Critical Closing Pressure Mediates the Association of Intracranial Artery Calcification with White Matter Hyperintensities

Abstract

Introduction

This study sought to investigate the severity of intracranial artery calcification (IAC) in relation to white matter hyperintensities (WMHs), and whether the association was mediated by cerebral autoregulation (CA).

Methods

A total of 144 patients with cerebral small vessel disease were included in this study. The severity of WMH was assessed using Fazekas scores in FLAIR-magnetic resonance imaging images. On non-contrast head computed tomography images, the severity of IAC was measured by IAC scores and further classified as intimal or medial calcification. As proxy of CA, critical closing pressure (CrCP) was determined by analyzing blood pressure-flow velocity relationships in the middle cerebral artery. Mediation analyses were conducted examine the proportion of mediation of CrCP on the association between IAC and WMH.

Results

IAC scores were found to be associated with WMH scores (β 0.364; 95% confidence interval [CI], 0.133–0.409; p < 0.001). After multivariable adjustment, a statistically significant association was observed between IAC scores and higher CrCP values (β, 0.329; 95% CI, 0.129–0.528; p = 0.001). Mediation analyses revealed that CrCP partially mediated (10.3%) the association between higher IAC scores and increased WMH severity. The proportion of mediation was driven by a medial calcification pattern (13.9%).

Conclusion

This hospital-based study demonstrated the association between higher IAC scores and the severity of WMH in patients with cerebral small vessel disease, which can be partially mediated by CA as indicated by CrCP, especially for the patients with predominantly medial calcification.

Introduction

Intracranial arterial calcification (IAC) is a common finding on head computed tomography (CT) scans and is linked to ischemic stroke and cognitive impairment [1–4]. Although IAC is typically seen in large intracranial vessels, increasing evidence shows links of IAC with white matter hyperintensities (WMHs), a specific feature of cerebral small vessel disease (CSVD) [5, 6]. The mechanism underlying this association remains largely unclear.

A key feature of IAC is its location within the arterial wall [7]. Intimal calcification reflects focal atherosclerotic change, whereas medial calcification represents a non-atherosclerotic process associated with increased arterial stiffness and pulse pressure. Intimal calcification develops within atherosclerotic plaques because of lipid accumulation, inflammation, and endothelial dysfunction, leading to luminal narrowing and reduced downstream perfusion. In contrast, medial calcification arises from osteogenic transformation of vascular smooth muscle cells and calcium-phosphate deposition in the arterial media, resulting in loss of elasticity. These differences in pathophysiology suggest that medial calcification may have a stronger effect on cerebral hemodynamics and autoregulatory function, providing a biological basis for the phenotypic stratification of IAC. There are numerous methods currently available for evaluating CA, such as transfer function analysis, autoregulation index, and the pulsatility index [8–10]. Over recent years, the measurement of critical closing pressure (CrCP) using transcranial Doppler (TCD) has emerged as a significant method for quantifying CA function. CrCP, a theoretical pressure threshold below which blood vessels are presumed to collapse and cerebral blood flow approaches zero, has been used in several dynamic CA models [11–13]. Moreover, CrCP, acting as an indicator of cerebrovascular tension, has been identified as an independent predictor of WMH and CSVD burden [14]. Considering the effects of IAC on both cerebral autoregulation (CA) and CSVD, we hypothesized that CrCP might mediate the association between IAC and WMH.

Thus, we aimed in this study to test the hypothesis that CrCP mediates the association between IAC and WMH. Additionally, we investigated the mediation of CrCP in different IAC patterns to identify potential mechanisms explaining the observed association between IAC and WMH.

Methods

Study Population

From 1 May 2021 to 31 September 2022, we enrolled patients in a consecutive manner with CSVD who were hospitalized in the Department of Neurology, The Second Affiliated Hospital of Guangzhou Medical University. All participants underwent brain CT, TCD, and magnetic resonance imaging (MRI) during admission. The exclusion criteria were as follows: (1) inadequate temporal bone window for TCD; (2) poor CT imaging quality; (3) a history of stroke; (4) arrhythmias that may affect the evaluation of cerebral blood flow; (5) a history of radiotherapy for head and neck cancer.

