Association Between PM2.5 Components and Cerebral Hypoperfusion: A Mediation Role of Inflammatory Biomarkers Authors: Xingyan Xu, Xinyue Lu, Nan Wang, Liuyan Zheng, Jianhui Guo, Tongjin Xu, Yiming Guo, Hao Li, Yi Yang, Le Yang Published: 2026 ABSTRACT ABSTRACTBackgroundCerebral hypoperfusion, an acute manifestation of intracranial arterial stenosis (ICAS), is particularly concerning among older East Asian adults given its high prevalence. Systemic inflammation triggered by PM2.5 exposure has been linked to cerebrovascular diseases, yet evidence on its specific components of hypoperfusion in this vulnerable group remains limited. We explore associations between PM2.5 and its components exposure with cerebral hypoperfusion, and assess mediation by inflammatory indicators.MethodsWe analyzed 967 patients with symptomatic ICAS at the First Affiliated Hospital of Soochow University. Hypoperfusion was evaluated using multimodal CT imaging and quantified with MIStar software. Three‐year exposure to PM2.5 and its components was estimated using the TAP database. Associations were evaluated using generalized additive models and restricted cubic spline regression, while weighted quantile sum regression assessed mixture effects. Mediation analysis quantified inflammatory markers' roles in these relationships.ResultsPM2.5 was significantly association with cerebral hypoperfusion (OR = 4.344, 95% CI: 3.323–5.734), with black carbon (BC) showing the strongest association (OR = 4.396, 95% CI: 3.327–5.869), followed by sulfate (SO4 2−) (OR = 4.359), organic matter (OM) (OR = 4.278), nitrate (NO3 −) (OR = 3.552), and ammonium (NH4 +) (OR = 3.313). OM (44.2%) and BC (42.2%) contributed 86% of the total mixture effect. Neutrophils mediated 5.37% of the total effect between PM2.5 and cerebral hypoperfusion, whereas leukocytes (4.30%) and monocytes (2.73%) demonstrated significant mediation effects.ConclusionProlonged exposure to PM2.5 and its components, particularly OM and BC, was associated with increased cerebral hypoperfusion risk, partially mediated by inflammatory responses. These findings highlight the need for component‐targeted air quality policies to reduce cerebrovascular risk in vulnerable populations. Introduction Intracranial arterial stenosis (ICAS) is an increasingly recognized cerebrovascular condition among older adults, with a particularly high prevalence in East Asian populations [1]. Acute cerebral hypoperfusion, a devastating consequence of ICAS, often precedes infarction and may initially present with atypical symptoms such as headache, representing a potentially reversible phase of ischemic injury [2, 3, 4]. Hemodynamic disturbances during this phase can lead to white matter rarefaction and are also associated with lacunar infarctions, likely reflecting shared small vessel disease mechanisms, which are closely associated with cognitive decline, vascular dementia, and an elevated risk of recurrent ischemic stroke, particularly in individuals with underlying intracranial vascular disease [5, 6, 7]. In China, rapid urbanization, together with factors such as industrial emissions and evolving emission control policies, has intensified public concerns about air pollution, particularly the cerebrovascular effects of fine particulate matter (PM2.5) [8, 9]. Increasing evidence suggests that cerebral hemodynamic instability may be a key mechanism through which PM2.5 exerts its harmful effects [10, 11]. Supporting this hypothesis, the results of Pan et al. showed that every 10 μg/m3 increase in PM2.5 concentration significantly reduces blood flow velocity in the middle cerebral artery [12]. Notably, individuals with pre‐existing intracranial vascular disease may be especially vulnerable to such exposures, further heightening public health concerns. In recent years, the pathophysiological mechanisms underlying PM2.5‐induced cerebrovascular dysfunction have become a focus of scientific inquiry [13, 14]. Research suggests that inhaled particles may affect the brain primarily through systemic circulation following pulmonary absorption, with subsequent disruption of the blood–brain barrier. In addition, experimental studies have proposed a potential direct translocation route via the olfactory or trigeminal pathways; however, this mechanism remains under investigation and its relevance in