Introduction

Stroke is a common acute cerebrovascular disease in the clinical practice, which is caused by the sudden rupture of blood vessels in the brain or the obstruction of cerebral blood vessels resulting in blood not flowing into the corresponding areas of the brain, thus causing damage to the corresponding brain tissues. Currently, stroke has become the second leading cause of death worldwide [1]. In the latest global statistics on heart disease and stroke, an average of 1 in 21 deaths died of stroke [2]. Meanwhile, U.S. stroke prevalence also increased by 7.8% from 2011–2013 to 2020–2022[3]. In recent years, the diagnosis and treatment of stroke has made significant progress, but the number of stroke deaths is still on the rise [4]. How to improve the diagnosis and treatment of stroke is an important part of improving the occurrence and development of stroke.

Inflammation is a biological phenomenon that is recognized as a series of reactions, processes, or states in the body that regulate homeostasis[5]. Numerous studies show that cellular and molecular mediators in the inflammatory response are generally involved in a wide range of biological processes, including tissue remodeling, metabolism, thermogenesis, and neurological function[6]. Modern medicine often uses some quantitative tests to reflect the overall or local inflammatory situation in the body, such as white blood cell(WBC), hypersensitive C-reactive protein(hs-CRP), interleukin-6(IL-6) and other indicators. Systemic immune-inflammation index(SII) is a new type of composite inflammatory response index that can be used to predict inflammation, with good stability, and can assess the local or systemic immune response[7]. Recent studies have shown that SII is not only be used as a prognostic factor in cancer, but also has a certain predictive value for cerebrovascular diseases [8,9,10].

There is a growing body of evidence suggesting that inflammatory markers are closely associated with stroke[11]. Lei Zheng et al[12] showed a positive correlation between the level of cadmium metal in urine and plasma CRP levels at dose concentrations, and concluded that CRP played a 10.1% role in the association between cadmium and stroke through mediated analysis. This suggests that CRP, an inflammatory response marker, plays an associative role in cadmium and stroke, which may lead to stroke through this pathway. Valerian L Altersberger et al[13] also found that WBC was an important independent predictor of intravenous thrombolysis (IVT) in stroke patients, and higher WBC and CRP were positively associated with higher NIHSS, which could assess patient prognosis. These studies show that inflammatory markers can effectively reflect brain tissue damage to a certain extent. However, the association between SII, a novel composite index of inflammatory response, and stroke has not been well studied.

As a result, we investigate the association between SII and stroke through a cross-sectional analysis on data from investigating participants in the National Health and Nutrition Examination Survey (NHANES).

Method

Study population

The NHANES is a representative survey of the US national population conducted by the Centers for Disease Control and Prevention (CDC) that uses a sophisticated, multistage, and probabilistic sampling methodology to provide a wealth of information about the general US population’s nutrition and health[14]. All study procedures were authorized by the National Center for Health Statistics’ ethical review board prior to data collection, and all participants gave their signed, informed consent. The entirety of NHANES data is accessible to the general public and can be freely obtained from: https://www.cdc.gov/nchs/nhanes/index.htm (accessed on 8 May 2024).

In this study, cross-sectional data of 25,531 participants from three consecutive cycles of the NHANES (2015–2020) were initially included. Participants with complete stroke and SII data were included in the study analysis. The exclusion criteria were set as follows: (1) participants without SII data (n = 5264); (2) participants with no stroke status data (n = 6980). After manual data filtering, we ultimately selected a total of 13,287 participants for subsequent analyses. A detailed flow chart of study participant recruitment was presented in Fig. 1.

Fig. 1
figure 1

Flowchart of participant selection. Abbreviations: NHANES, National Health and Nutrition Examination Survey; SII, Systemic immune-inflammation index.

