Association of neutrophil to lymphocyte ratio with the severity of coronary artery lesions in patients undergoing coronary angiography for the first time

Introduction

Coronary artery disease (CAD) remains the dominant cause of global cardiovascular mortality. Atherosclerotic plaque burden is driven by a chronic, low-grade inflammatory cascade that begins in adipose tissue, spills into the systemic circulation, and ultimately damages the coronary endothelium (1). Both visceral and ectopic fat depots are known to secrete pro-inflammatory cytokines, including interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), as well as reactive oxygen species—molecules that up-regulate endothelial adhesion molecules, promote leukocyte recruitment, and accelerate lipid deposition within the arterial wall (2). The Gensini score can quantify the anatomical extent of this process. However, a reproducible blood biomarker is still needed to reflect the underlying inflammatory environment, and the neutrophil-to-lymphocyte ratio (NLR) might be a potential indicator worthy of in-depth investigation.

NLR integrates innate neutrophil and adaptive lymphocyte immune activity into a single, inexpensive parameter. Experimental data link neutrophil-derived myeloperoxidase and lymphopenia-induced loss of anti-inflammatory interleukin-10 (IL-10) to endothelial dysfunction, plaque micro-calcification, and impaired collateral formation (3).

Clinically, NLR has been shown to correlate with main adverse cardiovascular events (MACEs) (4). However, there remains a paucity of research exploring the specific association between baseline NLR and the severity of CAD among hospitalized patients undergoing coronary angiography (CAG) for the first time.

Beyond visceral fat, sex-specific adipose depots also modulate vascular risk. Among pre-menopausal women, compared to higher breast density, greater evidence of adipose tissue at the breast gland level (lowest breast density, category A) is associated with higher rates of MACEs (5), which is due to same pathways, endothelial dysfunction and via over-activation of sodium-glucose cotransporter-2 (SGLT-2) receptors at level of breast adipose tissue (6,7). Over-expression of SGLT-2 in breast adipocytes amplifies local oxidative stress and down-regulates sirtuin 1 (SIRT-1), mechanisms that parallel those observed in abdominal subcutaneous fat and that may potentiate systemic neutrophil activation (7). Whether NLR captures this depot-specific inflammatory signal in first-time angiography patients has not been explored.

Therefore, we conducted a single-center retrospective analysis aiming to clarify the association between NLR and the severity of CAD measured by the Gensini score in patients undergoing CAG for the first time, providing references for the role of NLR in assessing the severity of CAD. We present this article in accordance with the STROBE reporting checklist (available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-aw-537/rc).

Methods

Research subjects

This was a retrospective study. Patients admitted to the Cardiology Department of Zhengzhou People’s Hospital from July to December 2024 who underwent CAG for the first time were consecutively selected as the research subjects. Initially, 305 patients were identified eligible. After rigorous screening, 229 cases were finally included (see Figure 1).

Figure 1 This flow chart outlines the enrollment process for patients included in the study, detailing the sequential inclusion and exclusion criteria applied to first-time CAG inpatients. ACS, acute coronary syndrome; CABG, coronary artery bypass grafting; CAG, coronary angiography; CHD, coronary heart disease; PCI, percutaneous coronary intervention.

Inclusion criteria

• Age >18 years old.
• Inpatients undergoing CAG for the first time.

Exclusion criteria

• Patients with incomplete medical record data: more than 20% of data are missing.
• Patients who had previously undergone coronary intervention or coronary artery bypass grafting (CABG), had a history of coronary heart disease (CHD), underwent emergency CAG upon admission, or were patients with acute coronary syndrome (ACS).
• Patients with a history of CAG performed at other hospitals (i.e., prior CAG outside the current institution).
• Patients with a history of or currently suffering from severe cardiovascular and cerebrovascular diseases: severe arrhythmia or history of cardiopulmonary resuscitation, structural heart disease, cardiac insufficiency [left ventricular ejection fraction (LVEF) <35%], pulmonary heart disease, pericardial effusion, acute cerebrovascular accident, pulmonary embolism, aortic dissection, infective endocarditis, etc.
• Patients who have used drugs affecting blood indices within 3 months before admission: lipid-regulating drugs (such as statins, fibrates, etc.), anti-platelet drugs, anticoagulant drugs, steroids, immunosuppressants, etc.

