Cerebral small vessel disease (CSVD) represents a cluster of pathologies with a heterogeneous etiology, which can affect the brain vascular system, such as small arteries, capillaries, and small veins. CSVD can cause lacunar infarction and cerebral hemorrhage, which accounts for 20–30% of symptomatic strokes (1, 2). Typical CSVD lesions are characterized by white matter hyperintensities of presumed vascular origin, lacunes, microbleeds, enlarged perivascular space (EPVS), and microinfarcts (3). CSVD is recognized as an important cause of cognitive impairment, dementia, increased risk of stroke (4), and worse outcome after stroke (5). It is the most common cause of vascular dementia, contributing to about 50% of dementias worldwide, and, is a massive health burden of stroke and dementia (4). The possible mechanism of CSVD is arteriolosclerosis, lipohyalinosis, or fibrinoid necrosis of small vessels.

Enlarged perivascular space, a neuroimaging hallmark of CSVD, is associated with morphological features of CSVD (3). It is generally most prominent in the inferior basal ganglia (BG) and centrum semiovale (CSO). Epidemiological evidence indicates that the overall EPVS burden is associated with the vascular dementia, Alzheimer's disease, post-stroke depression, and post-stroke cognitive impairment (6, 7). The CSO perivascular space (PVS) develops in middle age and increases with age (8). PVS in the cerebral hemisphere white matter is more common in patients with cerebral amyloid angiopathy (9), while that in the BG is associated with deep perforating artery arteriolosclerosis and subcortical vascular cognitive impairment (10). There are some known risk factors for EPVS, such as old age, hypertension, estimated glomerular filtration rate (eGFR), and hyperuricemia (11). However, other risk factors remain to be further investigated.

Insulin resistance (IR) is a pathological condition resulting from the reduced insulin sensitivity in peripheral tissues (12). Studies have indicated that IR can enhance the activation, adhesion, and aggregation of platelets, and promote atherosclerosis, which may further lead to ischemic stroke (13, 14). Therefore, IR should be accurately measured in clinic and the damage caused by IR should be closely monitored (15, 16). The Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) has been proposed as a surrogate biomarker of IR. However, the HOMA-IR depends on insulin levels that are not usually measured in clinical practice (17).

Insulin resistance can alter the systemic lipid metabolism. Excessive lipolysis can increase the synthesis of hepatic triglyceride (TG) and decrease the levels of high-density lipoprotein (HDL) and the appearance of small dense low-density lipoprotein (LDL) (16, 18). Based on this, the triglyceride-glucose (TyG) index is suggested as a convenient marker for IR (17, 19). Previous studies have confirmed that the TyG index has high correlation with hyperinsulinaemic-euglycaemic clamps (the gold standard technique for assessing IR) and HOMA-IR (20). TyG index has been found to be associated with metabolic syndromes, cardiovascular diseases, cerebrovascular diseases, arterial stiffness, and carotid atherosclerosis (19, 21–24).

An increased TyG index is significantly correlated with both macrovascular diseases and microvascular diseases (17). Many studies that focused on the relationship between IR and CSVD have used HOMA-IR as indicator. For example, HOMA-IR is found to be associated with total CSVD score (25), silent brain infarcts (26), and white matter hyperintensity (WMH) (27), and EPVS (28). However, most of the studies were conducted in a relatively healthy population and did not exclude patients with diabetes. Recently, a cross-sectional study suggested that the TyG index had a slightly stronger association with the prevalence of silent brain infarct and the volume of WMH than HOMA-IR (27). However, little is known about the association of the TyG index with EPVS. If there is a close relationship between TyG index and EPVS, the high-risk group of EPVS will be able to be identified more easily.

Therefore, this study aims to investigate whether the TyG index is associated with EPVS burden independent of other clinical risk factors in non-diabetic adult population. Our findings may provide experimental evidence for further understanding the pathogenesis of EPVS and may provide guidance for early screening, early diagnosis, early prevention, and treatment of CSVD.

This is the first study to evaluate the relationship between IR and EPVS burden using the TyG index. In 531 patients of our study, 51.22% were observed to have moderate to severe EPVS. After adjusting for possible confounding factors, we found that the TyG index was an independent predictor of moderate to severe EPVS. In addition, our study demonstrated that the association of TyG index with moderate to severe EPVS was more significantly in low age group, male group, normotensive group, and high LDL-C group by stratified analysis. Further analysis of EPVS distribution, the TyG index was found to be associated with moderate to severe EPVS in the CSO, but not in the BG.

Several studies have reported a positive association of IR with CSVD (25–28). Nam et al. conducted a retrospective study on 2,615 Korean subjects and found that the TyG index was strongly associated with CSVD (27). In addition, the TyG index has a better predictive value for IR than HOMA-IR (27). Lee et al. found that IR was independently associated with the presence and severity of silent lacunar infarction in a healthy Korean population (26). Wu et al. found that IR assessed using HOMA-IR was an independent risk factor for EPVS in a Chinese non-diabetic population (28). However, the relationship between IR and EPVS is rarely reported. In our study, we reported an association between the TyG index and EPVS in non-diabetic patients, and this study further refines previous research results.