Demographic information (e.g., age and gender) and medical histories were abstracted from participants’ electronic medical records. Blood sampling was done at morning after 8–12 h fasting duration. Low-density lipoprotein-cholesterol, triglyceride, and hemoglobin A1c were measured. Obesity was defined as BMI ≥25 kg/m² based on the cutoffs for Asian population. Hypertension was a systolic blood pressure ≥140 mm Hg or diastolic pressure ≥90 mm Hg or use of antihypertensive medication. Diabetes mellitus was hemoglobin A1c ≥ 6.0%, or a history of diabetes or use of antidiabetic therapy. This study protocol was reviewed and approved by the Medical Ethics Review Committee of The Second Affiliated Hospital of Guangzhou Medical University, approval number [2021-YJS-KS-04]. All participants provided written informed consent.

CT Acquisition and IAC Assessment

The scan was performed by using a 64-row multidetector CT scanner without contrast administered. Axial images were acquired with the following parameters: 120 kVp, 170 mA s, 1-s rotation time. CT images were independently evaluated by two readers who were blinded to any clinical information of all the participants.

Visual grading method was used to assess the scores of IAC [15]. Seven main intracranial arteries (bilateral internal carotid artery C2-C7 segments, bilateral middle cerebral artery [MCA], bilateral vertebral artery V4 segments, and basilar artery) were assessed. As previously described, the presence of calcification were above 130 Hounsfield units, the severity of IAC was evaluated by extent and thickness of calcification in individual cerebral arteries, a highest composite CT score of 0–2, 3–5, and 6–8 was classified as mild, moderate, and severe degree of IAC, respectively. The most calcified vessels were used for the final score. IAC patterns were classified using a previously established IAC scoring method [16]. Points were assigned as follows: calcification circularity: absent (0); dots (1); <90 degrees (2); 90–270 degrees (3); 270–360 degrees (4). Calcification thickness: thick ≥1.5 mm (1); thin <1.5 mm (3). Calcification morphology: indistinguishable (0); irregular/patchy (1); continuous (4). Scores of 1–6 indicated intimal calcification, and 7–11 indicated medial calcification.”

MRI Acquisition and Analysis

All research subjects underwent magnetic resonance (Achieva 1.5T MRI, Philips) examination. The MRI was set with a slice thickness of 6 mm and an interval of 1 mm. The axial cross-sectional fast gradient echo T1WI, fast spin echo T2WI, echo convert recovery sequence T2FLAIR, diffusion weighted imaging. The fields of view are 210 mm, 210 mm, 250 mm, and 240 mm; the matrices are 256 × 256, 256 × 256, 192 × 256, and 288 × 288.

CSVD is diagnosed based on its characteristic MRI features, which mainly include four imaging features: (1) lacunar infarction, (2) WMH, (3) enlarged perivascular space, and (4) cerebral microbleed. In this study, the presence of at least one of these features was defined as indicative of CSVD. WMH: hyperintensity on T2 and T2FLAIR sequences can be divided into periventricular hyperintensity and deep WMH. The Fazekas scale was used for grading, and the scores of periventricular hyperintensity and deep WMH were added together to form the final score. 1–2 was classified as mild, 3–4 as moderate, and 5–6 as severe.

Hemodynamic Data Acquisition and Analysis

An experienced technician measured hemodynamic parameters using TCD and continuous blood pressure monitor. Participants were in a supine position with their heads elevated after a 15-min rest. The TCD baseline examination was conducted with a commercial device by positioning a 2-MHz transducer at the temporal window above each zygomatic arch to assess blood flow in the MCA.

Continuous blood pressure was recorded using a tonometric monitor (CBM-7000; Colin Corporation, Japan). Initial measurements were validated against standard blood pressure readings taken with an automated arm cuff (Omega 1400 series; In vivo Laboratories Inc., Orlando, FL, USA). Continuous blood pressure signals were transmitted to the TCD machine via a dedicated cable, allowing simultaneous display of the blood pressure curve and MCA blood flow spectrum envelope in the monitoring trend window. Once the waveforms stabilized (with a change rate of less than 10% per minute), the trend graph was recorded continuously for 5 min, and the data were synchronized to the TCD machine’s hard drive.

