Perivascular spaces (PVS) are fluid-filled tubular channels distributed along the course of cerebral blood vessels and serve as a critical pathway for the clearance of metabolic waste in the central nervous system (CNS) (1, 2). The recently proposed “glymphatic-like system” theory suggests that PVS plays a vital role in maintaining the homeostasis of the CNS microenvironment by facilitating the transport and exchange of cerebrospinal fluid and interstitial fluid (3). PVS has been associated with various neurological disorders, particularly cerebrovascular diseases. For example, PVS may reflect the early stage of cerebral small vessel disease. Many studies have demonstrated a substantial association between PVS and an increased risk of hemorrhagic and ischemic stroke, dementia, cerebral amyloid angiopathy (CAA), etc. (4–6). Furthermore, abnormal PVS distributions have been observed in patients with neurodegenerative diseases such as Alzheimer’s disease and Parkinson’s disease (7, 8). Additionally, there is a potential association between PVS and neuroinflammatory demyelinating diseases, including multiple sclerosis and neuromyelitis optica spectrum disorders (9, 10). Risk factors for PVS include traditional vascular risk factors such as age, hypertension, diabetes, smoking, and dyslipidemia. Moreover, chronic neuroinflammation and genetic susceptibility also contribute to the development of PVS (5, 11–14). However, the mechanisms underlying PVS and its causal relationships with neurological disorders remain poorly understood and necessitate further investigation.

Uric acid (UA) is the final product of human purine metabolism, primarily produced in the liver. It may exhibit complex dual biological effects, functioning as both an antioxidant and a pro-oxidant (15). Notably, recent studies have also demonstrated that UA can induce neuroinflammation (16) and promote brain atrophy (17), suggesting its involvement in PVS. High-density lipoprotein cholesterol (HDL-C) exerts a protective role in demyelinating nerve diseases by reducing blood–brain barrier (BBB) permeability or mitigating microglial activation (18, 19). However, excessively high HDL-C levels (>2.2 mmol/L) can promote oxidative stress and inflammatory responses (20), which are the mechanisms underlying PVS. The uric acid to high-density lipoprotein cholesterol ratio (UHR) overcomes the limitations of a single indicator and facilitates a more comprehensive and multi-faceted assessment by combining the effects of UA and HDL-C (21). It can more accurately predict the risk of diseases such as coronary artery stenosis and abdominal aortic aneurysm (22, 23). Studies have demonstrated that UHR is associated with cardiovascular diseases such as arteriosclerosis (24), coronary artery disease (25), hypertension (26), myocardial infarction (27), and atrial fibrillation (28). An increase in UHR is related to an increased risk of adverse events such as stroke in patients with diabetes (29) and coronary heart disease (30). Moreover, the all-cause mortality and cardiovascular mortality of individuals may increase with the increase of UHR, indicating a non-linear association (31, 32). Additionally, UHR is also correlated with vascular risk factors such as diabetes (33), insulin resistance (34), non-alcoholic fatty liver disease (35), chronic kidney disease (36), and metabolic syndrome (37). The extensive relationship between UHR and these diseases highlights its value in the risk assessment of cardiovascular and cerebrovascular diseases. However, the association between UHR and PVS remains inadequately explored.

This study aims to investigate the association between UHR and the burden of PVS for the first time and further assess the potential dose–response relationship. Our findings may provide new evidence for the role of metabolic-inflammatory regulation in the formation of PVS.

This study systematically investigated the association between UHR and moderate-to-severe PVS burden for the first time. The findings indicate that UHR levels in the moderate-to-severe PVS group were significantly higher than those in the none-to-mild PVS group. UHR was independently and positively correlated with moderate-to-severe PVS burden, revealing a linear dose–response relationship, which suggests that UHR may serve as a potential risk factor for moderate-to-severe PVS burden. Furthermore, UHR was associated with a linear, dose-dependent increase in the risks of moderate-to-severe BG-PVS and moderate-to-severe CSO-PVS burdens. Subgroup analyses indicated that the relationship between UHR and PVS burden remained robust across most subgroups except for the subgroup with LDL-C levels < 3.4 mmol/L. Interaction analysis suggested that LDL-C had a moderating effect on the association. Including UHR in the basic model significantly enhanced predictive performance for PVS, BG-PVS, and CSO-PVS burdens. Finally, two sensitivity analyses were conducted to further validate the robustness of the core conclusions. The results indicate a positive association between UHR and PVS burden, suggesting that UHR may serve as a potential indicator for predicting the severity of PVS.

