The Association of Systemic Inflammatory Response Index and Neutrophil-to-High-Density Lipoprotein Ratio Mediated by Fasting Blood Glucose with 90-Day Prognosis in Acute Ischemic Stroke Patients Open Access

Stroke is one of the leading causes of disability and death worldwide [1]. Following acute ischemic stroke (AIS), the immune system elicits a neuroinflammatory response in the brain that produces irreversible damage to neurocytes [2, 3]. Numerous studies have documented that various subtypes of leukocytes, such as monocytes, neutrophils, and lymphocytes, are involved in the inflammatory response [4, 5]. The neutrophil-to-lymphocyte ratio (NLR), a universally recognized biomarker, has been reported to be significantly correlated with a poor prognosis in patients with AIS [6‒8]. Notwithstanding, it still has its limitations; for instance, it is merely a ratio of two types of white blood cells. A prior study from our laboratory team determined that systemic inflammatory response index (SIRI), an inflammatory index, calculated as monocyte counts × NLR, is independently associated with the 90-day prognosis in patients with AIS who have undergone intravenous thrombolysis [9]. In addition, high-density lipoprotein (HDL) has been established to be closely correlated with the development of ischemic stroke [10, 11]. Neutrophil to high-density lipoprotein ratio (NHR), an indicator composed of the ratio of peripheral blood neutrophils to HDL level, is a novel potential inflammatory marker that can reflect endothelial dysfunction, atherosclerosis, and thrombosis [5, 12‒14].

A large number of AIS patients experience transient elevations in blood glucose levels, which may be a stress response to acute illness or an adaptation of inflammation to ischemic injury [15]. In addition to neuroinflammation, fasting blood glucose (FBG) level is positively associated with the risk of stroke [16, 17]. Specifically, high FBG levels are associated with poor clinical and mortality outcomes in patients with AIS [18‒20]. Of note, hyperglycemia has been shown to be associated with elevated levels of inflammatory markers [21, 22]. Thus, patients with a high neuroinflammatory response may exhibit high FBG levels, which may adversely affect stroke prognosis. The aforementioned findings imply that FBG levels play a mediating role in the relationship between neuroinflammation and stroke prognosis. However, the relationship between inflammation, FBG levels, and AIS prognosis has not been established so far. Therefore, the aim of this study was to investigate the relationship between SIRI, NHR, and the prognosis of patients with AIS and further examine the mediating role of FBG.

This study enrolled patients with AIS from the Nanjing Stroke Registry between January 2017 and July 2021. Eligible participants meet the following criteria: (1) 18 years or older; (2) a diagnosis of AIS according to the World Health Organization criteria, with confirmation of stroke lesion via brain magnetic resonance imaging; (3) admission within 48 h of onset. The exclusion criteria were as follows: (1) Incomplete data on fasting glucose level and monocyte count; (2) patients with inflammatory diseases or infectious diseases; (3) mRS score >2 before onset; (4) lost to follow-up 90 days after onset; (5) patients who underwent thrombolysis or other interventions. After excluding patients who did not meet the inclusion criteria, a total of 2,828 patients were analyzed (Fig. 1).

On the first day of admission, the demographic characteristics (age and gender), presence of vascular risk factors [23‒25] (history of smoking and alcohol, diabetes mellitus [DM], hypertension [HT], atrial fibrillation [AF], hyperlipidemia [HL], and chronic kidney failure), medical history (use of antiplatelets and statins), and laboratory data of patients were collected. Laboratory data consisted of HDL levels, platelet counts, neutrophil counts, monocyte counts, lymphocyte counts, triglyceride (TG) levels, FBG levels, SIRI, and NHR.