Full size image

SII

The systemic immune-inflammatory index is a novel composite inflammatory response index, which is the exposure variable in our study. In our study, we employed Lymphocyte, neutrophil, and platelet counts when calculating the SII. All indicators were measured by complete blood count using automated hematology analyzing devices (Coulter®DxH 800 analyzer) and presented as ×103 cells/ml. The SII was calculated by the following formula[8,15,16]:

$$\text{SII} = \text{platelet count} {\times} \text{neutrophil count}/\text{lymphocyte count}.$$

Stroke

Stroke was assessed by whether or not you have been informed of the occurrence of a stroke in the medical conditions questionnaire. “Have you ever been told by a doctor or health professional that you had a stroke?” If participants answered “yes” to the question, they were judged to have had a stroke[17,18]. By making this determination, we considered stroke as an outcome indicator.

Covariates

Being based on previous publications and biological considerations, we collected as many covariates with known confounding effects on stroke as possible[14]. Covariates were obtained by standardized questionnaires and face-to-face interviews, including age, gender, Body Mass Index (BMI), race, education level, ratio of family income-to-poverty ratio(PIR), smoking status, drinking status, high blood pressure, diabetes, and coronary heart disease. Among them, Race/ethnicity were divided into five categories: non-Hispanic White, non-Hispanic Black, other Hispanic, Mexican American, and other races. Education levels were divided into three categories: below high school, high school, and above high school. Body mass index (BMI), calculated as weight in kilograms (kg) divided by the square of height in meters (m2), is widely used for estimating overweight/obesity status. Participants who smoked over 100 cigarettes throughout their lifetime were defined as smokers[19]. High blood pressure, diabetes, and coronary heart disease are assessed by the responses in the medical conditions questionnaire.

Statistical analysis

The statistical study was carried out using the statistical computing and graphics software R (version 4.1.3) and EmpowerStats (version:2.0). Baseline tables for the study population were statistically described by stroke and the SII quartiles. The association between SII and stroke was analyzed by multivariate linear regression and the outcome was represented by OR [95%CI]. Model 1 did not adjust for any covariates; Model 2 adjusted for age, gender, and race; Model 3 adjusted for all covariates. By adjusting the variables, smoothed curve fits were done simultaneously. A threshold effects analysis model was used to examine the relationship and inflection point between SII and stroke. Finally, the same statistical study methods described above were conducted for the gender subgroups. It was determined that P < 0.05 was statistically significant.

Results

Baseline characteristics of the participants

The study involved 13,287 participants with an average age of 50.15 ± 17.55 years. Among these participants, 48.21% were men, 51.79% were women, 34.19% were non-Hispanic white, 23.83% were non-Hispanic black, 14.15% were Mexican American, 11.64% were other Hispanic, and 16.19% were from other race. Of them, 611 participants (4.598%) had a history of stroke. The mean SII ± SD concentrations were 518.14 ± 343.98 (1,000 cells/µl).

Table 1 lists all clinical characteristics of the participants with stroke as a column-stratified variable. Compared to non-stroke individuals, stroke patents tended to be older (65.12 years vs. 49.81 years), smokers (58.92% vs. 41.05%), less educated (26.51% vs. 20.35%), and more likely to have high blood pressure (74.30% vs. 36.13%), diabetes(36.99% vs. 17.17%), Coronary heart disease (18.82% vs. 3.59%), and the higher SII levels (590.81 vs. 514.64). Moreover, PIR (1.98 vs. 2.18) and alcohol status (2.10 vs. 2.39) were all lower in the stroke group.

Table 1 Baseline characteristics of the study participants grouped by stroke status.
Full size table

Table 2 lists all clinical characteristics of the participants with the quartiles of SII. Among all participants, the prevalence of stroke increased with the higher SII level. The mean SII ± SD concentrations were 518.14 ± 343.98 (1,000 cells/µl), with the values for the different quartiles as follows: quartile 1: < 315.58; quartile 2: 315.58–448.00; quartile 3: 448.00–628.43; and quartile 4: ≥ 628.43 (1,000 cells/µl). Compared to those with the lowest quartile of SII, individuals with the highest quartile of SII were more likely to be older, women, non-Hispanic white, BMI, and smoking status, and they were more likely to have high blood pressure, and diabetes(P < 0.05).