Although there is no study on the association between UHR and PVS, the relationships between UA, HDL-C, and PVS have been widely reported. For example, a study involving 665 patients with lacunar stroke found that UA was positively correlated with both the BG-PVS burden and the CSO-PVS burden, suggesting that UA is an independent risk factor for PVS (45). Li et al. (46) indicated that UA was associated with an increased risk of PVS, identifying it as a key predictor for its occurrence. A population-based multimodal imaging study demonstrated a significant linear association between elevated UA levels and an increased risk of midbrain enlarged PVS, suggesting that UA may play a crucial role in the pathogenic mechanisms of PVS in specific brain regions (47). In contrast, the existing evidence regarding the association between HDL-C and PVS is contradictory. One study found that CSO-PVS may be positively correlated with UA and negatively correlated with HDL-C (48). A multicenter study revealed that decreased levels of apolipoprotein E in plasma and the HDL fraction were significantly correlated with the degree of CSO-PVS, indicating that low apolipoprotein E content in HDL may increase the risk of PVS. However, Evans et al. (49) recently found that increased HDL-C levels were associated with a more severe PVS burden in the hippocampus, CSO, and midbrain regions but not in the BG region. These contradictory conclusions may arise from differences in study design, population characteristics, and control of confounding variables. This study introduces the novel inflammatory metabolic indicator UHR for the first time, finding that higher UHR levels were independently associated with an increased risk of moderate-to-severe PVS. This finding supports the complex involvement of metabolic-inflammatory imbalance in PVS. Furthermore, UHR has the potential to serve as a biomarker for assessing the risk of moderate-to-severe PVS, providing substantial support for its diagnosis. Additionally, given the lack of standardized clinical thresholds for UHR in our specific patient population, we objectively and systematically stratified patients used quartile grouping. This approach allows us to avoid the potential pitfalls of arbitrary threshold setting while ensuring an even distribution of participants across UHR groups, which enhances the statistical power of our findings. This is also consistent with previous studies in cardiovascular and cerebrovascular domains, which have effectively utilized quartile or quintile groupings to delineate UHR levels among patient cohorts (21, 32, 50–52).

Currently, UHR has been established as a reliable inflammatory indicator, and its role in metabolic inflammatory diseases has attracted increasing attention. UHR reflects the balance between the antioxidant effect of UA and the anti-inflammatory effect of HDL-C (53). An increase in this ratio indicates elevated oxidative stress and inflammatory response in the body, which aligns closely with the pathological mechanism of PVS, where the imbalance between reactive oxygen species (ROS) generation and the antioxidant defense mechanism can lead to cellular damage and vascular endothelial dysfunction. Although direct evidence of a causal association between UHR and PVS is lacking, UA and HDL-C, as core components of UHR, provide important insights into this potential relationship. UA mitigates the generation of free radicals by scavenging ROS and chelating metal ions, thereby protecting cells from oxidative damage and exerting a neuroprotective effect (54). However, excessive UA accumulation can cause its antioxidant properties to shift into a pro-oxidant effect, thereby contributing to PVS: (1) It promotes inflammation through enhancing ROS generation, activating NLRP3 inflammasome, and stimulating pro-inflammatory signaling (55–57); (2) It triggers the peroxynitrite anion-mediated lipid oxidation while reducing the production and bioavailability of nitric oxide in endothelial cells, exacerbating vascular endothelial dysfunction (58). (3) It increases BBB permeability (59), aggravating the formation of PVS. These mechanisms induce vascular endothelial injury, matrix remodeling, and impaired fluid drainage, which collectively promote the development of PVS (60). Conversely, HDL-C possesses anti-inflammatory and antioxidant properties that can counteract the damaging effects of elevated UA. However, in cases where the UHR is elevated, the protective role of HDL-C may be compromised, particularly in the presence of chronic inflammation often linked with high LDL-C levels (61). This interplay between elevated UA and dysfunctional HDL-C may exacerbate endothelial dysfunction and promote the expansion of PVS. Further investigation into these complex interactions is warranted, as elucidating the underlying biological mechanisms could provide valuable insights into therapeutic strategies aimed at managing PVS and related cerebrovascular conditions.