humans is not fully established [13, 15, 16]. PM2.5 exposure may also activate inflammatory monocytes and promote systemic vascular dysfunction via TLR4 and NADPH oxidase signaling pathways [17], triggering inflammation and endothelial damage, thereby increasing the risk of stroke and cognitive decline [14]. Although the precise mechanisms require further elucidation, systemic inflammatory cascades are likely key mediators linking environmental exposure to cerebrovascular impairment. PM2.5 is a complex mixture of airborne particles with aerodynamic diameters ≤ 2.5 μm, comprising diverse chemical constituents such as sulfate (SO₄2−), nitrate (NO₃−), ammonium (NH₄+), organic matter (OM), and black carbon (BC). Due to heterogeneous emission sources, these components exhibit substantial spatial and temporal variability and distinct toxicological profiles based on their physicochemical properties [18, 19]. Most existing studies have focused on the aggregate health effects of PM2.5, overlooking the differential impacts of individual components on cerebral circulation and the complex mechanisms through which they may act synergistically. Despite being a frontier topic, the association between PM2.5 constituents and the risk of cerebral hypoperfusion remains poorly understood, and mechanistic studies are scarce. This knowledge gap limits the ability of policymakers to implement targeted regulations to control emissions from specific sources. In this context, the present study investigates both individual and combined exposure to PM2.5 and its major components (SO4 2−, NO3 −, NH4 +, OM, BC) in relation to the risk of cerebral hypoperfusion. We further quantify the relative contributions of these components and examine the mediating role of immune‐inflammatory markers. The findings aim to inform evidence‐based policymaking and guide public health interventions by clarifying the roles of specific PM2.5 components and inflammatory pathways in cerebrovascular risk. Methods Study Design and Population We conducted a retrospective analysis of consecutive patients diagnosed with symptomatic ICAS at the First Affiliated Hospital of Soochow University between January 2018 and December 2022. Eligible patients were adults with imaging‐confirmed intracranial arterial stenosis (≥ 50% in at least one major intracranial artery), assessed using a combination of CTA, MRA, or transcranial Doppler (TCD), who experienced neurological symptoms corresponding to the stenotic artery (e.g., ischemic stroke, transient ischemic attack, dizziness, or limb numbness). To ensure rigorous patient selection, all available imaging data were systematically reviewed and incorporated into the screening process. This study was approved by the Institutional Review Board of the First Affiliated Hospital of Soochow University ((2025) Ethics Approval 235), and all participants provided informed consent for relevant examinations and questionnaire surveys. The methods used in this research follow the principles outlined in the Declaration of Helsinki. Clinical trial number: not applicable. A total of 1680 patients were initially screened for inclusion. The following exclusion criteria were applied: (1) patients under 18 years of age; (2) patients with incomplete clinical records or insufficient imaging data for analysis; (3) patients with intracranial arterial stenosis of less than 50% who remained asymptomatic during the screening period; (4) patients without a stable residence in Suzhou or those who had relocated within the past five years; (5) patients diagnosed with malignancies or significant hepatic or renal impairment; and (6) patients with diagnosed autoimmune or rheumatologic disorders, or those with a documented infection within the past two weeks. The selection of participants is shown in Figure S1. All patients received standardized medical management based on current clinical guidelines, including dietary modification, blood pressure control, and intensive lipid‐lowering therapy. Data Collection Relevant data were obtained from electronic medical records and administrative databases, including patient demographics (age and gender), smoking status, and pre‐existing conditions such as hypertension, diabetes, and coronary heart disease. Blood samples were collected on the first day after admission in a fasted state for each