Abstract

Objectives:

This study is designed to assess the association between the systemic immune-inflammation index (SII) and all-cause mortality among patients with ischemic stroke (IS).

Methods:

A single-center retrospective cohort study was conducted, including 1,156 IS patients discharged from Dandong Central Hospital from January to December 2024. The formula for SII is as follows: SII = (neutrophil count × platelet count)/lymphocyte count. Multivariate Cox regression model, subgroup analysis, sensitivity analysis, receiver operating characteristic (ROC) curve, and Kaplan–Meier survival analysis were used to evaluate the association between SII and all-cause mortality.

Results:

During the median follow-up period of 14.23 months, a total of 97 (8.4%) patients with all-cause mortality were identified. Multivariate Cox regression analysis showed that after adjusting for a variety of confounding factors, for every one-standard-deviation increase in SII, the risk of all-cause mortality increased by 17.2% [hazard ratio (HR) = 1.172, 95% confidence interval (CI): 1.019–1.348, p = 0.026]. Moreover, the risk of all-cause mortality in the high SII group was 1.543 times that of the low SII group (HR = 1.543, 95% CI: 1.017–2.340, p = 0.041). Multiple subgroup analyses and sensitivity analyses reverified the stability of these results. ROC curve analysis indicated that SII had a certain predictive value for all-cause mortality (overall population, AUC = 0.605; male, AUC = 0.609; female, AUC = 0.600), and the predictive value of SII was higher than that of platelet-to-lymphocyte ratio (PLR) and monocyte to high-density lipoprotein cholesterol ratio (MHR) (overall population, AUC of PLR = 0.579; AUC of MHR = 0.487). Furthermore, SII combined with NIHSS score and mRS score could improve its predictive value for all-cause mortality in patients with IS (overall population, AUC = 0.683). The Kaplan–Meier survival curve analysis revealed significant differences in the cumulative risk of all-cause mortality among different SII groups, and the cumulative risk of all-cause mortality was higher in the high SII group (p < 0.05).

Conclusion:

Higher levels of SII were found to be significantly associated with an elevated risk of all-cause mortality in IS patients.