Using Rune Aaslid’s method, we used our own offline software to calculate the value of CrCP [17]. We selected cerebral blood flow wave spectrums with complete envelopes over six consecutive cardiac cycles (covering at least one respiratory cycle).

To address the time delay between pressure and flow velocity curves at the radial artery and MCA, the flow velocity curves were adjusted by an average of 54 milliseconds. The correct time delay compensation was determined through iterative regression analysis until the hysteresis in the blood pressure/flow velocity plots was eliminated. The least squares method was applied to analyze the relationship line between blood pressure and flow velocity, with the pressure axis intercept of these plots representing the CrCP of cerebral circulation. Six consecutive cardiac cycles of each measurement period were randomly selected, and the extrapolated CrCP data from all heartbeats within these cycles were averaged for further analysis.

Statistical Analysis

The Kolmogorov-Smirnov formal test was used to assess the data distribution. Continuous variables with normal distribution were presented as mean and standard deviation or median and interquartile range when nonnormal distribution, categorical variables were presented as numbers and percentages. Our patients were divided according to the severity of calcification (mild, moderate, severe) and patterns of calcification (intimal and medial calcification). Covariables were selected based on their biological importance as cerebrovascular risk factors, and included age, sex, body mass index, smoking habits, diabetes, history of cardiovascular diseases, and use of antihypertensive treatment.

Our analyzes consisted of (1) examining the association of IAC scores and CrCP values with WMH by applying multivariable regression modeling; (2) addressing the association between IAC scores and CrCP values; and (3) the mediation analyses. All analyzes were conducted in the whole sample, and we also combined participants into 2 groups: (1) participants with absent IAC plus participants with intimal calcification and (2) participants with absent IAC plus participants with medial calcification.

Statistical analyses were performed using the SPSS software version 26 (IBM, Chicago, IL, USA). Forest plots were performed using the statistical software Stata 16.0 Version.

Results

Characteristics of the Study Population

Among the 157 consecutive patients initially examined or considered for inclusion in this study, 2 were excluded because of incomplete clinical data, 2 for image artifacts on CT, 4 for poor TCD signal quality. In addition, we also excluded individuals who had serious health conditions, such as atrial fibrillation, infection, anemia, or cancer (n = 5). After exclusion, a total of 144 participants remained eligible in the present study. Among the included patients, the mean age was 63.3 years old, and 33.3% of the participants were women. The basic characteristics are presented in Table 1.

Table 1.

Characteristics of the study participants

Characteristic	All participants (n = 144)	Mild IAC (n = 71)	Moderate IAC (n = 53)	Severe IAC (n = 20)
Age, years	63.3±9.7	58.0±8.8	66.0±7.5	72.0±7.8
Female, n (%)	48 (33.3)	31 (43.7)	15 (28.3)	2 (10.0)
Current smoking, n (%)	44 (30.6)	15 (21.1)	20 (37.7)	9 (45.0)
BMI, kg/m2	23.8±2.9	23.5±3.0	23.9±2.5	24.4±3.1
Hypertension, n (%)	64 (44.4)	19 (26.8)	31 (58.5)	13 (65.0)
Diabetes mellitus, n (%)	27 (18.8)	6 (8.5)	12 (22.6)	9 (45.0)
LDL-C, mmol/L	2.8±1.0	2.9±0.9	2.6±0.9	2.4±0.8
TG, mmol/L	1.6±1.0	1.5±0.9	1.5±0.8	1.9±1.7
HbA1c, %	6.05±1.2	5.8±0.8	6.2±1.4	6.7±1.6
CrCP, mm Hg	27.9±10.2	25.1±12.3	31.6±9.7	35.7±9.4
WMH, n (%)
No-mild	95 (66.0)	60 (84.5)	28 (52.8)	8 (40.0)
Moderate-severe	49 (34.0)	11 (15.5)	25 (47.2)	12 (60.0)

Categorical variables are shown as number (%); continuous variables as mean ± standard deviation.