This study also evaluated the association between UHR and PVS burden across different brain regions. The results revealed a linear relationship between UHR and both moderate-to-severe BG-PVS burden and moderate-to-severe CSO-PVS burden. In the categorical model for UHR, the Q2, Q3, and Q4 groups exhibited an increased risk of moderate-to-severe BG-PVS compared to the Q1 group, whereas only the Q4 group demonstrated an increased risk of moderate-to-severe CSO-PVS. One possible explanation is the anatomical differences in PVS structures across various brain regions. The lenticulostriate and thalamostriate arteries in the BG region are surrounded by a double-layered pia mater, enabling the subpial space to communicate directly with the subarachnoid space. As a result, PVS in the BG region may be more susceptible to systemic metabolic indicators. Conversely, the penetrating superficial and cortical arteries leading to the CSO region are covered by a single layer of pia mater, with the subpial space in this area connecting indirectly to the subarachnoid space. This relatively closed drainage structure may limit the influence of peripheral metabolic factors on CSO-PVS; however, it may also render this region more vulnerable to reduced clearance efficiency. Furthermore, the PVS burden in different regions may involve distinct pathophysiological processes. For example, BG-PVS is associated with hypertensive arterial lesions, while CSO-PVS is more closely linked to CAA (62). Endothelial dysfunction related to hypertension and systemic inflammation disrupts the BBB, leading to a series of reactions, including plasma protein leakage into the PVS, impaired vasodilation, myelin formation and repair disorders, and obstruction of perivascular fluid drainage, which promote PVS formation in the BG region (63). Consistently, numerous studies have found a strong correlation between elevated UHR levels and hypertension (64, 65). It is hypothesized that UHR can exacerbate the oxidative stress cascade, mediating damage to hypertensive cerebral arteries, which results in decreased pulsatility in perforating arteries and diminished clearance efficiency of the glymphatic system. According to the clearance hypothesis, CSO-PVS indicates chronic poor drainage around pial arterial vessels, making individuals more susceptible to amyloid protein clearance defects and the development of CAA (66). Lipid metabolism may also play a role in the occurrence of CAA. The apolipoprotein E ε4 gene is a genetic risk factor for CAA, influencing the conformation and clearance of amyloid-β (67). It is posited that increased UHR levels may decrease apolipoprotein-dependent amyloid-β clearance efficiency, exacerbating CSO-PVS.

Furthermore, this study evaluated the association between UHR and the moderate-to-severe PVS burden in different subgroups. The results suggested that the association between UHR and the moderate-to-severe PVS burden remained significant in most subgroups, except for the subgroup with LDL-C < 3.4 mmol/L. Notably, our interaction analysis identified a significant interaction effect between LDL-C levels and UHR, suggesting that the relationship between UHR and PVS burden may vary depending on LDL-C levels. This finding underscores the importance of considering lipid profiles when evaluating the effects of UHR on vascular health. Mechanistically, elevated LDL-C is known to contribute to the development of atherosclerosis, leading to endothelial dysfunction and impaired vascular homeostasis. Such conditions may exacerbate the negative effects of increased UHR on the PVS burden. Specifically, high LDL-C levels can promote oxidative stress and inflammation within the vascular system, which could diminish the protective effects of HDL-C and increase susceptibility to vascular damage. Research suggests that the LDL-C to HDL-C ratio is a critical indicator of cardiovascular health, where a higher ratio correlates with an increased risk of atherosclerotic events and poorer vascular outcomes (68). Additionally, the antioxidant and anti-inflammatory properties of HDL-C can be compromised in the presence of elevated LDL-C levels, leading to impaired endothelial repair mechanisms (61). Therefore, when LDL-C levels are elevated, the protective effects of HDL-C on vascular structures may be diminished, amplifying the detrimental effects associated with high UHR. This dynamic interaction may elucidate the observed association between UHR and PVS burden, emphasizing the need for future studies that explore these interrelationships further to provide enhanced context for lipid management strategies in patients at risk for cerebrovascular diseases.