1 Introduction

Ischemic stroke (IS) is a major public health problem worldwide, with significant variations in incidence, mortality, and disability rates across different regions and populations. According to the Global Burden of Disease Study, although age-standardized incidence and mortality rates are on a downward trend, the absolute number of IS cases is still increasing due to population aging and the widespread presence of high-risk factors (1). The high mortality risk of IS is closely related to multiple factors. Studies have shown that cardiogenic embolic stroke accounts for approximately one-fourth of all IS, and its severity and disability rate are higher than those of non-cardiogenic strokes (2). In addition, the presence of chronic diseases such as heart disease, diabetes, and hypertension significantly increases the risk of death after stroke (3). In young patients, coronary heart disease is considered a major predictor of new vascular events and death after IS (4). These research results emphasize the importance of managing cardiovascular risk factors after stroke. The risk factors for IS are diverse, including non-modifiable (age, gender, race) and modifiable (hypertension, smoking, diabetes, hyperlipidemia) indicators (5), with established subtype-specific risk profiles in young adults (6). Moreover, recent studies have also revealed the impact of inflammation, environmental pollution and low-density lipoprotein cholesterol (LDL-C) on the incidence of IS (1, 7). The systemic immune-inflammation index (SII), an emerging inflammatory biomarker integrating neutrophil, platelet, and lymphocyte counts, has drawn widespread attention in recent years for its crucial clinical significance in the incidence, progression, risk assessment, and prognosis of stroke (8). Specifically, in a systematic review and meta-analysis, investigators discovered that elevated SII levels were significantly correlated with suboptimal functional outcomes, increased mortality rates, and the incidence of hemorrhagic transformation among stroke patients (9). In the acute IS subgroup, elevated SII levels were independently associated with 90-day poor outcomes in patients receiving intravenous thrombolysis, and SII was significantly positively correlated with the admission National Institutes of Health Stroke Scale (NIHSS) score, suggesting its predictive value for mortality risk (10). Among hospitalized elderly stroke patients (≥ 60 years old, including ischemic and hemorrhagic stroke), non-survivors had significantly higher SII levels than survivors, and multivariate analysis confirmed that SII is an independent predictor of in-hospital all-cause mortality (11). However, in patients with intracerebral hemorrhage (ICH) (n = 320), the predictive efficacy of SII did not surpass that of the neutrophil-to-lymphocyte ratio (NLR), and current evidence does not support its use as an independent predictive indicator for mortality risk in ICH patients (12). It should be noted that although several studies [e.g., in populations with asthma, hypertension, acute myocardial infarction (AMI), and cardiorenal metabolic (CKM) syndrome] have shown an association between SII and stroke prevalence or cardiovascular mortality, they did not directly assess the mortality endpoint in stroke patients (13–16). In contrast, studies on the progression of cerebral small vessel disease (CSVD) suggest that SII may be involved in stroke-related pathological processes (e.g., progression of CSVD burden, new cerebral microbleeds) (17, 18), which are indirectly associated with prognosis. Moreover, SII has exhibited high diagnostic and predictive utility in gauging the disease severity of patients with large-artery atherosclerotic stroke (19). Compared to the NLR, SII demonstrated greater diagnostic efficacy (19). Beyond this, SII not only excels in predicting the severity of stroke but also presents independent predictive value in evaluating the risk of stroke-associated pneumonia (SAP). For example, stroke patients with higher SII levels are more likely to develop SAP, indicating that SII can be utilized as an early-detection tool for SAP, which enables clinicians to implement appropriate intervention strategies in a timely manner (20). In addition, the applications of SII in other cardiovascular diseases offer valuable references for its use in the context of stroke. For example, in patients with acute myocardial infarction and peripartum cardiomyopathy, SII has been validated as a key metric for prognosis assessment, and it can effectively predict the risk of in-hospital mortality and the recovery of left ventricular function (21, 22).

However, few studies have systematically compared the discriminatory ability of SII with other inflammatory composite indices such as platelet-to-lymphocyte ratio (PLR) and monocyte to high-density lipoprotein cholesterol ratio (MHR) for all-cause mortality in IS patients, nor have they explored the incremental value of combining SII with core clinical prognostic markers including NIHSS and modified Rankin Scale (mRS) scores. Collectively, these findings underscore that SII, as a readily obtainable blood parameter, plays a pivotal role in the incidence, progression, risk assessment, and prognosis of stroke. Against this backdrop of existing research, the present study endeavored to assess the correlation between SII and all-cause mortality in IS patients, and to determine the optimal cut-off value of SII for characterizing the discriminatory ability for all-cause mortality in this population. The objective was to provide novel insights and a theoretical foundation for the prevention, diagnosis, and treatment strategies targeting inflammation in IS patients, as well as an objective and quantifiable threshold for clinical risk stratification.

2 Methods2.1 Study design and participants

This single-center, retrospective cohort study was carried out at Dandong Central Hospital. The subjects of this study were discharged IS patients. Between January 2024 and December 2024, 1,156 subjects were consecutively recruited according to the following inclusion and exclusion criteria. The relevant flow chart was presented in Figure 1.

Flow chart of enrollment for the study population of IS. IS, ischemic stroke.

Inclusion criteria: (1) Patients diagnosed with IS by computed tomography (CT) or magnetic resonance imaging (MRI); (2) Patients with complete baseline clinical characteristics, demographic data, and laboratory test results; (3) Patients admitted within 24 h of symptom onset.