patient. Routine blood tests, including complete blood counts and common biochemical markers such as blood glucose, liver and kidney function, and electrolytes, were also performed and documented. Cerebral Hypoperfusion Assessment All patients underwent non‐contrast CT (NCCT), CT angiography (CTA), and CT perfusion (CTP) using 128‐slice dual‐source scanners. For CTA, 50–80 mL of non‐ionic contrast medium (Omnipaque, 300 mg/mL) was administered via an 18‐G needle into the antecubital vein at 5 mL/s. Time–density curves were generated and transferred to the Syngo.VIA workstation for initial processing using the VPCT Neuro‐STROKE platform. Images were further analyzed with MIStar software. Perfusion parameters—including cerebral blood flow (CBF), cerebral blood volume (CBV), mean transit time (MTT), and time to maximum (Tmax)—were calculated using a deconvolution algorithm. Low regional CBF (rCBF) was defined as < 35 mL/100 g/min (i.e., 2 standard deviations below the normal range), indicating oligemia. Perfusion abnormalities were assessed on pseudo‐color maps, and the largest dimension of each lesion was selected for quantification. MIStar was used to measure the volume (mL) and proportional distribution of perfusion deficits across the whole brain and specific lobes. To ensure reliability, two independent radiologists evaluated the perfusion maps and manually assessed the location and extent of cerebral hypoperfusion and ischemic core regions. Inter‐rater agreement and consistency with software‐generated results were assessed using the kappa (κ) statistic. Final perfusion maps were displayed on a grayscale CT background, with pseudo‐color overlays indicating perfusion abnormalities: red for ischemic core, green for ischemic penumbra (i.e., regions with delayed MTT but preserved or mildly reduced CBV), while normal tissue appeared in standard grayscale. Based on integrated radiological and quantitative imaging assessments, patients were categorized into hypoperfusion and non‐hypoperfusion groups. It should be noted that, although all patients had symptomatic ICAS, some exhibited normal perfusion on CT maps because acute symptoms may correspond to small tissue volumes that do not reach the rCBF threshold, or perfusion deficits were too subtle to be detected by CTP. The hypoperfusion group was defined as patients with any red or green regions on perfusion maps, reflecting either irreversible ischemia or potentially salvageable tissue with impaired perfusion. This classification served as the basis for subsequent analyses of associations between cerebral perfusion status and PM2.5 component exposure. Air Pollution Exposure Assessment The research team linked participants' residential addresses to their corresponding levels of residential ambient air pollution exposure. Pollution data were obtained from the TAP database (http://tapdata.org.cn/?page_id=88), which has been independently validated and widely utilized in studies assessing the adverse health effects of air pollution [20, 21]. To accurately characterize the spatiotemporal variation of pollution closely associated with emission sources, the database integrates machine learning algorithms, land use spatial data, and satellite‐derived aerosol optical depth (AOD) parameters. These inputs include population density, road networks, vegetation cover, land surface types, and elevation, and the framework is based on a well‐established air pollution assessment model [22]. Previous studies have identified organic matter (OM), black carbon (BC), ammonium (NH4 +), nitrate (NO3 −), and sulfate (SO4 2−) as the major chemical components of PM2.5 pollution in China [23]. The estimation of PM2.5 chemical composition was constrained by total PM2.5 mass concentrations and derived using operational CMAQ (Community Multiscale Air Quality) simulations. To correct the inherent biases in CMAQ model outputs, the following two‐step approach was applied: first, the dust emission simulation module was refined; second, an XGBoost (Extreme Gradient Boosting) algorithm was used, in combination with ground‐based observations of PM2.5 components, to build a calibration model that adjusted the relative contributions of each PM2.5 constituent. This method enabled the generation of more accurate interconversion factors among PM2.5 components, ultimately producing