgery[J]. Nutrition. 2010;26(4):349–53. [DOI] [PubMed] [Google Scholar]
- 16.Feguri GR, Franco AC. Postoperative Metabolic, Inflammation, and Atrial Fibrillation Control: a Relentless Battle. Braz J Cardiovasc Surg. 2019;34(2):6–7. [DOI] [PMC free article] [PubMed]
- 17.Hernandez A, Shaw AD. Age and inflammation after cardiac surgery[J]. Br J Anaesth. 2017;119(4):562–3. [DOI] [PubMed] [Google Scholar]
- 18.Nagao M, Nakajima H, Toh R, et al. Cardioprotective effects of high-density lipoprotein beyond its anti-atherogenic action[J]. J Atheroscler Thromb. 2018;25(10):985–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Barker G, Winer JR, Guirgis FW, et al. HDL and persistent inflammation immunosuppression and catabolism syndrome[J]. Curr Opin Lipidol. 2021;32(5):315–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Özkan S, Özdemir F, Uğur O, et al. The effects of the metabolic syndrome on coronary artery bypass grafting surgery[J]. Cardiovasc J Afr. 2017;28(1):48–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Başyiğit F, Çöteli C. Relationship between the neutrophil to HDL-C ratio and anatomical significance of coronary artery stenosis in patients with documented myocardial ischemia. Eur Rev Med Pharmacol Sci. 2022;26(9):3179–3184. [DOI] [PubMed]
- 22.Chen G, Yang N, Ren J, et al. Neutrophil counts to high-density lipoprotein cholesterol ratio: a potential predictor of prognosis in acute ischemic stroke patients after intravenous thrombolysis[J]. Neurotox Res. 2020;38(4):1001–9. [DOI] [PubMed] [Google Scholar]
- 23.Xu W, Wang P, Wan J, et al. Neutrophil to high-density lipoprotein cholesterol ratio predicts deep venous thrombosis in spontaneous intracerebral hemorrhage. Sci Rep. 2025;15(1):25355. [DOI] [PMC free article] [PubMed]
- 24.Lamichhane P, Agrawal A, Abouainain Y, et al. Utility of neutrophil-to-high-density lipoprotein-cholesterol ratio in patients with coronary artery disease: a narrative review[J]. J Int Med Res. 2023;51(4):3000605231166518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cuthbertson BH, Croal BL, Rae D, et al. N-terminal pro-B-type natriuretic peptide levels and early outcome after cardiac surgery: a prospective cohort study[J]. Br J Anaesth. 2009;103(5):647–53. [DOI] [PubMed] [Google Scholar]
- 26.Duchnowski P, Śmigielski W. Usefulness of myocardial damage biomarkers in predicting cardiogenic shock in patients undergoing heart valve surgery[J]. Pol Heart J. 2024;82(4):423–6. [DOI] [PubMed] [Google Scholar]
- 27.Mokhtar AT, Begum J, Buth KJ, et al. Cardiac troponin T is an important predictor of mortality after cardiac surgery[J]. J Crit Care. 2017;38:41–6. [DOI] [PubMed] [Google Scholar]
- 28.Huang JB, Chen YS, Ji HY, et al. Neutrophil to high-density lipoprotein ratio has a superior prognostic value in elderly patients with acute myocardial infarction: a comparison study[J]. Lipids Health Dis. 2020;19(1):59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Chen Y, Jiang D, Tao H, et al. Neutrophils to high-density lipoprotein cholesterol ratio as a new prognostic marker in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention: a retrospective study[J]. BMC Cardiovasc Disord. 2022;22(1):434–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Saltiel AR, Olefsky JM. Inflammatory mechanisms linking obesity and metabolic disease. J Clin Invest. 2017;127(1):1–4. [DOI] [PMC free article] [PubMed]
- 31.Huang Z, Chien K, Yang C, et al. Peripheral differential leukocyte counts in humans vary with hyperlipidemia, smoking, and body mass index[J]. Lipids. 2001;36(3):237–45. [DOI] [PubMed] [Google Scholar]
- 32.Hoofnagle AN, Vaisar T, Mitra P et al. HDL Lipids and Insulin Resistance[J]. [DOI] [PubMed]
- 33.Han CY, Tang C, Guevara ME, et al. Serum amyloid A impairs the antiinflammatory properties of HDL. J Clin Invest. 2016;126(1):266–281. [DOI] [PMC free article] [PubMed]
- 34.hdl cholesterol efflux capacity and incident cardiovascular events[Z]. [DOI] [PMC free article] [PubMed]
- 35.Anti-inflammatory. In: Deroissart J, Porsch F, Koller T, et al. editors. Handbook of Experimental Pharmacology. Cham: Springer International Publishing; 2021. pp. 359–404. and immunomodulatory therapies in atherosclerosis[M]//. [DOI] [PubMed]
- 36.Karr S. Epidemiology and management of hyperlipidemia[J]. [PubMed]
- 37.Ridker PM, Gotto AM, Lorenzatti AJ, et al. Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein[J]. New England Journal of Medicine; 2008. [DOI] [PubMed]
- 38.Abebe A, Kumela K, Belay M, Kebede B, Wobie Y. Mortality and predictors of acute kidney injury in adults: a hospital-based prospective observational study. Sci Rep. 2021;11(1):15672. [DOI] [PMC free article] [PubMed]
- 39.Squiccimarro E, Stasi A, Lorusso R, Paparella D. Narrative review of the systemic inflammatory reaction to cardiac surgery and cardiopulmonary bypass. Artif Organs. 2022;46(4):568–577. [DOI] [PMC free article] [PubMed]
- 40.Hu A, Zhang K, Sun W, et al. Combined impact of neutrophil-to-high-density lipoprotein cholesterol ratio (NHR) and cognitive function on all-cause mortality in older adults: a population-based study. Lipids Health Dis. 2025;24(1):83. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Publicly available datasets were analyzed in this study. These data can be found at [https://mimic.mit.edu/].