Table 2 Baseline characteristics of the study participants grouped by SII quartiles.
Full size table

Association of SII with stroke

Because the effect value is not apparent, SII/100 is used to amplify the effect value by 100 times. Table 3 shows the multivariate regression analysis between SII/100 and stroke. In the crude [1.04(1.03, 1.06)] and partially adjusted [1.03(1.02,1.05)] models, the SII and stroke show a significant positive association. Upon complete adjustment, the aforementioned positive association remained statistically significant [1.02(1.01, 1.04)]. This positive association remained stable after transforming the SII into quartiles (all P for trend < 0.05). Participants in the highest SII quartile had a 71% increased prevalence of stroke compared to those in the lowest quartile [1.71(1.34, 2.18)].

Table 3 The association between SII and stroke .
Full size table

Subgroup analyses

Table 4 shows subgroup analyses and interaction tests stratified by age, gender, race, education level, smoking status, high blood pressure, diabetes, and coronary heart disease. Our outcomes showed that the relationship between the SII and stroke was not dependent on the above factors (all P for interaction > 0.05).

Table 4 Subgroup analysis for the association between SII and Stroke.
Full size table

In subgroup analyses stratified by gender, our results suggest that the positive association between SII and stroke is significant in women [1.04(1.01, 1.06)], but not statistically significant in fully adjusted model for men, as shown in Table 5. Further, we performed a smooth curve fit to describe the nonlinear relationship between the SII and stroke (Figs. 2 and 3). After adjusting for variables: age, gender, race, education level, smoking status, alcohol status, BMI, PIR, high blood pressure, diabetes, and coronary heart disease, our findings suggest that there exists a nonlinear relationship between SII and stroke with an inflection point of 740.00(1000 cells/µl) by using a two-stage linear regression model. After stratifying the analysis by gender, an approximate inverted U-shaped curve was also present in men and women, with inflection points of 772.20 (1,000 cells/µl) and 3551.18 (1,000 cells/µl), respectively (Table 6).

Table 5 The association between SII and stroke in subgroup analyses stratified by gender.
Full size table
Fig. 2
figure 2

The association between SII and stroke.The solid red line represents the smooth curve fit between variables.Blue bands represent the 95% Confidence Interval from the fit.

Full size image
Fig. 3
figure 3

The association between SII and stroke stratified by gender.

Full size image
Table 6 Threshold effect analysis of SII on stroke using a linear regression model.
Full size table

Discussion

Among the 13,287 NHANES participants included in our study, we observed that individuals with stroke had a clearly higher mean SII level than those without stroke. Based on this phenomenon, we confirmed that higher SII was associated with a higher risk of stroke. Then, the results of the subgroup analyses and interaction testing indicated that this connection was similar across population. A nonlinear relationship between SII and stroke was also discovered, with an inflection point of 740.00(1000 cells/µl). After stratifying the analysis by gender, an approximate inverted U-shaped curve was also present in men and women, with inflection points of 772.20 (1,000 cells/µl) and 3551.18 (1,000 cells/µl), respectively. This indicated that SII was an independent crisis factor for stroke when the SII was less than 740.0(1000 cells/µl).

Nowadays, a growing number of studies have found that SII plays an important measure of chronic inflammation, and is significantly associated with cardiovascular-related diseases [20,21]. For example, Dong W et al[22] revealed that SII is an independent risk factor for coronary heart disease. Cao Y et al[23] found that a positive association between SII and all-cause mortality and cardiovascular disease in a cross-sectional study of 8524 participants. Liao M et al[24] also reported that carotid atherosclerosis was a risk factor for ischemic stroke, which was significantly positively correlated with SII. Jin N et al[25] showed that a nonlinear relationship between SII and hypertension, with each standard deviation increase in SII increasing the prevalence of hypertension by 9%. In addition, Cheng W et al[26] found an association between higher levels of SII and increased stroke prevalence in patients with asthma, suggesting SII as a potential predictor of stroke in patients with asthma. In our study, we found a positive correlation between SII and stroke, which is consistent with the poor prognosis of SII and cardiovascular-related disease described in earlier studies.