refined results. The spatial resolution of PM2.5 concentration data was set at 1 km × 1 km, whereas the spatial resolution of individual PM2.5 chemical component data was set at 10 km × 10 km. Based on the previous studies [24], this study used the three‐year average concentration as the metric for assessing long‐term PM2.5 exposure levels. Statistical Analysis Participant characteristics were summarized using median (P25, P75) for continuous variables and count (percentage) for categorical variables. Differences between the cerebral hypoperfusion and non‐cerebral hypoperfusion groups were assessed using the Mann–Whitney U test for continuous variables and χ2 tests for categorical variables. The association between PM2.5 exposure and cerebral hypoperfusion was evaluated using generalized additive models. Additionally, a restricted cubic spline regression model was applied to examine potential nonlinear relationships between PM2.5, its chemical components, and the risk of cerebral hypoperfusion. To assess the impact of mixed PM2.5 component exposure, we used weighted quantile sum (WQS) regression. In this approach, PM2.5 component concentrations were first converted into quantiles, and the WQS model then assigned weighted scores to each component, generating an overall weighted index (WQS index) that reflected the cumulative exposure burden. Component weights indicated their relative contribution to the index. The model was run with 2000 bootstrap resampling iterations, and predefined directional assumptions (positive/negative) were used to explore both potential protective and adverse effects of PM2.5 exposure. Subgroup analyses were conducted based on age, sex, smoking status, and pre‐existing conditions, including hypertension and diabetes (Since the number of patients with coronary heart disease in this study population is 36, representing 3.7% of the total, we have opted not to conduct subgroup analyses pertaining to CHD history in this study.) to test the robustness of the PM2.5‐cerebral hypoperfusion association. Additionally, we used the ‘mediation’ package in R to investigate whether peripheral immune‐inflammatory markers, specifically absolute counts of white blood cells, neutrophils, monocytes, and lymphocytes from complete blood counts, mediated this relationship, with bootstrap resampling applied for validation. Finally, we conducted sensitivity analyses to determine the reliability and stability of our primary outcomes: (1) using 1‐year average exposure concentrations of PM2.5, and (2) using 5‐year average exposure concentrations of PM2.5. Potential confounders were included if previously reported in the literature or identified as relevant to the primary outcome. All outcome models were adjusted for key covariates, including age, sex, smoking status, and pre‐existing conditions. A two‐sided p‐value < 0.05 was considered statistically significant. All analyses were performed in R (version 4.3.3) using the “mgcv”, “rms”, and “visreg” packages. Results Demographic Characteristics A total of 967 individuals participated in this study (Table S1). Of these participants, 666 (68.9%) were diagnosed with cerebral hypoperfusion, while 301 (31.1%) had no cerebral hypoperfusion. The sample comprised 605 (62.6%) males and a median age of 69.0 years [61.0–76.0]. Participants with cerebral hypoperfusion were more likely to be male (67.0% vs. 52.8%, p < 0.001), older (median age 70.0 vs. 68.0 years, p = 0.002), and have a higher prevalence of hypertension (72.1% vs. 63.5%, p = 0.009) compared to those without cerebral hypoperfusion. Additionally, participants with cerebral hypoperfusion had significantly higher exposure to air pollutants, including PM2.5 (43.57 vs. 35.67 μg/m3, p < 0.001), SO4 2− (8.00 vs. 6.65 μg/m3, p < 0.001), NO3 − (9.44 vs. 8.08 μg/m3, p < 0.001), NH4 + (5.88 vs. 5.03 μg/m3, p < 0.001), OM (9.15 vs. 7.55 μg/m3, p < 0.001), and BC (1.91 vs. 1.52 μg/m3, p < 0.001). Association Between PM2.5 , Its Components and Cerebral Hypoperfusion Table 1 presents the associations between PM2.5, its constituents, and cerebral hypoperfusion using both univariate and multivariable logistic regression analyses. All pollutants showed significant positive associations with impaired cerebral hypoperfusion. After adjusting for age, gender, hypertension, diabetes, coronary heart