Exclusion criteria: (1) Patients with incomplete baseline data; (2) Patients admitted more than 24 h after the onset of symptoms; (3) Patients with severe hematological diseases; (4) Patients with severe hepatic or renal dysfunction; (5) Patients with advanced malignant tumors; (6) Patients with severe autoimmune diseases; (7) Patients receiving hormone or anti-inflammatory treatment; (8) Patients with missing blood test indices; (9) Patients who died during hospitalization; (10) Patients lost to follow-up. This study protocol was approved by the Ethics Committee of Dandong Central Hospital (Approval No. DDSZXYY-2025-48) and was conducted in accordance with the principles of the Declaration of Helsinki. As this was a retrospective study and all data were analyzed anonymously, the requirement for written informed consent from patients was reviewed and waived by the Ethics Committee of Dandong Central Hospital.

2.2 Data collection and variable definitions

Within the electronic medical record system of our hospital, data encompassing demographics, anthropometrics, comorbidities, prior medication history, and blood biomarkers were collected for all patients included in this study. Demographic data incorporated patient age, sex, smoking, and drinking status. Smoking was defined as a history of regular smoking or current active smoking (regardless of subsequent cessation). Drinking was defined as a history of regular alcohol intake or ongoing consumption (regardless of abstinence attempts). Anthropometric measurements included height, weight, body mass index (BMI), systolic blood pressure (SBP), and diastolic blood pressure (DBP). BMI was calculated as weight (in kilograms) divided by the square of height (in meters).

Comorbid conditions included hypertension, diabetes, hyperlipidemia, previous cerebral infarction, and the duration of cerebral infarction. Hypertension was defined as a well-established prior diagnosis of hypertension, an SBP of ≥ 140 mmHg or a DBP of ≥ 90 mmHg during the current hospitalization, or ongoing antihypertensive medication treatment (23). Diabetes was defined as a definite prior diagnosis of diabetes, a fasting blood glucose (FBG) level of ≥ 7.0 mmol/L, a glycated hemoglobin (HbA1c) level of ≥ 6.5%, current receipt of antidiabetic medication, or a blood glucose level of ≥ 11.1 mmol/L detected 2 h after an oral glucose tolerance test (24). Hyperlipidemia was defined as a definite prior diagnosis of hyperlipidemia, a triglyceride level of ≥ 2.3 mmol/L, a total cholesterol (TC) level of ≥ 6.2 mmol/L, or a LDL-C level of ≥ 4.1 mmol/L during the current hospitalization (25). A history of cerebral infarction was defined as a documented prior diagnosis of cerebral infarction. The duration of cerebral infarction was recorded as the time elapsed from the initial onset of cerebral infarction symptoms to the current hospitalization. Prior medication history involved information on patients’ previous use of antihypertensive drugs, antidiabetic agents, lipid-lowering medications, anti-platelet drugs, among others. Blood biomarker data included white blood cell count (WBC), neutrophil count, lymphocyte count, monocyte count, hemoglobin concentration, platelet count, FBG, uric acid, estimated glomerular filtration rate (eGFR), homocysteine, triglycerides, TC, LDL-C, and high-density lipoprotein cholesterol (HDL-C). Additionally, we calculated two other inflammatory composite indices for comparative analysis: PLR, defined as platelet count divided by lymphocyte count; and MHR, defined as monocyte count divided by HDL-C level, with all cell count units expressed as ×109/L. All these biomarkers were obtained through a standardized procedure. Specifically, trained nursing staff in our hospital collected venous blood samples from the antecubital veins of patients who had fasted for at least 8 h. All blood samples were derived from the patients’ first routine hematological tests after admission (with fasting time ≥ 8 h), which were generally completed within 24 h of hospital admission. From a clinical background and practical perspective, the patients included in this study were mainly in the acute or subacute phase of IS, as they were admitted within 24 h of symptom onset. These samples were then promptly transferred to the central laboratory of our hospital, where laboratory technicians conducted measurements following established protocols. Baseline stroke severity was assessed using the NIHSS score, which was evaluated by certified neurologists at admission. Pre-stroke residual functional impairment was assessed using the mRS score, which was collected through medical record review and patient/family interview to reflect the functional status before the index stroke onset. Both NIHSS score and pre-stroke mRS score were fully collected, verified for data integrity, and included as key covariates in all statistical analyses.