BMI, body mass index; LDL-C, low-density lipoprotein-cholesterol; TG, triglyceride; HbA1c, glycated hemoglobin; CrCP, critical closing pressure; WMH, white matter hyperintensity.

Association of IAC and CrCP with Moderate-Severe WMH

As shown in Table 2, in all participants, the severity of IAC was associated with WMH values in univariate regression analysis (β, 0.545; p < 0.001). In multivariate analyses, IAC scores remained independently associated with the WMH scores in all adjusted models (β, 0.456; p < 0.001 in model 1; β, 0.364; p < 0.001 in model 2). Similarly, a significant relationship was observed between WMH scores and higher CrCP values in univariate (β, 0.328; p < 0.001) and multivariate analyses (β, 0.240; p = 0.003; β, 0.211; p = 0.01).

Table 2.

Association of IAC and CrCP with WMH scores

Models	IAC scores β (95% CI)	p value	CrCP β (95% CI)	p value
Unadjusted	0.545 (0.302, 0.509)	<0.001	0.033 (0.027, 0.078)	<0.001
Model 1	0.456 (0.214, 0.464)	<0.001	0.024 (0.014, 0.063)	0.003
Model 2	0.364 (0.133, 0.409)	<0.001	0.021 (0.008, 0.060)	0.010

Model 1: adjusted for age and gender.

Model 2: further adjusted for history of hypertension, history of diabetes and smoking status, BMI, TG, and LDL.

Relations of IAC Scores with CA Assessed by CrCP

The relations of IAC scores with CrCP values based on linear regression models are presented in Table 3. Similarly, a significantly relationship was also found between elevated IAC scores and CrCP in univariate regression analysis (β, 0.380; p < 0.001), and it remained significant after adjustment for models 1, 2, and 3 (β, 0.346; p < 0.001 in model 1; β, 0.359; p < 0.001 in model 2; β, 0.329; p = 0.001).

Table 3.

Associations of IAC scores with CA assessed by CrCP

Models	β	CI of β	Sig
Univariate	0.380	0.225, 0.534	<0.001
Model 1	0.346	0.159, 0.533	<0.001
Model 2	0.359	0.162, 0.556	<0.001
Model 3	0.329	0.129, 0.528	0.001

Model 1: adjusted for age and gender.

Model 2: further adjusted for history of hypertension, history of diabetes and smoking status.

Model 3: further adjusted for BMI, TG, LDL.

Mediation Analysis: IAC→ CA → WMH

To further test the potential mediating effects of CrCP on the IAC-WMH association, we conducted causal mediation analysis and revealed that the association between IAC and WMH was partially mediated by CrCP in all participants, accounting for a mediation ratio of 10.3% (Fig. 1). The coefficients of direct effect and indirect effect were 0.340 (p < 0.001) and 0.035 (p < 0.05), respectively. When considered IAC as a continuous variable (IAC scores), we observed that the proportions of mediation (%) by CrCP on the association between IAC and WMH were higher in participants with predominant medial calcification (7.9%) (Fig. 2) compared to those with predominant intimal calcification (4.5%) (Fig. 2). However, indirect effects were not significant in either group (p > 0.05). When considering IAC as a categorical variable (intimal, medial, absent), we found that the proportion of mediation (13.9%, indirect effect = 0.080, p < 0.05) was higher in participants with intimal calcification and non-calcification IAC (−8.5%, indirect effect = −0.072, p > 0.05. Fig. 3).

Fig. 1.

Mediation effects of CrCP on the IAC-WMH association in all participants.

Fig. 2.

Mediation effects of CrCP on the IAC-WMH association among participants with medial and intimal calcification.

Fig. 3.

Participants with medial calcification vs. intimal calcification and calcification absent and participants with intimal calcification vs. medial calcification and calcification absent.

Discussion

The aims of this study were to investigate the cross-sectional association of IAC and WMH among patients with CSVD and to test whether CA function assessed by CrCP could be a mediator in this relationship. Our findings suggest that elevated IAC scores were associated with the severity of WMH, and that this association may be partly mediated by CA, particularly in patients with a predominance of medial calcification.