The determination of the severity of coronary artery lesions

In our study, the Gensini model was used mainly for the following reasons. Firstly, it can better quantify the degree of coronary artery lesions. By meticulously scoring the location and stenosis degree of lesions in different coronary artery branches, a comprehensive quantitative value can be obtained, facilitating accurate analysis and comparison in the study. Secondly, the Gensini model’s scoring system is relatively comprehensive and detailed. It takes into account all major coronary artery branches and has high accuracy in comprehensively evaluating coronary artery lesions. Compared with some simple models, it can more accurately reflect the true situation of coronary artery lesions, as its scoring basis is not only the single stenosis degree but also multiple factors such as lesion location. Additionally, in the study, it is relatively easy to conduct correlation analysis between the Gensini score and other factors like inflammatory indicators, providing a more reliable quantitative basis for exploring the relationship between coronary artery lesions and inflammation.

The results of CAG (severity of coronary artery lesions) were judged by at least two experienced senior-level physicians with comparable angiography skills, considering the results of quantitative CAG. Importantly, the physicians were blinded to patients’ NLR data to minimize observer bias. The Gensini score for the degree of coronary artery stenosis was quantitatively calculated according to the Gensini scoring comparison table (Table 1) (8,9). The Gensini score is the sum of the products of the coefficients of all coronary artery segments and the scores of the corresponding stenotic segments.

Statistical analysis

Software and variable description

All analyses were conducted using the statistical software packages R (http://www.R-project.org, The R Foundation) and Free Statistics software version 2.1. Categorical variables were presented as proportions (%), and continuous variables were described with the mean [standard deviation (SD)] or median [interquartile range (IQR)], depending on the situation. Due to non-normal distribution of variables, we used both parametric (Pearson correlation) and non-parametric (Spearman correlation) methods to ensure robustness of results. Given the violation of normality assumptions, we validated our findings using bootstrap resampling methods and reported robust standard errors. Missing data are imputed using multiple imputation.

Sample size calculation was performed based on prior literature: assuming a correlation coefficient (r) of 0.2 between NLR and Gensini score, a significance level (α) of 0.05, and a power (1−β) of 0.8, the minimum required sample size was estimated to be 193 using Zstats software. Our final enrolled sample size of 229 exceeded this estimate, ensuring sufficient statistical power to detect the hypothesized association.

Data collection and variable definition

Blood samples were collected from peripheral veins within 24 hours before angiography and at least 6 hours after acute symptom onset (to minimize acute-phase interference), with sampling performed for the first time after admission.

NLR was analyzed as both continuous and categorical variables (divided into four quartiles: Q1 <1.75, Q2 1.75–2.32, Q3 >2.32–3.48, Q4 >3.48) to explore its association with coronary lesion severity. Subgroup and curve fitting analyses were also conducted. Covariate information included age, gender, BMI, DM, hypertension, smoking status, white blood cell (WBC), platelet count (PLT), albumin (ALB), creatinine (Cr), uric acid (UA), triglyceride (TG), HDL and low-density lipoprotein (LDL). The relationships between these covariates and coronary lesion severity (represented by Gensini score) were analyzed.

Multivariable regression analysis

Multivariable linear regression models were constructed to assess the association between NLR and the Gensini score.

Model 1: adjusted for age, sex, and BMI (demographic factors).

Model 2: further adjusted for diabetes, hypertension, and smoking status (established CHD risk factors).

Model 3: additional adjustment for WBC, PLT, ALB, Cr, UA, TG, HDL, and LDL (metabolism- and inflammation-related laboratory variables).

RCS analysis with three knots was used to explore potential non-linear relationships. All statistical analyses were conducted using R version 4.3.2, with a two-sided P<0.05 considered statistically significant. No adjustment for multiple comparisons was performed due to the exploratory nature of the study. Model construction was guided by clinical experience and relevant literature (10,11).

Ethical statement

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Research Ethics Committee of Zhengzhou People’s Hospital (No. 2026-KY-001201). Written informed consent was waived due to the retrospective nature of the study and anonymization of all patient data, which posed no risk to patient privacy.

Results

Baseline characteristics of patients

A total of 229 patients who underwent CAG for the first time were enrolled in this study. Among them, 151 (65.9%) were male, and the mean age was 62.9±10.4 years. There were no significant differences in age, gender, BMI, DM, hypertension, smoking status, PLT, UA, HDL, LDL, etc. among different NLR quartile groups (P>0.05). Nevertheless, significant differences were found in WBC, ALB, Cr, TG, and Gensini score among different NLR quartile groups (P<0.05). All continuous variables are presented as mean ± SD or median (IQR), and categorical variables are presented as n (%). Specific data are shown in Table 2.