2.3 Definition and grouping of systemic immune-inflammation index

The SII is calculated using peripheral blood cell counts and serves to reflect the integrated state of systemic immunity and inflammation. The formula for SII is as follows: SII = (neutrophil count × platelet count)/lymphocyte count (26). Herein, all cell count units are expressed as ×109/L. Two different thresholds were used to stratify the study subjects into high and low SII groups: (1) Median-based grouping: The median value of SII in the study population (620.81) was used for grouping, with low SII group (SII ≤ 620.81, n = 578) and high SII group (SII > 620.81, n = 578), to ensure approximately equal sample size between the two groups and avoid selection bias caused by significant sample size disparity. (2) Optimal cut-off value-based grouping: The optimal cut-off value of SII for characterizing the discriminatory ability for all-cause mortality in IS patients was determined by the maximum Youden index (Youden index = sensitivity + specificity − 1) derived from the ROC curve. It should be noted that this cut-off value was obtained through exploratory analysis, and the ROC curve was not used to establish a clinical predictive model. The optimal cut-off value was determined to be 986.83, with low SII group (SII ≤ 986.83, n = 851) and high SII group (SII > 986.83, n = 305).

2.4 Follow-up and outcome indicators

The starting point of the follow-up was the date of the patient’s discharge. The ending point of the follow-up was determined as the time of all-cause mortality or September 2025, with precedence given to the earlier event. We defined all-cause mortality during the follow-up period with the follow-up endpoint set from patient discharge to September 2025, and this was primarily based on the following considerations. First, the present study focused on the long-term mortality risk of stroke patients after discharge, which is more closely aligned with clinical follow-up practices and the prognostic assessment of IS patients in clinical settings. Second, in this study, we had already excluded patients who died during hospitalization. Therefore, if the follow-up were to start from the onset of stroke, it would inevitably include patients who died in hospital, making it impossible to analyze the prognostic findings related to long- and medium-term mortality in out-of-hospital settings. Third, our choice of using the discharge date as the starting point for follow-up is also consistent with the approach reported in the majority of relevant studies. The follow-up strategy involved multiple approaches. First, we meticulously examined the outpatient and emergency department records, as well as the inpatient medical files, of patients who had made several visits to our hospital after being discharged. Additionally, telephone follow-up was carried out. Through these telephone contacts, we gathered follow-up information regarding the patients’ post-discharge status from either the patients themselves or their family members. The key outcome measure in this study was all-cause mortality, which was defined as death triggered by any disease or incident. Based on the occurrence of all-cause mortality, the study cohort was partitioned into two groups: the all-cause mortality group (n = 97) and the group with non-all-cause mortality (n = 1,059).