The smoking history was defined as ≥1 cigarettes per day for more than 6 months or meeting previous smoking criteria and having quit for less than 6 months [26]. The history of drinking alcohol was defined as >4 drinks in a day or >14 drinks a week in men; >3 drinks a day or >7 drinks a week in women [26]. DM is diagnosed by taking hypoglycemic agents or FPG ≥7.0 mmol/L or 2-h PG ≥11.1 mmol/L during OGTT or HbA1c ≥6.5% [27]. HT is diagnosed by taking antihypertensive agents or blood pressure consistently >130/80 mm Hg (1 mm Hg = 0.133 kPa) [28]. The diagnostic criteria for HL included the use of lipid-regulating medications, or meeting any of the following thresholds: TC ≥6.2 mmol/L, LDL ≥4.1 mmol/L, TG ≥2.3 mmol/L, or HDL <1.0 mmol/L [29]. The diagnostic criteria for AF are the use of anticoagulants or a 12-lead electrocardiogram indicating AF after admission [30]. The diagnostic criteria for chronic renal insufficiency (CRF) are the use of prescribed kidney medication or renal impairment ≥3 months or glomerular filtration rate (GFR) <60 mL/min/1.73 m² persisting for 3 months [31].

Blood samples from AIS patients fasting for at least 8 h within 24 h of admission were collected, and the glucose levels and blood cell counts were determined. The latter included neutrophil counts, lymphocyte counts, monocyte counts, and HDL levels. Next, the cell counts were employed to calculate composite inflammatory biomarkers. In other words, NLR was calculated as neutrophil counts/lymphocyte counts, SIRI was calculated as monocyte counts × NLR, and NHR was calculated as neutrophil counts/HDL levels.

Stroke severity was assessed using the National Institutes of Health Stroke Scale (NIHSS) upon admission. The primary clinical outcomes of enrolled patients were assessed by the mRS score at the 90-day follow-up visit and comprised good outcomes, poor outcomes, and death. A good outcome was defined as an mRS score between 0 and 2, whereas a poor outcome was represented by a score of 3–6 [32]. A total of 2,828 AIS patients were included in the final analysis. Approximately 1,952 patients manifested good outcomes, while 845 patients experienced poor outcomes. Among patients with poor outcomes, 31 patients died within 90 days.

This study enrolled 2,828 patients with AIS. NHR, SIRI, and FBG levels were measured and inputted into the database. Data were expressed as mean ± SD, IQR. Logistic regression analysis was adopted to determine the relationships between NHR (mean ± SD, IQR) and SIRI (mean ± SD, IQR) and poor outcomes. The Mann-Whitney test or t test was used to compare continuous variables based on data distribution, whereas χ² tests were used to compare categorical data. The Spearman correlation test was used to examine the association between each indicator and the mRS score. Risk variables with a p value below 0.05 in the univariate analysis were incorporated into the multivariate analysis. The results were presented as odds ratios (ORs) and 95% confidence intervals (CIs). The relationship among baseline NHR, SIRI, and the functional outcome was fitted by smooth curve fitting, and reverse stepwise logistic regression analysis was used to evaluate ORs and the corresponding 95% CIs. Mediation hypotheses of FBG on the relationship between inflammatory cytokines, which included NHR and SIRI, and poor outcomes were tested using the bias-corrected bootstrap method with 2,828 samples to calculate CIs (95%). p < 0.05 was considered statistically significant. An indirect effect was regarded as significant when the CI did not include zero. All statistical analyses were performed using the Statistical Package for the Social Sciences 25.0 (SPSS; IBM, USA) and the Free Statistics analytic platform. A two-sided p < 0.05 was deemed statistically significant.

A total of 3,653 AIS patients were enrolled in this study from January 2017 to July 2021. Roughly 825 patients were excluded for the following reasons: incomplete data on fasting glucose levels and monocyte counts (n = 143); patients suffering from inflammatory diseases or infectious diseases (n = 121); mRS score >2 prior to onset (n = 105); lost to follow-up 90 days after onset (n = 139); patients underwent thrombolysis or other intervention (n = 317). Finally, a total of 2,828 patients were included in the present study.