Among the 229 included cases, 109 patients underwent percutaneous coronary intervention (PCI) surgery, such as stent implantation or balloon dilation. 47 patients received conservative drug treatment, including intensive anti-platelet, lipid-regulating, and anti-angina medications. And 73 patients received health guidance, like diet adjustment and moderate exercise, before discharge after hospitalization for examination and symptomatic treatment.

Distributional plots to support assumptions underlying statistical tests

There is a significant linear relationship between NLR and Gensini score (P<0.001). The scatter plot shows an obvious linear trend. The residual plot shows no obvious non-linear pattern (see Figures 2,3).

Figure 2 This figure presents the distribution profiles, normality assessments, and quartile-stratified comparisons of the NLR and Gensini score (a metric for coronary lesion severity), to support the statistical assumptions of subsequent analyses. (A) Histogram of NLR (blue bars) with a superimposed theoretical normal distribution curve (red); NLR exhibits a non-normal distribution. (B) Histogram of Gensini score (red bars) with a superimposed theoretical normal distribution curve (green); Gensini score follows a non-normal distribution. (C) Quantile-quantile plot for NLR normality testing; the observed quantiles (blue line) deviate from the theoretical normal quantiles (red line), confirming non-normality. (D) Quantile-quantile plot for Gensini score normality testing; the observed quantiles (blue line) deviate from the theoretical normal quantiles (red line), confirming non-normality. (E) Box plot of NLR stratified by its quartiles (Q1–Q4); NLR levels increase incrementally across higher quartiles. (F) Box plot of Gensini score stratified by NLR quartiles (Q1–Q4); Gensini scores show a graded elevation with increasing NLR quartiles. NLR, neutrophil-to-lymphocyte ratio.

Figure 3 This figure presents the association between NLR and Gensini score (coronary lesion severity), along with regression model diagnostic plots. (A) Scatter plot of NLR (x-axis) and Gensini score (y-axis), with a linear regression line (β=10.64). The Pearson correlation coefficient (r=0.583, P<0.001) indicates a positive linear association. (B) Scatter plot of log-transformed NLR (x-axis) and log-transformed Gensini score (y-axis), with a linear regression line. The correlation (r=0.598, P<0.001) suggests improved linearity after transformation. (C) Residual plot (residuals vs. fitted values of Gensini score) for the regression model; residuals (mean =0.000, SD =22.20) show no obvious pattern, supporting the model’s homoscedasticity assumption. (D) Quantile-quantile plot of regression residuals; the observed quantiles (blue line) deviate from theoretical normal quantiles (red line), indicating non-normality of residuals (consistent with non-normal distributions of raw variables). NLR, neutrophil-to-lymphocyte ratio; SD, standard deviation.

Association between covariates and Gensini score

The analysis showed that elevated levels of DM [β=16.3, 95% confidence interval (CI): 6.91–25.69, P<0.001], smoking (β=10.23, 95% CI: 1.65–18.81, P=0.02), WBC (β=2.69, 95% CI: 1.16–4.23, P<0.001), and UA (β=0.04, 95% CI: 0–0.08, P=0.04) were associated with an increase in the Gensini score. In contrast, an increase in HDL (β=−20.53, 95% CI: −34.97, −6.08, P=0.006) was associated with a decrease in the Gensini score (specific data are shown in Table 3).

Distributional plots to support assumptions underlying statistical tests

Although normality assumptions were violated, the large sample size (n=229) allows application of the central limit theorem. Results from non-parametric methods confirmed the robustness of our findings (see Figures 4,5).

Figure 4 This figure presents Gensini score (coronary lesion severity) comparisons across NLR quartiles, along with variance homogeneity and distribution analyses. (A) Violin plot showing the distribution of Gensini scores across NLR quartiles (Q1–Q4). The mean Gensini score increases incrementally with higher NLR quartiles (Q1: 38.9; Q2: 40.7; Q3: 64.3; Q4: 78.3). (B) Bar plot displaying the mean (± SD) Gensini score across NLR quartiles. Sample sizes per quartile: Q1 (n=58), Q2 (n=57), Q3 (n=57), Q4 (n=57); the plot reflects the graded elevation of Gensini score with increasing NLR. (C) Bar plot of Gensini score standard deviation by NLR quartiles, with a Levene test result (P<0.001) indicating unequal variances across groups (violating the homogeneity of variance assumption). (D) CDF curves of Gensini scores for each NLR quartile. The CDF curves diverge with increasing Gensini score, indicating that higher NLR quartiles are associated with a greater cumulative probability of severe coronary lesions. CDF, cumulative distribution function; NLR, neutrophil-to-lymphocyte ratio; SD, standard deviation.