disease, and smoking in the multivariable analysis, the associations remained significant. The adjusted OR for PM2.5 was 4.344 (95% CI: 3.323–5.734, p < 0.001). Among the constituents, BC showed the strongest association (OR = 4.396, 95% CI: 3.327–5.869, p < 0.001), followed by SO4 2− (OR = 4.359, 95% CI: 3.309–5.803, p < 0.001), OM (OR = 4.278, 95% CI: 3.278–5.635, p < 0.001), NO3 − (OR = 3.552, 95% CI: 2.838–4.477, p < 0.001), and NH4 + (OR = 3.313, 95% CI: 2.612–4.239, p < 0.001). Sensitivity analyses revealed a robust association between the 1‐ and 5‐year mean exposure to PM2.5, its constituents, and the risk of cerebral hypoperfusion (Table S2, Table S3). TABLE 1Associations between PM2.5 and its constituents and cerebral hypoperfusion.ExposureUnivariate logistic regressionMultivariable logistic regressionOR (95% CI) p‐valueOR (95% CI) p‐valuePM2.5 4.154 (3.197–5.446)< 0.0014.344 (3.323–5.734)< 0.001SO4 2− 4.141 (3.165–5.473)< 0.0014.359 (3.309–5.803)< 0.001NO3 − 3.396 (2.730–4.252)< 0.0013.552 (2.838–4.477)< 0.001NH4 + 3.151 (2.500–4.004)< 0.0013.313 (2.612–4.239)< 0.001OM4.077 (3.145–5.331)< 0.0014.278 (3.278–5.635)< 0.001BC4.211 (3.206–5.584)< 0.0014.396 (3.327–5.869)< 0.001 Note: Multivariable analyses were adjusted for age, gender, hypertension, diabetes, coronary heart disease, and smoking. OR and their 95% CI of cerebral hypoperfusion associated with each IQR increase in PM2.5 exposures.Abbreviations: BC, black carbon; CI, confidence interval; IQR, interquartile range; NH₄+, ammonium; NO₃−, nitrate; OM, organic matter; OR, odds ratio; PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; SO₄²−, sulfate. Associations between PM2.5 and its constituents and cerebral hypoperfusion. Note: Multivariable analyses were adjusted for age, gender, hypertension, diabetes, coronary heart disease, and smoking. OR and their 95% CI of cerebral hypoperfusion associated with each IQR increase in PM2.5 exposures. Abbreviations: BC, black carbon; CI, confidence interval; IQR, interquartile range; NH₄+, ammonium; NO₃−, nitrate; OM, organic matter; OR, odds ratio; PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; SO₄²−, sulfate. The probability density distributions (Figure 1A,B) show that BC had the lowest concentration range with the highest density peak, while other constituents exhibited broader distributions, and total PM2.5 displayed a bimodal pattern. The weighted quantile sum (WQS) analysis (Figure 1C) demonstrated that the overall effect of PM2.5 on cerebral hypoperfusion was primarily attributed to two constituents. OM contributed the largest proportion (0.442), closely followed by BC (0.422), together accounting for approximately 86% of the total effect. SO4 2− had a moderate contribution (0.089), while NH4 + and NO3 − showed minimal influence with weights of 0.024 and 0.023, respectively, suggesting that organic matter and black carbon may be the most critical components of PM2.5 affecting cerebral hypoperfusion. Distributions and relative importance of PM2.5 and its components in cerebral hypoperfusion. Abbreviations: WQS, weighted quantile sum; PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; NH₄+, ammonium; BC, black carbon; OM, organic matter; SO₄2−, sulfate; NO₃−, nitrate. (A) Probability density distribution of BC, NH4 +, NO3 −, OM, and SO4 2− concentrations. (B) Probability density distribution of total PM2.5 mass concentration. (C) Association of the constituents of PM2.5 with Cerebral hypoperfusion in WQS analyses. The models were adjusted for age, gender, hypertension, diabetes, coronary heart disease, and smoking. According to the concentration‐response curve (Figure 2), we found that associations of exposure to PM2.5 mass, SO4 2−, OM, and BC and risk of cerebral hypoperfusion were non‐linear over the entire exposure range (p‐values for non‐linear relationships < 0.001 for PM2.5, SO4 2−, OM, and BC; p = 0.005 for NH4 +), showing steeper increases in risk at higher concentration levels. For NO3 −, the relationship with cerebral hypoperfusion risk was more gradual and appeared roughly linear at lower concentrations, with increasing steepness at higher exposure levels (p‐value for non‐linear relationship = 0.458). Concentration‐response associations between long‐term PM2.5 and its five constituents exposures and cerebral hypoperfusion using restricted