Previous studies have confirmed that UHR has a certain predictive effect on chronic inflammatory diseases such as coronary artery stenosis, prediabetes, and sacroiliitis (69–71), suggesting that UHR can serve as a convenient, accessible, and non-invasive clinical assessment tool. In our study, the incremental predictive value of UHR for the moderate-to-severe PVS burden was evaluated using the NRI and IDI, which indicated that the predictive performance for the moderate-to-severe PVS/BG-PVS/CSO-PVS was significantly improved after introducing UHR into the basic model. Finally, sensitivity analyses revealed that log2UHR was independently associated with the moderate-to-severe PVS burden. After excluding the patients who had received lipid-lowering drugs, the association between UHR and the moderate-to-severe PVS burden remained significant. This further verified the robustness of the association between UHR and PVS burden.

The clinical utility of UHR in evaluating cerebrovascular disease risk is promising, particularly as a simple and cost-effective biomarker that can supplement existing diagnostic practices. Given the association between UHR and the burden of PVS, incorporating UHR measurements could enhance risk stratification for patients presenting with symptoms indicative of cerebrovascular conditions or those with established risk factors such as hypertension, diabetes, and dyslipidemia. To facilitate its integration into clinical workflows, routine UHR evaluation could be considered as part of the standard metabolic panel during initial patient assessments. Specifically, clinicians could utilize UHR alongside traditional risk factors to evaluate patients’ vascular health more comprehensively. By establishing thresholds for UHR that correlate with increased PVS burden, clinicians can identify individuals who may require more frequent monitoring, referrals for imaging studies, or targeted therapeutic interventions.

The study, however, is not without limitations. Primarily, it focused on assessing the PVS burden in the BG and CSO regions, neglecting less common distribution areas like the cerebellum, corpus callosum, and hippocampus. Future research should comprehensively examine PVS distribution across the entire brain to thoroughly explore the relationship between UHR and the spatial specificity of PVS burden. Secondly, due to the limited sample size, there may be insufficient statistical power, potentially weakening the observed associations. Thirdly, this is a single-center study with all patients recruited from a single hospital in China. The clinical characteristics, ethnic backgrounds, and health system practices of the study population might not be representative of broader populations, thereby affecting the generalizability of our findings. Future research involving multicenter designs and diverse populations is essential to validate these findings and enhance their applicability to various clinical settings. Fourthly, while multiple confounders have been adjusted for in our analyses, it is important to recognize that other unaccounted or unmeasured confounders may still exist. These variables could potentially influence the observed association between UHR and PVS burden, warranting caution in the interpretation of our findings. Lastly, the study is a cross-sectional analysis, limiting our ability to establish a causal relationship between UHR and PVS burden. Additionally, the temporal sequence between UHR levels and PVS burden remains unclear. More extensive longitudinal studies are warranted to better understand how UHR influences the onset and advancement of PVS.

In conclusion, this study provides novel insights into the role of UHR in the burden of PVS in the brain. The findings reveal a positive association between UHR levels and the risk of moderate-to-severe PVS burden. Notably, the association was consistent across different brain regions, including the BG and CSO, and remained robust in subgroup analyses. The inclusion of UHR significantly improved the predictive performance for PVS burden, underscoring its potential as a non-invasive, accessible biomarker for assessing PVS severity. Moreover, the study highlights the complex interplay between metabolic-inflammatory factors and PVS formation, emphasizing the need for further research to elucidate the underlying mechanisms and clinical implications of UHR in PVS-related neurological disorders. Future studies should focus on larger, longitudinal cohorts to validate these findings, establish causality, and explore the therapeutic implications of targeting UHR in the management of PVS-associated conditions.