2.5 Statistical analysis

All statistical analyses were conducted using SPSS 27.0 software (IBM Corporation, Armonk, New York, USA). Categorical variables were presented as frequencies (percentages). The chi-square test or Fisher’s exact test was employed to determine the differences between two groups. For all continuous variables, normality was examined using the Shapiro–Wilk test. Given that all continuous variables did not conform to a normal distribution, they were expressed as the median (interquartile range), and the Mann–Whitney U test was utilized to assess the differences between groups. For the handling of missing data, this study was a retrospective analysis based on the hospital electronic medical record system. During participant enrollment, we excluded patients with missing parameters required for SII calculation (including neutrophil count, platelet count, and lymphocyte count) and those lost to follow-up. We ensured no missing data for the independent and dependent variables, and complete data were also available for demographic characteristics, anthropometric measures, comorbidities, and medication history. However, a small number of individual hematological indicators had minor missing values, with a missing rate of less than 5% for each variable. We therefore used mean imputation, a conventional simple imputation method commonly adopted in clinical research, to address these limited missing data. In summary, the vast majority of variables in this study had no missing values, and the few remaining missing data were imputed using the clinically common mean imputation approach. Univariate Cox regression analysis was carried out to evaluate the correlation of each variable with all-cause mortality. Subsequently, variables with a p value less than 0.05 were selected to establish three multivariate Cox regression models with incremental adjustment for confounding factors: Model 1: Adjusted for age, baseline NIHSS score, and pre-stroke mRS score; Model 2: Further adjusted for hyperlipidemia, a history of previous cerebral infarction, lipid-lowering medications, and anti-platelet drugs on the basis of Model 1; Model 3 (fully adjusted model): Additionally adjusted for SBP, DBP, FBG, uric acid, and eGFR on the basis of Model 2. The significant association between SII and all-cause mortality was further evaluated within these three multivariate Cox regression models. Next, ten subgroups were formed based on five variables: age, smoking status, drinking, diabetes, and a history of previous cerebral infarction. The stratified association between SII and all-cause mortality was re-evaluated in the fully adjusted model. In the sensitivity analysis, patients without hypertension and those without hyperlipidemia were separately excluded. The correlation between SII and all-cause mortality was then re-verified among these subsets of patients. Descriptive and exploratory receiver operating characteristic (ROC) curve analysis was performed for the following purposes: (1) To assess and compare the discriminatory ability of SII, PLR and MHR for all-cause mortality in the overall population, male subgroup and female subgroup; (2) To evaluate the incremental discriminatory value of combining SII with NIHSS score, mRS score, or both for all-cause mortality. The area under the curve (AUC) with 95% CI was calculated for each ROC curve. We explicitly state that the ROC curve and AUC values were not applied for the development or validation of a formal clinical predictive model, and only served to descriptively characterize the discriminatory performance of biomarkers for the study endpoint. Finally, the Kaplan–Meier survival curve with log-rank test was applied to evaluate the differences in the cumulative risk of all-cause mortality among different SII groups. All statistical tests were two-tailed, and a p value less than 0.05 was considered to indicate a statistically significant difference.

3 Results3.1 Baseline characteristics grouped by the median of SII

The baseline characteristics grouped by the median of SII were presented in Table 1. The total sample size was 1,156 individuals. The median age was 70 years (interquartile range: 64.00–77.00), among which 645 were male, accounting for 55.8% of the total population. Based on the median of SII, the subjects were divided into two groups: the low SII group (n = 578) and the high SII group (n = 578). When compared with the low SII group, the high SII group exhibited a higher prevalence of antihypertensive drugs use, higher levels of BMI, WBC, neutrophil count, platelet count (p < 0.05). Conversely, the high SII group had a lower prevalence of smoking history, a shorter duration of cerebral infarction, and lower lymphocyte counts (all p < 0.05). Notably, the high SII group had a significantly higher pre-stroke mRS score than the low SII group (p = 0.010), while there was no statistically significant difference in baseline NIHSS score between the two groups (p = 0.180), the high SII group also had a significantly higher proportion of all-cause mortality (10.7% vs. 6.1%, p = 0.004). However, other variables such as age, sex, drinking, hypertension, diabetes, hyperlipidemia, and previous history of cerebral infarction did not show statistically significant differences between the two groups (p > 0.05).