cubic spline. Abbreviations: PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; NH₄+, ammonium; BC, black carbon; OM, organic matter; SO₄2−, sulfate; NO₃−, nitrate. (A) PM2.5, (B) SO4 2−, (C) NO3 −, (D) NH4 +, (E) OM, and (F) BC. Solid lines represent odds ratios, dashed lines represent 95% confidence intervals, and histograms show the distribution of pollutant concentrations. Models were adjusted for age, sex, hypertension, diabetes, coronary heart disease, and smoking status. Mediation Effects of Immune Inflammation Factors in the Relationships of PM2.5 and Its Components With Cerebral Hypoperfusion Mediation analysis was conducted to investigate whether immune inflammation factors mediated the associations between PM2.5, its components, and cerebral hypoperfusion. For PM2.5, neutrophils mediated 5.37% of the total effect (p < 0.001), leukocytes mediated 4.30% (p < 0.001), and monocytes mediated 2.73% (p < 0.001). For OM, neutrophils mediated 5.50% of the total effect (p < 0.001), leukocytes mediated 4.26% (p < 0.001), and monocytes mediated 2.71% (p < 0.001). For BC, neutrophils mediated 5.19% of the total effect (p < 0.001), leukocytes mediated 4.66% (p < 0.001), and monocytes mediated 2.96% (p < 0.001) (Figure 3). The other components (SO4 2−, NO3 −, NH4 +) showed similar patterns, with neutrophils consistently demonstrating the strongest mediation effect across all pollutants (Figure S2). Mediation of association between PM2.5, OM, BC and cerebral hypoperfusion by immune inflammation factors, including Leukocytes, Monocytes, Lymphocytes, and Neutrophils. Abbreviations: PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; BC, black carbon; OM, organic matter. Adjusted for age, gender, hypertension, diabetes, coronary heart disease, and smoking. p < 0.05 was considered statistically significant. *: p‐values adjusted by using the FDR multiple correction method. Subgroup Analysis Subgroup analysis stratified by age, gender, hypertension, diabetes, and smoking status in multi‐pollutant models consistently showed a significant link between PM2.5 component mixture and cerebral hypoperfusion (p < 0.05, Figure 4). For PM2.5 and all five chemical components (including NH4 +, BC, OM, SO4 2−, and NO3 −), we observed numerically stronger associations among males (OR: 4.954, 95% CI: 3.462–7.213 for PM2.5) compared with females (OR: 3.719, 95% CI: 2.471–5.727), and among individuals with hypertension (OR: 5.112, 95% CI: 3.657–7.258) compared to those without (OR: 3.295, 95% CI: 2.100–5.303). However, none of these subgroup differences reached statistical significance (all P for interaction > 0.05). Figure 5 presents the relative weights of PM2.5 components in different subgroups, with BC showing the highest weight in individuals aged < 60 years (0.67) and OM showing substantial weights in diabetic individuals (0.402). More details can be found in Table S4. Subgroup analyses of ORs (95% CI) for cerebral hypoperfusion with per IQR increase in exposure to PM2.5 and its constituents by age groups, gender, hypertension, diabetes, and smoking. Abbreviations: PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; NH₄+, ammonium; BC, black carbon; OM, organic matter; SO₄2−, sulfate; NO₃−, nitrate; OR, odds ratio; CI, confidence interval; IQR, interquartile range. Subgroup analyses of relative weights of PM2.₅ constituents for cerebral hypoperfusion associated with per quartile increase in co‐exposures of PM2.₅ constituents by age groups, gender, hypertension status, diabetes status, and smoking status in the WQS models. Abbreviations: PM2.₅, particulate matter with aerodynamic diameter ≤ 2.5 μm; NH₄+, ammonium; BC, black carbon; OM, organic matter; SO₄2−, sulfate; NO₃−, nitrate. Discussion In this study involving symptomatic ICAS patients, we observed that long‐term exposure to PM2.5 mass and its constituents (NH4 +, OM, BC, SO4 2−, and NO3 −) was associated with an increased risk of cerebral hypoperfusion. Weighted Quantile Sum (WQS) regression indicated that BC and OM contributed the most to this risk, with respective weights of 0.422 and 0.442. Additionally, leukocytes, monocytes, and neutrophils partially mediated the associations between PM2.5, BC, OM, and cerebral hypoperfusion. Stratified analyses by sex, age group, hypertension, diabetes, and smoking status further supported the