Figure 5 This figure presents data transformation effects (to improve normality) and non-parametric/robustness analyses of associations between NLR and Gensini score. (A) Histogram of log-transformed NLR (green bars) with an overlaid theoretical normal distribution curve (red); the distribution is closer to normality than raw NLR. (B) Histogram of log-transformed Gensini score (pink bars) with an overlaid theoretical normal distribution curve (green); the distribution approximates normality better than raw Gensini score. (C) Heatmap of Spearman rank correlations among key variables; NLR and Gensini score show a moderate positive correlation (ρ=0.46), consistent with prior linear analyses. (D) Distribution of bootstrap-derived correlation coefficients (1,000 resamples) between log(NLR) and log(Gensini). The mean correlation (0.583) with 95% CI (0.497–0.666) confirms the robustness of the association. BMI, body mass index; CI, confidence interval; HDL, high-density lipoprotein; NLR, neutrophil-to-lymphocyte ratio; WBC, white blood cell.

Relationship between NLR and the severity of coronary artery lesions

NLR was divided into quartiles (Q1: <1.75, Q2: 1.75–2.32, Q3: >2.32–3.48, Q4: >3.48) for trend analysis. This categorization was selected for three reasons. Firstly, it ensured a balanced sample distribution (with 56–58 participants in each quartile) to prevent statistical bias. Secondly, it was in line with the clinical thresholds of NLR (high NLR >3.0–4.0) reported in previous CAD studies (12,13). Thirdly, it had the ability to show a graded association between NLR and the Gensini score.

Without model adjustment, NLR was significantly associated with the severity of coronary artery lesions (β=3.37, 95% CI: 1.67–5.08, P<0.001). After adjusting for confounding factors in different models, NLR remained significantly associated with the severity of coronary artery lesions, albeit with variations in the association strength and P-value. When analyzed as a categorical variable, in some cases, the associations between different NLR quartile groups and cardiovascular disease (CVD) were significantly different under different adjusted models, and the trend test showed a significant trend after adjustment with models 1-3 (P<0.05) (specific data are presented in Tables 4,5).

After false discovery rate (FDR) correction for multiple comparisons, Q3 and Q4 remained significantly associated with the Gensini score (q<0.05).

Curve fitting

The results of the curve fitting analysis indicated that the overall P value was 0.02, while the P value for non-linearity was 0.38, suggesting a linear relationship between NLR and the severity of coronary artery lesions. RCS used three knots at 10th, 50th, and 90th percentiles of NLR; AIC =1,847, goodness-of-fit χ² P=0.22, supporting linearity (refer to Figure 6).

Figure 6 This figure presents a RCS curve fitting to evaluate the relationship between NLR and Gensini score (coronary lesion severity). Spline knots: set at the 10th (1.75), 50th (2.32), and 90th (3.48) percentiles of NLR. Model metrics: overall association P=0.02 (significant positive relationship); non-linearity P=0.38 (no evidence of non-linear association, supporting a linear trend); AIC =1,847, goodness-of-fit χ² P=0.22 (model fits the data well). Curve interpretation: The red curve (fitted RCS line) shows a positive trend between NLR and Gensini score; the pink shaded area represents the 95% confidence interval, indicating stable estimation of the association. AIC, Akaike information criterion; NLR, neutrophil-to-lymphocyte ratio; RCS, restricted cubic spline.

Subgroup analyses

To evaluate the robustness of the association between NLR and the Gensini score, we performed pre-specified subgroup analyses stratified by age, sex, BMI, DM, hypertension, and smoking status (Figure 7). The positive association between NLR and CAD severity was consistent across all subgroups, with no significant interaction detected (P for interaction >0.05 for all). Specifically: In patients <65 years, each unit increase in NLR was associated with a 4.32-point increment in the Gensini score (95% CI: 0.54–8.11); in patients ≥65 years, the corresponding increment was 1.50 (–1.30, 4.30). The effect estimates were similar in men and women (β= 2.51 vs. 2.28). Comparable associations were seen regardless of BMI, DM, or hypertension status. Among smokers, each unit rise in NLR led to a 3.74-point increase in the Gensini score (95% CI: 0.37–7.11), numerically higher than in non-smokers (0.59; –2.04, 3.22), but the interaction P value was 0.21.