VariablesTotal populationLow SIIHigh SIIp valueN1,156578578Age, years70.00 (64.00, 77.00)70.00 (63.00, 76.00)71.00 (64.00, 78.00)0.082Sex, n (%)0.214Male645 (55.8%)333 (57.6%)312 (54.0%)Female511 (44.2%)245 (42.4%)266 (46.0%)Smoking, n (%)435 (37.6%)237 (41.0%)198 (34.3%)0.018Drinking, n (%)342 (29.6%)183 (31.7%)159 (27.5%)0.122Hypertension, n (%)1,048 (90.7%)518 (89.6%)530 (91.7%)0.225Diabetes, n (%)521 (45.1%)250 (43.3%)271 (46.9%)0.214Hyperlipidemia, n (%)1,070 (92.6%)536 (92.7%)534 (92.4%)0.823Previous cerebral infarction, n (%)626 (54.2%)301 (52.1%)325 (56.2%)0.157Course of cerebral infarction, day1.00 (0.19, 3.38)2.00 (0.25, 5.00)1.00 (0.19, 3.00)<0.001Antihypertensive drugs, n (%)821 (71.0%)393 (68.0%)428 (74.0%)0.023Antidiabetic drugs, n (%)420 (36.3%)200 (34.6%)220 (38.1%)0.221Lipid-lowering drugs, n (%)1,066 (92.2%)534 (92.4%)532 (92.0%)0.826Antiplatelet drugs, n (%)1,041 (90.1%)525 (90.8%)516 (89.3%)0.376BMI, kg/m224.54 (22.49, 27.06)24.49 (22.49, 26.99)24.69 (22.49, 27.34)<0.001SBP, mmHg151.00 (139.00, 166.00)150.00 (138.00, 163.00)151.00 (138.00, 166.00)0.202DBP, mmHg84.00 (78.00, 91.00)84.00 (77.00, 90.00)85.00 (78.00, 91.00)0.184WBC, x109/L6.93 (5.73, 8.66)6.12 (5.21, 7.28)7.87 (6.63, 9.55)<0.001Neutrophil count, x109/L4.78 (3.71, 6.23)3.79 (3.11, 4.62)6.01 (4.90, 7.75)<0.001Lymphocyte count, x109/L1.48 (1.08, 1.89)1.74 (1.39, 2.21)1.19 (0.92, 1.61)<0.001Monocyte count, x109/L0.37 (0.29, 0.48)0.37 (0.30, 0.46)0.38 (0.29, 0.49)0.429Hemoglobin, g/L138.00 (127.00, 149.00)139.00 (128.00, 150.00)137.00 (126.00, 148.00)0.053Platelet count, x109/L198.00 (166.00, 238.00)188.00 (155.00, 216.00)218.00 (181.00, 255.00)<0.001FBG, mmol/L5.70 (5.00, 7.30)5.60 (5.00, 7.18)5.74 (5.00, 7.43)0.417Uric acid, μmol/L325.00 (266.00, 389.75)320.00 (266.00, 381.75)335.00 (265.00, 400.00)0.223eGFR, mL/min/1.73m2101.49 (80.64, 122.41)102.70 (85.29, 121.16)99.85 (77.01, 123.96)0.090Homocysteine, μmol/L12.60 (10.50, 15.80)12.60 (10.40, 15.60)12.80 (10.50, 15.80)0.279Triglycerides, mmo/L1.34 (0.97, 1.91)1.36 (1.00, 1.92)1.33 (0.97, 1.85)0.052Total cholesterol, mmo/L4.43 (3.65, 5.23)4.43 (3.58, 5.15)4.42 (3.69, 5.29)0.693LDL-C, mmol/L2.71 (2.10, 3.32)2.70 (2.08, 3.28)2.73 (2.12, 3.35)0.792HDL-C, mmol/L1.13 (0.95, 1.32)1.11 (0.95, 1.33)1.14 (0.86, 1.32)0.424mRS score3.00 (1.00, 4.00)2.00 (1.00, 3.00)3.00 (1.00, 4.00)0.010NIHSS score2.00 (1.00, 4.00)2.00 (1.00, 4.00)2.00 (1.00, 4.00)0.180All-cause mortality, n (%)0.004Yes97 (8.4%)35 (6.1%)62 (10.7%)No1,059 (91.6%)543 (93.9%)516 (89.3%)

Baseline characteristics grouped by median of SII.

SII, Systemic-immune inflammation index; BMI, Body mass index; SBP, Systolic blood pressure; DBP, Diastolic blood pressure; WBC, White blood cell count; FBG, Fasting blood glucose; eGFR, Estimated glomerular filtration rate; LDL-C, Low-density lipoprotein cholesterol; HDL-C, High-density lipoprotein cholesterol; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.