elevated risk of cerebral hypoperfusion with higher exposure to PM2.5 and its components. To our knowledge, this is the first study to examine the associations between PM2.5 constituents and cerebral hypoperfusion specifically in symptomatic ICAS patients (mechanism illustrated in Figure 6). Potential underlying mechanism of the links between PM2.5 components and cerebral hypoperfusion. Long‐term exposure to organic matter (OM) and black carbon (BC), two key components of PM2.5, showed the strongest associations with the risk of cerebral hypoperfusion. This finding is consistent with previous evidence linking these constituents to cerebrovascular outcomes. For instance, Chen et al. reported that OM and NH4 + were the dominant contributors to cerebrovascular hospitalizations, while Cai et al. found that environmental exposure to BC, OM, and SO4 2− increased stroke‐related mortality in hospitalized patients [25, 26, 27]. Similarly, Xu et al. (2023) highlighted OM as the main factor driving reductions in cerebral blood flow velocity in a multi‐pollutant model [11]. There is no consensus regarding the most significant PM2.5 component. At the same time, the relationship between PM composition and cerebral hypoperfusion in existing studies remains to be fully elucidated due to the limited availability of imaging data. Unlike earlier studies that focused primarily on clinical endpoints, the present research incorporated imaging‐based assessment of cerebral perfusion, providing more direct evidence of PM2.5‐related cerebrovascular effects. BC primarily originates from vehicle emissions and biomass combustion, whereas OM arises from both combustion sources and secondary atmospheric reactions [28, 29]. Their adverse impact may be mediated by vascular inflammation and endothelial dysfunction [10, 15], which can compromise cerebral perfusion. These findings not only support the results of previous studies but also expand current understanding by identifying specific PM2.5 components contributing most to cerebral hypoperfusion. Moreover, our mediation analysis confirmed that immune indicators—specifically leukocytes, monocytes, and neutrophils—partially mediate the association between PM2.5 exposure and cerebral hypoperfusion. This finding is consistent with previous studies that have proposed systemic inflammatory responses as a potential mechanism linking PM2.5 exposure to cerebrovascular dysfunction. PM2.5 may penetrate the brain via the blood–brain barrier or olfactory nerve, triggering neuroinflammation and cognitive impairment in animal models [30, 31, 32]. The observed brain damage likely involves systemic oxidative stress and inflammation that disrupt cerebral vascular integrity [30, 33]. Although the proportion of mediation was relatively modest, these effects may still be biologically meaningful, especially in the context of multiple coexisting pathological pathways. Our findings thus support and extend existing evidence on inflammation‐mediated mechanisms in PM2.5‐related brain injury. Although this study offers new evidence on the associations between PM2.5 components (especially BC and OM) and cerebral hypoperfusion, several limitations remain. First, the cross‐sectional design precludes causal inference, and longitudinal studies are needed to confirm these findings. Second, exposure estimates were based on residential addresses and did not account for personal time‐activity patterns, microenvironments, or occupational exposures, which may lead to exposure misclassification. Future research could consider personal monitoring or spatiotemporal modeling to enhance accuracy. Third, some PM2.5 components, such as OM and BC, share similar sources and may be correlated, potentially affecting model estimates. Lastly, the study lacked a comparison group without ICAS and did not fully account for potential confounders, including detailed lipid profiles, medication use, long‐term temporal trends, co‐pollutants, and additional inflammatory markers (e.g., C‐reactive protein), which may introduce residual confounding and affect the robustness of the observed associations. Future studies with more comprehensive covariate adjustment are warranted to strengthen causal interpretation. Conclusion Our findings provide new insights into the link between PM2.5 and its components with cerebral hypoperfusion