Figure 7 This forest plot presents the subgroup analysis of the association between NLR and Gensini score (coronary lesion severity), with interaction tests for effect modification. Subgroups: Stratified by age (≤65/>65 years), gender, BMI (≤25/>25 kg/m²), DM (no/yes), hypertension (no/yes), and smoking status (no/yes). Metrics: Each entry shows subgroup sample size (n), regression coefficient (95% CI), and P value for interaction (assessing whether subgroup modifies the association). Interpretation: all interaction P values (range, 0.21–0.88) are >0.05, indicating no significant effect modification by the subgroups; the association between NLR and Gensini score is consistent across strata (note: the smoking subgroup shows a stronger positive association, though not statistically distinct). BMI, body mass index; CI, confidence interval; DM, diabetes mellitus; NLR, neutrophil-to-lymphocyte ratio.

Collectively, these data suggest that the positive relationship between NLR and angiographic CAD severity is stable across diverse clinical phenotypes in first-time CAG patients.

ROC analyses

To assess the discriminative capacity of the NLR for predicting a high Gensini score (≥50), receiver operating characteristic (ROC) curve analysis was performed. This yielded an area under the curve (AUC) of 0.723 (95% CI: 0.651–0.795) and an optimal cut-off value of 2.365 (Figure 8). We also performed 1,000-bootstrap internal validation to assess the robustness of the ROC model, with the validated AUC determined as 0.718 (95% CI: 0.650–0.786).

Figure 8 ROC curve evaluates the performance of NLR in predicting a high Gensini score (≥50, indicating severe coronary lesions). Key metrics: the AUC is 0.723 (95% CI: 0.651–0.795), indicating moderate predictive ability of NLR for severe coronary lesions. Optimal cutoff: the optimal NLR cutoff value is 2.365, corresponding to 68% sensitivity and 70% specificity (marked by the orange dot on the curve). Reference: the dashed diagonal line represents the “no-discrimination” reference (AUC =0.5). AUC, area under the curve; CI, confidence interval; NLR, neutrophil-to-lymphocyte ratio; ROC, receiver operating characteristic.

For clinical context, the optimal cut-off value of 2.365 for NLR aligns with the 75th percentile of NLR in our cohort (Q3=2.32, Q4=3.48), meaning patients with NLR >2.365 (upper 25% of the study population) had a 2.1-fold higher risk of high Gensini score (≥50) compared to those with NLR ≤2.365 (OR =2.10, 95% CI: 1.12–3.94, P=0.02). In routine clinical practice, this threshold is readily applicable: since NLR is derived from standard complete blood count (CBC) results (typically available within 1 hour of sampling), clinicians can quickly identify patients at high risk of severe CAD (Gensini ≥50) at the time of first CAG referral, prioritizing urgent diagnostic or interventional management.

Moreover, this cut-off value is consistent with the range of critical values previously determined in other studies that are associated with obstructive CAD and MACEs (14). In the future, we will continue to explore conducting multicenter prospective studies on diverse populations (such as different age groups, ethnicities, or comorbidity profiles) to confirm its generalizability across various clinical settings.

Performance metrics and residual diagnostics

The performance metrics and residual diagnostics of the multivariable linear regression models for Gensini score are shown in Table 6. R²/Adjusted R²: explain the proportion of Gensini score variance accounted for by predictors; adjusted R² corrects for variable count. Model 3’s adjusted R²=0.165 is consistent with typical explanatory power of inflammatory/metabolic markers in observational studies. F-statistic/P-value: Confirm overall model validity; all P<0.001 indicate significant linear associations between predictors and Gensini score. Residual normality: Shapiro-Wilk P>0.05 supports normal residual distribution, meeting linear regression assumptions. Residual homoscedasticity: Breusch-Pagan P>0.05 confirms homogeneous residual variance, ensuring unbiased coefficient estimates.