3.2 Baseline characteristics grouped by all-cause mortality

As presented in Table 2, subjects were categorized into two groups according to the occurrence of all-cause mortality: the non-all-cause mortality group (n = 1,059) and the all-cause mortality group (n = 97). In comparison with the non-all-cause mortality group, the all-cause mortality group demonstrated a higher age, a greater proportion of previous cerebral infarction, and elevated levels of SBP, DBP, WBC, neutrophil count, FBG, uric acid, and homocysteine (p < 0.05). Conversely, the all-cause mortality group had a lower proportion of hyperlipidemia, a shorter duration of cerebral infarction, a lower utilization rate of lipid-lowering drugs and anti-platelet drugs, as well as reduced levels of lymphocyte count, hemoglobin, and eGFR (p < 0.05). The all-cause mortality group had significantly higher baseline NIHSS scores (p < 0.001) and pre-stroke mRS scores (p < 0.001) than the non-all-cause mortality group. More significantly, the level of SII was markedly higher in the all-cause mortality group (p < 0.001). Nevertheless, other variables including sex, smoking status, drinking, hypertension, and diabetes did not exhibit statistically significant differences between the two groups (p > 0.05).

VariablesNon all-cause mortalityAll-cause mortalityp valueN1,05997Age, years70.00 (63.00, 76.00)76.00 (70.00, 83.00)<0.001Sex, n (%)0.505Male594 (56.1%)51 (52.6%)Female465 (43.9%)46 (47.4%)Smoking, n (%)401 (37.9%)34 (35.1%)0.584Drinking, n (%)312 (29.5%)30 (30.9%)0.762Hypertension, n (%)957 (90.4%)91 (93.8%)0.264Diabetes, n (%)473 (44.7%)48 (49.5%)0.361Hyperlipidemia, n (%)989 (93.4%)81 (83.5%)<0.001Previous cerebral infarction, n (%)561 (53.0%)65 (67.0%)0.008Course of cerebral infarction, days1.00 (0.21, 4.00)1.00 (0.13, 2.50)0.021Antihypertensive drugs, n (%)751 (70.9%)70 (72.2%)0.795Antidiabetic drugs, n (%)384 (36.3%)36 (37.1%)0.867Lipid-lowering drugs, n (%)986 (93.1%)80 (82.5%)<0.001Antiplatelet drugs, n (%)971 (91.7%)70 (72.2%)<0.001BMI, kg/m224.60 (22.51, 27.04)23.44 (21.48, 28.55)0.371SBP, mmHg151.00 (139.00, 165.00)151.00 (140.00, 172.00)0.031DBP, mmHg84.00 (78.00, 90.00)89.00 (74.00, 97.00)0.042WBC, ×109/L6.79 (5.70, 8.33)7.68 (5.48, 9.99)<0.001Neutrophil count, ×109/L4.69 (3.68, 5.93)5.77 (3.65, 7.98)<0.001Lymphocyte count, ×109/L1.50 (1.11, 1.89)1.31 (0.96, 1.72)0.002Monocyte count, ×109/L0.37 (0.29, 0.46)0.39 (0.29, 0.41)0.124Hemoglobin, g/L139.00 (128.00, 149.00)128.00 (113.00, 143.00)<0.001Platelet count, ×109/L199.00 (168.00, 235.00)200.00 (151.00, 245.00)0.496FBG, mmol/L5.61 (5.00, 7.17)6.30 (5.40, 8.59)0.001Uric acid, μmol/L324.50 (265.25, 385.00)363.00 (274.00, 454.00)0.024eGFR, mL/min/1.73m2101.21 (82.27, 121.86)88.77 (52.73, 115.59)<0.001Homocysteine, μmol/L12.50 (10.40, 15.50)14.90 (11.81, 17.70)0.003Triglycerides, mmol/L1.34 (0.99, 1.87)1.32 (0.96, 1.97)0.465Total cholesterol, mmo/L4.42 (3.63, 5.18)4.74 (4.00, 5.53)0.266LDL-C, mmol/L2.70 (2.09, 3.28)3.04 (2.53, 3.70)0.199HDL-C, mmol/L1.13 (0.95, 1.33)1.14 (1.03, 1.25)0.825mRS score3.00 (1.00, 4.00)3.00 (2.00, 4.00)<0.001