The baseline characteristics of the included patients (according to the NHR quartiles) are detailed in Table 1. Compared to the low NHR group, patients with higher NHR were more likely to suffer from diabetes, HT, AF, and CRF.

In the univariate regression analysis, SIRI and NHR were significantly associated with poor outcomes at 90 days (p < 0.001). Model I was adjusted for gender and age (OR [95% CI] = 1.17 (1.12–1.23), p < 0.001). Model II was adjusted for gender, age, platelet count, TOAST classification, OCSP classification, admission NIHSS scores, and history of HT, diabetes, AF, CRF, smoking, and alcohol (OR [95% CI] = 1.07 (1.02–1.11), p = 0.005). FBG was introduced in model II to construct model III (OR (95% CI) = 1.07 (1.02–1.11), p = 0.005) (Table 2). Multivariate logistic regression analysis was carried out, and the results exposed that high SIRI and NHR values were independently associated with a poor prognosis at 3 months. The results remained consistent after the addition of adjusted covariates insinuating that both SIRI and NHR were positively associated with poor outcomes.

Similarly, logistic regression analyses were performed for patients who died within 3 months (Table 3). As anticipated, the analyses demonstrated that SIRI was positively correlated with mortality outcomes. When NHR was introduced into the univariate regression analysis, there was a significant association between NHR and mortality outcomes, but this statistical significance weakened as the number of covariates was increased. We speculate that this may be ascribed to the fact that the small number of patients experiencing death during the 90-day follow-up period might have compromised the statistical results. However, the overall trend indicated a positive linear correlation between SIRI and NHR with mortality outcomes. In short, univariate and multivariate logistic regression analyses determined that high SIRI and NHR (p < 0.05) were independent risk factors for a poor prognosis at day 90.

Restricted cubic spline plots were drawn to assess the relationship between SIRI, NHR, and poor outcomes at 90 days. Further restricted cubic spline analysis uncovered a positive linear correlation between SIRI, NHR, and the risk of poor outcomes (p for nonlinearity = 0.132, 0.143) (Fig. 2).

Patients were stratified into subgroups according to gender, age, and history of alcohol and smoking, HT, DM, AF, HL, CRF, and Oxfordshire Community Stroke Project (OCSP) classification. As displayed in Figure 3, the influence of different subgroups on stroke prognosis was consistent, and there was no significant difference between different groups.

Following quartiles of SIRI and NHR, the FBG levels of each group were compared (Fig. 4). The observations unveiled that SIRI and NHR values were positively correlated with FBG levels.

Figure 5 depicts the relative total, direct, and indirect effects of the mediating role of FBG on the relationship between inflammatory cytokines and poor outcomes at 90 days in the mediation models. Our mediation hypothesis was validated, considering that bootstrapping revealed significant relative indirect effects for poor outcomes (indirect effect = 0.0038, 95% CI 0.001–0.008; direct effect = 0.1719, 95% CI 0.1258–0.2179), indicating that FBG mediated the association between SIRI and poor outcomes. The association between NHR and poor prognosis also exhibited a mediating effect, which was estimated at approximately 4.8%. These mediating effects are also presented in Figure 5.

The results of this observational single-center cohort study demonstrated that SIRI and NHR were positively associated with 90-day poor clinical and mortality outcomes in AIS patients. Furthermore, mediation analyses revealed that FBG partially mediates the relationship between SIRI, NHR, and stroke prognosis within 90 days of follow-up in patients with AIS.

There is currently a paucity of articles on the associations between SIRI and NHR and stroke outcomes. It is worthwhile emphasizing that SIRI and NHR are novel composite inflammatory indicators [33, 34]. Compared with traditional inflammatory indicators, the former had been shown to be related to the prevalence of metabolic syndrome [35]; the latter can be used to distinguish three different inflammatory responses and reflect the inflammatory state [36]. Our previous study has established a relationship between SIRI and stroke outcomes [9], but the value of SIRI was only explored in AIS patients who underwent intravenous thrombolysis, and the sample size was relatively small. Herein, the value of SIRI and NLR in AIS patients, as well as the mediating effect of FBG, was explored in a large sample size study.