For the mechanism underlying the association between the TyG index and EPVS, we propose several possible explanations. First, endothelial dysfunction should be considered. Normal endothelium is essential for maintaining glucose and insulin homeostasis in the body, while insulin has a vasodilatory effect on blood vessels (30). When IR occurs in the body, excessive insulin and glucose may enhance the vasoconstrictive effects of insulin and may damage the vessel walls, resulting in a diminished ability of the endothelium to perform the physiological effects of vasodilation, vascular permeability, and maintenance of vascular tension (31, 32). When the above conditions occur in the cerebral vessels, they may cause damage to the blood-brain barrier; and, when the blood-brain barrier is damaged, components within the blood can leak into the PVS, leading to the enlargement of the PVS (33).

Second, inflammatory injury may be involved in the relationship between the TyG index and EPVS. Increased subclinical inflammation and oxidative stress are often found in patients with IR (23, 34). If the vascular endothelium is damaged by inflammation, blood contents may leak into the PVS, and these substances cannot be cleared through the vessel wall, eventually leading to the development of CSVD (35). Our study showed no association between the TyG index and WBC count. However, we found that the distribution of hs-CRP in different grades of EPVS was significantly different in the baseline data (Table 1).

Third, atherosclerosis may be the mechanism of action of IR on EPVS. In previous studies, it has been found that IR is associated with the development of atherosclerosis, and the development and rupture of plaque (13, 14, 18, 23, 34). Our study failed to confirm the relationship between the TyG index and atherosclerosis, but we observed that the TyG index was related to TC and LDL-C (Table 5). Patients with IR are often accompanied by other comorbidities, such as hypertension, diabetes, and low HDL-C levels. The TyG index may reflect their combined impact on the disease.

Finally, the pathogenesis of brain arteriolosclerosis (B-ASC) should also be considered. B-ASC is pathologically characterized mainly by thickening of the vessel wall and morphological changes, with increased tortuosity of small arteries (36). B-ASC is considered to be the underlying cause of WMH (37). In the BG and other regions associated with tortuous vessels, the CSVD markers of WMH, EPVS, and cerebral microbleeds can often be observed. In some cases, these may be related to B-ASC directly. A recent study found that intracranial arteriosclerosis was significantly associated with PVS (38). The pathological changes of intracranial arteriosclerosis are similar to B-ASC, which may be correlated with EPVS. In a quantitative ultrastructural study of blood vessels, it was found that the intracranial capillary basement membrane was markedly thickened in patients with diabetes (39). IR is the pathological basis of diabetes and is likely to be involved in the capillary wall thickening due to diabetes. Therefore, we cannot neglect the role of B-ASC between IR and EPVS.

There was a correlation between TyG index and EPVS distributed in the CSO, but not in the BG. However, the mechanism for this result is unclear. The association of CSO-EPVS with cerebral amyloid angiopathy is possibly due to β-amyloid deposition in the perivascular space, which cannot be cleared (40). We suppose that IR may primarily affect cerebral vessels close to the cortical areas, while have insignificant effects on the deep cerebral vessels. Further studies are needed to validate this speculation.

This study has several strengths. First, we used the TyG index to investigate the association between IR and EPVS burden, which has been reported to be more strongly associated with CSVD than HOMA-IR in a previous study (27). Similarly, the TyG index is more strongly associated with various adverse metabolic states than HOMA-IR (41). Second, the stratified regression analysis and subcomponent layer analysis were performed in this study to further validate the TyG index as an independent risk factor for moderate to severe EPVS. Third, our study was conducted in a non-diabetic population, thus excluding the influence of confounding factors, such as diabetes and the use of hypoglycemic agents.

There are some limitations in this study. First, this study was a retrospective observational study. Second, we excluded subjects with diabetes or taking hypoglycemic drugs, which may lead to the selection bias. However, these individuals represented only about 3.4% of the total enrolled population. Therefore, we believe that this is acceptable when interpreting our main results. Third, we used the TyG index to measure IR instead of using the hyperinsulinaemic-euglycaemic clamps, which is the gold standard in quantifying the insulin sensitivity. Because glucose clamps are expensive, time-consuming, and labor-intensive, they are not suitable for clinical and epidemiological studies. The TyG index is a more facile and currently recognized alternative (20). Fourth, because our study was a cross-sectional study, the causality could not be determined. Fifth, there was no correction for multiple independent tests, which may affect the robustness of our findings. Finally, we did not collect other imaging markers of CSVD, such as WMH, lacunae, cortical superficial siderosis, and cerebral microbleeds in this study. These markers may appear simultaneously in the same patient, and the relationship between IR and EPVS may be disturbed. Further prospective studies are needed to validate our findings.

In summary, a high TyG index was associated with a higher prevalence of moderate to severe EPVS in the non-diabetic population, independent of other clinical risk factors. Our findings suggest that the TyG index is an independent risk factor for EPVS.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Institutional Review Board of Quanzhou First Hospital Affiliated to Fujian Medical University (No. [2020]168). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

YC analyzed the data and wrote the paper. BC collected and analyzed the data. XZ and MX collected the data. XW provided the guidance for the manuscript and statistical analysis. JC conceived the study and revised the study. All authors have read and approved the submission of the manuscript.

This work was supported by grants from the Natural Science Foundation of Fujian Province (Nos. 2020J011284 and 2021J011407) and the Science and Technology Project of Quanzhou, Fujian (No. 2019C030R).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.