Discussion

Atherosclerosis is a lipid-accumulation and chronic inflammatory disease, and chronic inflammation can drive the progression of CAD (15,16). The NLR is an easily accessible and cost-effective biomarker of systemic inflammation, which can be used to predict the prognosis of CVDs (17,18). Oliva et al. demonstrated that baseline NLR is associated with adverse cardiovascular outcomes in CAD patients undergoing PCI (4). However, its direct association with the severity of de novo coronary artery lesions remains unclear, as most prior studies enrolled heterogeneous cohorts including patients with a history of revascularization or established CAD, which obscures the link between baseline immunity and atherosclerosis (19). We conducted this study in an attempt to fill this knowledge gap to some extent. Among 229 patients undergoing first-time CAG, there was a significant linear correlation between baseline NLR and Gensini score (P for non-linearity =0.38).

Regarding patients’ baseline characteristics, no significant differences were observed in age, gender, BMI, DM, hypertension, smoking status, PLT, UA, HDL or LDL across NLR quartile groups. However, white WBC count, ALB, Cr, TG, and Gensini score differed significantly among these quartiles. These variations likely reflect confounding effects—for instance, elevated WBC may directly drive higher NLR, while alterations in ALB, Cr, and TG may stem from shared systemic perturbations—rather than direct pathological associations between NLR and these indices. Notably, after adjusting for these covariates in multivariable models, NLR remained significantly associated with Gensini score (β=2.39, 95% CI: 0.28–4.51, P=0.02), confirming its robust link to coronary lesion severity.

In our analysis of covariates associated with Gensini score, type 2 diabetes mellitus (T2DM), smoking, elevated WBC count, and increased UA correlated with higher Gensini scores, while higher HDL was linked to lower scores. Notably, 28.4% of the cohort (65/229) had T2DM, and this subgroup exhibited a strong positive association with Gensini score (β=16.3, 95% CI: 6.91–25.69, P<0.001). This finding underscores that T2DM is robustly linked to more severe coronary artery lesions, consistent with its established role as a key driver of atherosclerotic progression (20). From a statistical perspective, post-hoc power was ≥80% for age <65 years, male, and DM subgroups—supporting the reliability of these subgroup analyses—while other subgroup assessments are considered exploratory due to wider CIs that limit definitive conclusions.

Indeed, the worse glycemic control is a negative prognostic factor in this class of patients. Studies by Sardu et al. (21,22) suggest that the potential mechanisms underlying the adverse prognostic impact in these patients with poor glycemic control may be multifactorial. Hyperglycemia can lead to endothelial dysfunction, increased oxidative stress, and abnormal platelet activation. All of these contribute to a pre-thrombotic state and impair vascular repair. In the context of PCI, these factors may further exacerbate the formation and development of thrombi, increasing the risk of adverse cardiovascular events. Additionally, poor glycemic control may also affect the body’s inflammatory response, intensifying the inflammatory processes associated with acute coronary syndromes.

In a retrospective cohort study (14), over 2 million adult patients with type 2 diabetes were included, among whom 552,000 (27%) were using SGLT-2 inhibitors. SGLT2i could ameliorate clinical outcomes in patients admitted with acute myocardial infarction alone (23), with this benefit amplified when combined with glucagon-like peptide-1 receptor agonists (GLP-1 RAs) (24); GLP-1 RA monotherapy yields similar cardiovascular advantages (24). These data align with our finding that baseline NLR—an inexpensive, widely available inflammatory biomarker—parallels the anatomic severity of coronary lesions. This convergence suggests that combined pharmacologic (SGLT-2i/GLP-1 RAs) and biomarker-guided (NLR) anti-inflammatory approaches could further enhance prognosis in diabetic patients with CAD.

Regarding the association between NLR and coronary lesion severity, a significant correlation was observed in the unadjusted model. After accounting for confounders across sequential multivariable models, the association remained significant, albeit with modest changes in effect size and P-values. When NLR was analyzed as a categorical variable (quartiles), distinct associations with CAD were observed in select adjusted models, and trend tests confirmed a significant graded relationship across all three model iterations. Notably, curve-fitting analyses supported a linear correlation between NLR and coronary lesion severity.

In this study, we took measures to exclude the confounding effects of acute stress responses, such as excluding patients with acute coronary syndrome who had stress-induced aggravated inflammation. Acute infections, which were included in severe comorbidities, and medication interference, through restricting pre-admission drug use, were also excluded. This approach ensured that the final study cohort was minimally influenced by these confounding factors, thus enhancing the internal validity of the study results. However, there are still some limitations in this study. It uses a single-center retrospective design and lacks longitudinal outcome data. Despite adjusting for multiple covariates, residual confounding may persist from unaccounted factors that could influence the NLR-coronary lesion severity association. Additionally, the current study has limitations regarding diagnostic performance: the NLR showed moderate sensitivity (68%) and specificity (70%) for predicting severe coronary lesions (Gensini score ≥50), which may restrict its standalone utility in clinical practice. This highlights the need for combining NLR with other biomarkers or imaging parameters to enhance diagnostic accuracy.