The interactions between SIRI, NHR, and FBG levels that ultimately lead to short-term poor outcomes in AIS patients remain elusive. We hypothesize that they are associated with the following phenomenon: endothelial cells are a core component of the vascular system [37]. According to prior cellular and animal experiments, vascular endothelial cells exposed to hyperglycemic conditions activate the nuclear factor κB signaling pathway, which promotes the expression of inflammatory factors, eventually leading to disruptions in the endothelial barrier, lipid accumulation, and vascular stenosis [38], which in turn limits blood supply to the brain tissue. Under ischemic and hypoxic conditions, glycolysis is enhanced in endothelial cells to meet the needs of energy and material metabolism. Nevertheless, excessive glucose cannot be satisfactorily metabolized, resulting in the synthesis of metabolites that damage the extracellular matrix and aggravate inflammatory cell infiltration into the vascular wall, ultimately triggering AIS [39]. Noteworthily, the time of inflammatory cells adhering to vascular endothelium was significantly longer in hyperglycemic mice compared to normoglycemic model mice, so vascular ischemic injury was aggravated, which triggered a series of inflammatory reactions [40]. These findings collectively provide insights into the underlying pathology involving inflammation and fasting glucose levels in stroke prognosis. Nonetheless, further large-scale studies are warranted to validate our findings.

However, the present study has several potential limitations that cannot be overlooked. First of all, this study was based on single-center retrospective data, but a large sample size was included to minimize the risk of selection bias. Second, the study did not collect data on oral hypoglycemic drugs taken prior to admission. Given that the main focus was on the FBG level at admission, we postulate that the effect of hypoglycemic agents on the results was reflected by FBG values. Furthermore, SIRI and NHR were not dynamically monitored, which could have offered insights into their dynamic changes during emergency hospital admission. Further research is necessitated to unravel the clinical significance of the relationship between systematic inflammation, fasting blood glucose levels, and functional outcomes in patients with AIS.

In summary, our study revealed that high SIRI and NHR were positively correlated with the risk of poor clinical and mortality outcomes in AIS patients. Moreover, the association of SIRI and NHR with poor outcomes at 90 days and mortality could potentially be partially mediated by FBG.

This study was approved by the Ethics Committee of the Affiliated Hospital of Nanjing University of Chinese Medicine (2017NL-012-01). Verbal informed consent was obtained from participants or their legal representatives.

The authors have no conflicts of interest to declare.

This work was supported by National Natural Science Foundation of China (Grant No. 82274428, 81973794) and Jiangsu Province Administration of Chinese Medicine (ZT202102); Project of National Clinical Research Base of Traditional Chinese Medicine in Jiangsu Province, China (JD2023SZ), to Yuan Zhu; and leading talents of Traditional Chinese Medicine of Jiangsu Province (SLJ0201) and Peak Academic Talent Project of Jiangsu Province Hospital of Chinese Medicine (y2021rc01) to Zhuyuan Fang. This publication was made possible by support from Brain Center in Jiangsu Province Hospital of Chinese Medicine.

Aimei Zhang and Yuan Zhu were mainly involved in study design, data analysis, data interpretation, and manuscript preparation. Junqi Liao, Dan Wu, and Xiaohui Yan were mainly involved in data acquisition and data analysis. Jingyi Chen, Qiuhua He, Fantao Song, Yunze Li, and Li were mainly involved in data acquisition. Zhaoyao Chen and Wenlei Li were mainly involved in data analysis and manuscript preparation. Qin Yang, Yuanzhu Fang, and Minghua Wu were mainly involved in study design, data interpretation, and manuscript preparation. The authors read and approved the final manuscript.

Aimei Zhang and Yuan Zhu contributed equally to this work.

The data that support the findings of this study are available from the corresponding author upon reasonable request.