In recent years, several studies have confirmed that inflammation is an important risk factor for stroke and is involved in its main pathological process [27]. A large number of previous studies have demonstrated the effect of NLR and PLR on the assessment of inflammation in stroke. SII is different from those indexes that includes platelet count, neutrophil count and lymphocyte count. SII can comprehensively reflect the three pathways of thrombosis, inflammatory response and adaptive immune response, and provide a more comprehensive assessment of inflammatory response[28–32]. In SII, neutrophil count and lymphocyte count represent the state of inflammation in the blood and platelet count reflects the condition of thrombosis. For example, Cai W et al[33] found that neutrophil constitution in peripheral blood increased soon after stroke onset of patients, and higher neutrophil count indicated detrimental stroke outcomes. Misirlioglu NF et al[34] elevated NLR and SII has been linked to worse outcomes in the context of stroke, including increased risk of stroke severity, larger infarct size, and higher mortality rates. Also, Du J et al[35] demonstrated that platelet count increases the risk for ischemic stroke.

Much evidence has shown that a single inflammatory marker is associated with stroke and reflect the immediate inflammatory state of the body. But, they alone are highly susceptible to the influence of their environment and lifestyle. SII is not the same as a single inflammatory marker. It weakens the instability of the change of a single index through the ratio, which reflects the immune inflammation of the patient’s body as a whole. What’s more, SII can assess the risk factors for stroke-related conditions, such as hypertension, diabetes, coronary artery disease, and atrial fibrillation[36–39]. SII is more accurate than traditional risk factors in assessing coronary artery conditions that can predicts cardiovascular events after coronary intervention[40]. Consistent with our findings, SII is associated with stroke prognosis. Therefore, SII can be used as a comprehensive, accurate and simple indicator to reflect inflammation in cardiovascular diseases, which deserves the attention of healthcare professionals. Inflammation is one of the main pathological mechanisms of stroke. When stroke occurs, local brain tissue is damaged, the blood-brain barrier is destroyed, neutrophils accumulate, inflammatory mediators are released, and lymphocytes undergo apoptosis, which in turn causes secondary damage to stroke and promotes neurological deterioration[41–43]. At the same time, the release of inflammatory factors from the injury site can also lead to abnormal platelet function, which can be directly activated and aggregated, resulting in the blood clots’ formation and stroke recurrence[44]. After stroke, its multiple complications are also related to inflammation [45, 46]. For example, post-stroke immunosuppression is a key factor in the development of stroke-related pneumonia [47]. Following the onset of stroke, a series of inflammatory responses are produced, resulting in immune deficiencies such as decreased peripheral lymphoid count, decreased monocyte activity, and decreased phagocytic activity [48]. Inflammatory response is also one of the main research hotspots in the pathological mechanism of post-stroke depression(PSD)[49]. When the pro-inflammatory factors increase during stroke, the inflammatory response is aggravated, which can affect the hypothalamic-pituitary-adrenal axis, reduce the brain-derived neurotrophic factor (BDNF), and enhance the neurocytotoxicity[50–52]. Finally, it promotes the occurrence of PSD[53]. Stroke can also trigger a cascade of reactions, further aggravating the inflammatory response. That causes damage to brain tissue structure and neurological function. Stoll G et al[54] found that platelets would guide lymphocytes to the site of vascular injury during cerebral ischemia, prompting activated platelets to form thrombus. Then, Cognitive function of brain tissue is impaired. As important indicators of inflammation, SII and NLR can indirectly assess the severity of neurological impairment in patients, which is valuable in the prediction of cognitive function after stroke[55].

Our study has some limitations. First, this is a cross-sectional analysis; thus, temporality cannot be ascertained. Furthermore, despite adjusting several relevant confounders, we were unable to rule out the impact of additional confounding factors; therefore, our findings should be interpreted with caution. Third, due to the limitations of the NHANES database, the covariates of this study did not include participants’ medications use; therefore, our findings may not fully reflect the true situation. Despite these limitations, our study has several advantages. Because we used a nationally representative sample, our study is representative of a multiethnic and gender-diverse population in the United States. In addition to this, the large sample size included in our study allowed us to perform a subgroup analysis.

Conclusion

Our findings imply that increased SII levels are linked to stroke. To confirm our findings, more large-scale prospective investigations are needed.