Future research should validate NLR’s prognostic value via multi-center prospective studies, explore its utility in combination with SGLT-2 inhibitors/GLP-1 receptor agonists based on cardiometabolic evidence, and investigate its role in guiding invasive versus conservative management decisions for first-time CAG patients.

Conclusions

This study demonstrates that NLR is significantly associated with the severity of coronary artery lesions in hospitalized patients undergoing CAG for the first time.

The detection of NLR is simple, rapid and cost-effective, and can provide reference for clinical decision-making and risk stratification to a certain extent. However, the elevation of NLR may be influenced by multiple factors. Therefore, in clinical application, NLR can be combined with existing methods to provide index reference for the clinic, so as to achieve complementarity with existing methods.

Acknowledgments

We sincerely thank the Clinical Information Center and Imaging Department of Zhengzhou People’s Hospital, especially Yuanqiao Shi, Shida Li, and Jiahui Zhang, for providing clinical data support. We would like to express our gratitude to Dr. Jie Liu (Department of Vascular and Endovascular Surgery, Chinese PLA General Hospital) for his consultation on the study design, statistical support, and comments on the manuscript. We also thank Fei Gao (Indiana University School of Medicine) for her assistance with the language polishing of this manuscript.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-aw-537/rc

Data Sharing Statement: Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-aw-537/dss

Peer Review File: Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-aw-537/prf