in symptomatic ICAS patients. We identified a significant positive association between PM2.5 component mixtures and hypoperfusion risk, with OM and BC as primary contributors. Elevated immunoinflammatory markers (neutrophils, leukocytes, and monocytes) mediated these associations. Our results highlight the crucial role of inflammatory biomarkers in PM2.5‐induced cerebral hypoperfusion, suggesting that improving air quality and regulating immune responses could help reduce the burden of cerebrovascular disease in China. Author Contributions Jianhui Guo: conceptualization, methodology, investigation, validation, formal analysis, writing – review and editing. Xinyue Lu: conceptualization, methodology, software, data curation, investigation, validation, formal analysis, writing – original draft, writing – review and editing, visualization. Yi Yang: conceptualization, methodology, investigation, validation, formal analysis, writing – review and editing, funding acquisition, visualization, writing – original draft, supervision, data curation, software. Tongjin Xu: conceptualization, methodology, investigation, validation, formal analysis, writing – review and editing. Nan Wang: conceptualization, methodology, software, data curation, formal analysis, validation, investigation, writing – original draft, writing – review and editing, visualization. Le Yang: conceptualization, methodology, software, supervision, data curation, formal analysis, validation, investigation, funding acquisition, visualization, writing – review and editing. Liuyan Zheng: conceptualization, methodology, validation, investigation, writing – review and editing, formal analysis. Hao Li: conceptualization, methodology, investigation, validation, formal analysis, writing – review and editing, software, data curation, supervision, visualization, funding acquisition, writing – original draft. Xingyan Xu: conceptualization, methodology, software, data curation, formal analysis, investigation, validation, writing – original draft, writing – review and editing, visualization. Yiming Guo: conceptualization, methodology, investigation, validation, formal analysis, writing – review and editing. Funding This work was supported by grants from Natural Science Foundation of Fujian Province (2025J08039), Fujian Provincial Health Commission Youth Scientific Research Project (2025QNA005, 2025QNA080), Fujian Provincial Science and Technology Innovation Joint Fund Project (2025Y9051, 2025Y9554), National Natural Science Foundation of China (82201446), Natural Science Foundation of Jiangsu Province of China (BK20220249), Suzhou Science, Education, and Strengthened Healthcare Project (QNXM2024027), the Suzhou Basic Research Pilot Project (SSD2024043), and the Young Key Personnel Training Program of Fujian University of Traditional Chinese Medicine (X2025039). Disclosure Declaration of generative AI and AI‐assisted technologies in the writing process: During the preparation of this work, the authors used ChatGPT in order to improve language and readability. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Ethics Statement This study was approved by the Institutional Review Board of the First Affiliated Hospital of Soochow University (2025) Ethics Approval No. 235, and all participants provided informed consent for relevant examinations and questionnaire surveys. The methods used in this research follow the principles outlined in the Declaration of Helsinki. The experiments comply with the current laws of the country in which they were performed. Clinical trial number: not applicable. Conflicts of Interest On behalf of all authors, the corresponding author states that there are no conflicts of interest. Supporting information Table S1: Description characteristics of participants in this study *. Table S2: Associations between one‐year exposure to PM2.5 and its constituents and cerebral hypoperfusion. Table S3: Associations between five‐year exposure to PM2.5 and its constituents and cerebral hypoperfusion. Table S4: Association between PM2.5, its components and cerebral hypoperfusion by subgroups. Figure S1: Study Design and Participant Flow Chart. Figure S2: Mediation of association between NH4+, SO42−, NO3− and cerebral hypoperfusion by immune inflammation factors, including Leukocytes, Monocytes, Lymphocytes, and Neutrophils.