Funding: This study was supported by the Joint Fund of Science and Technology R&D Program of Henan Province (No. 222103810054), Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2023ZD0503502), International Registration Research Project (Nos. ChiCTR2200061611, and ChiCTR1900027448), and Zhengzhou Science and Technology Benefiting Plan Project (No. 2022KJHM0035). The funders had no role in the research design, data collection and analysis, decision to publish, or the writing of the manuscript.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-aw-537/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Research Ethics Committee of Zhengzhou People’s Hospital (No. 2026-KY-001201). Written informed consent was waived due to the retrospective nature of the study and anonymization of all patient data, which posed no risk to patient privacy.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Waksman R, Merdler I, Case BC, et al. Targeting inflammation in atherosclerosis: overview, strategy and directions. EuroIntervention 2024;20:32-44. [Crossref] [PubMed]
2. Sardu C, Pieretti G, D'Onofrio N, et al. Inflammatory Cytokines and SIRT1 Levels in Subcutaneous Abdominal Fat: Relationship With Cardiac Performance in Overweight Pre-diabetics Patients. Front Physiol 2018;9:1030. [Crossref] [PubMed]
3. Cheng D, Talib J, Stanley CP, et al. Inhibition of MPO (Myeloperoxidase) Attenuates Endothelial Dysfunction in Mouse Models of Vascular Inflammation and Atherosclerosis. Arterioscler Thromb Vasc Biol 2019;39:1448-57. [Crossref] [PubMed]
4. Oliva A, Vogel B, Sartori S, et al. Association of Neutrophil-to-Lymphocyte Ratio With Clinical Outcomes After Percutaneous Coronary Intervention. Eur J Prev Cardiol 2025;zwaf597. [Crossref] [PubMed]
5. Sardu C, Gatta G, Pieretti G, et al. Pre-Menopausal Breast Fat Density Might Predict MACE During 10 Years of Follow-Up: The BRECARD Study. JACC Cardiovasc Imaging 2021;14:426-38. [Crossref] [PubMed]
6. Sacks D, Baxter B, et al. Multisociety Consensus Quality Improvement Revised Consensus Statement for Endovascular Therapy of Acute Ischemic Stroke. Int J Stroke 2018;13:612-32. [Crossref] [PubMed]
7. Sardu C, Gatta G, Pieretti G, et al. SGLT2 breast expression could affect the cardiovascular performance in pre-menopausal women with fatty vs. non fatty breast via over-inflammation and sirtuins’ down regulation. Eur J Intern Med 2023;113:57-68. [Crossref] [PubMed]
8. Gai MT, Yan SQ, Wang MY, et al. Comparison of Gensini score and SYNTAX score for predicting in-stent restenosis in patients with coronary artery disease and drug-eluting stent implantation. Sci Rep 2025;15:1077. [Crossref] [PubMed]
9. Zhang Y, Luo R, Sun X, et al. The relationship between Gensini score and rates of 30-day mortality in acute coronary syndrome patients in China. BMC Cardiovasc Disord 2025;26:3. [Crossref] [PubMed]
10. Li H, Chen M, Wang Y, et al. The Predictive Value of TyG Index and NLR for Risk of CHD and the Severity of Coronary Artery Lesions in Patients with Type 2 Diabetes Mellitus. J Inflamm Res 2024;17:11813-28. [Crossref] [PubMed]
11. He Y, Li L, Zhou T, et al. Association Between Inflammation Indices Derived From Complete Blood Count and Coronary Artery Calcification. J Inflamm Res 2025;18:3807-16. [Crossref] [PubMed]
12. Hou BQ, Chandrashekar AS, Jamal NH, et al. Admission Neutrophil-to-Lymphocyte Ratio Is Superior to WBC Count at Predicting the Presence and Severity of Pediatric Musculoskeletal Infection. J Bone Joint Surg Am 2025;107:868-77. [Crossref] [PubMed]
13. Shen J, Zhou J, Xue F, et al. Association between neutrophil-lymphocyte ratio and all-cause and cardiovascular mortality in osteoarthritis patients from the NHANES 1999-2018 cohort. Sci Rep 2025;15:34061. [Crossref] [PubMed]
14. Marfella R, Sardu C, D'Onofrio N, et al. SGLT-2 inhibitors and in-stent restenosis-related events after acute myocardial infarction: an observational study in patients with type 2 diabetes. BMC Med 2023;21:71. [Crossref] [PubMed]
15. Song B, Bie Y, Feng H, et al. Inflammatory Factors Driving Atherosclerotic Plaque Progression New Insights. J Transl Int Med 2022;10:36-47. [Crossref] [PubMed]
16. Tasouli-Drakou V, Ogurek I, Shaikh T, et al. Atherosclerosis: A Comprehensive Review of Molecular Factors and Mechanisms. Int J Mol Sci 2025;26:1364. [Crossref] [PubMed]
17. Antognoli L, Sabbatinelli J, Rosa MD, et al. Admission Neutrophil-to-Lymphocyte Ratio Predicts Short- and Long-Term Mortality in Hospitalized Older Adults: A Retrospective Multicenter Study. Aging Dis 2025; Epub ahead of print. [Crossref]
18. Wu ZM, Chen S, Dai Y, et al. Circulating HMGB2 Is Associated With Angiographic Coronary Collateralization in Diabetic Patients With Chronic Total Occlusion and Impairs Vessel Growth via NLRP3. Circ J 2025;89:1245-55. [Crossref] [PubMed]
19. Bao Q, Liu T, Song H, et al. Prognostic Role of Inflammatory Hematologic Indices in Predicting Acute Coronary Syndrome in Elderly Patients with Chronic Coronary Syndrome. J Inflamm Res 2025;18:9637-53. [Crossref] [PubMed]
20. Guan H, Zhao S, Fang X, et al. Frontier technologies for investigating endothelial heterogeneity and function in diabetic vascular disease: An updated review. Biomed Pharmacother 2025;191:118445. [Crossref] [PubMed]
21. Sardu C, D'Onofrio N, Mauro C, et al. Thrombus Aspiration in Hyperglycemic Patients With High Inflammation Levels in Coronary Thrombus. J Am Coll Cardiol 2019;73:530-1. [Crossref] [PubMed]
22. Sardu C, Barbieri M, Balestrieri ML, et al. Thrombus aspiration in hyperglycemic ST-elevation myocardial infarction (STEMI) patients: clinical outcomes at 1-year follow-up. Cardiovasc Diabetol 2018;17:152. [Crossref] [PubMed]
23. Marfella R, Prattichizzo F, Sardu C, et al. GLP-1 receptor agonists-SGLT-2 inhibitors combination therapy and cardiovascular events after acute myocardial infarction: an observational study in patients with type 2 diabetes. Cardiovasc Diabetol 2024;23:10. [Crossref] [PubMed]
24. Marfella R, Sardu C, Balestrieri ML, et al. Effects of incretin treatment on cardiovascular outcomes in diabetic STEMI-patients with culprit obstructive and multivessel non obstructive-coronary-stenosis. Diabetol Metab Syndr 2018;10:1. [Crossref] [PubMed]