IAC is thought to be closely linked to dementia, a link that may be explained in part by structural brain changes including CSVD. In a study involving 2,339 stroke‐free and dementia‐free participants from the general population, Bos et al. [18] found that IAC of both the anterior and the posterior cerebral circulation, increased the risk of dementia, partly mediated by increased WMH volume. In this study, we found that IAC correlated with WMHs, which is similar to the results of certain previous studies. Chen et al. [5] found a significant correlation between cerebral arterial calcification and CSVD, particularly with WMHs and lacunes. In the population-based Rotterdam Study, Bos et al. demonstrated a similar association, indicating that a larger burden of IAC may be linked to the accelerated progression of small vessel disease [19]. In our previous study, we found that in a cohort of acute stroke patients, medial IAC was correlated with the occurrence of WMH. Although the precise relationship between IAC and CSVD has been confirmed by numerous studies, the underlying mechanisms by which IAC influences small vessel pathology remain unclear.

Based on pathological and imaging studies, IAC can be classified into medial calcification and intimal calcification. Intimal calcification primarily contributes to the formation of atherosclerotic plaques, whereas medial calcification increases vascular stiffness and reduces vascular compliance [20–22]. CrCP, which has been defined as an arterial pressure threshold below which arterial vessels collapse, is an indirect measure of cerebrovascular tone and resistance. CrCP can represent CA because it reflects vascular smooth muscle tone, which adjusts cerebral blood flow in response to changes in arterial pressure. When autoregulation is intact, CrCP decreases during hypotension and increases during hypertension, maintaining stable perfusion. Unlike transfer function analysis or the autoregulation index, which describe the effects of autoregulation through pressure-flow coupling, CrCP quantifies the mechanism – the active modulation of vascular tone. Therefore, CrCP provides a mechanistic and physiologically indicator of cerebral autoregulatory capacity. In our study, we observed that higher IAC burden was associated with impaired CA, suggesting a potential pathway through which IAC could influence CSVD. This association appeared more evident among patients with predominant medial calcification. The relationship between CrCP, calcification, and WMH may be explained by stiffness-related hemodynamic changes. Medial calcification increases arterial stiffness and pressure pulsatility, which elevate CrCP, indicating higher vascular tone and reduced compliance. Elevated CrCP reflects impaired autoregulatory capacity and higher perfusion thresholds, which may cause chronic hypoperfusion or ischemic damage in small vessels. This mechanism may explain how arterial calcification contributes to WMH development through impaired cerebrovascular regulation.

CSVD is an important risk factor for stroke and dementia, which seriously threatens the health and quality of life of the elderly [23–27]. Due to its hidden onset and lack of early diagnosis and intervention methods, its management is more complicated. This study found that the degree of IAC can reflect the severity of CSVD, suggesting that quantitative assessment of arterial calcification may be helpful in clinical evaluation of CSVD risk. Additionally, this study shows that different patterns of IAC have different effects on CSVD, highlighting the need to develop targeted prevention strategies for CSVD for specific types of calcifications. Finally, although intensive blood pressure therapy is widely used for the treatment of cerebrovascular diseases, including CSVD, it is unclear whether it causes cerebral hypoperfusion or worsens prognosis due to differences in individual cerebral blood flow autoregulation. Therefore, individualized blood pressure regulation methods based on different calcification types and CA may have more clinically significant in the future.

The results of our analysis are subject to several limitations. First, the cross-sectional nature of our study allows the investigation of only associations, which needs to be further validated a causal association between IAC and WMH in prospective cohort studies. In addition, WMH itself may exacerbate dysregulation of cerebral hemodynamics by reducing vascular reactivity, thereby contributing to a self-reinforcing cycle of hypoperfusion → WMH → further hypoperfusion. Second, the sample size is small due to the challenges faced in imaging and hemodynamic data acquisition and analysis. The evaluation of CA requires a high-quality TCD blood flow spectrum. Therefore, many elderly participants with either no temporal window or a narrow temporal window could not be included in this study. Furthermore, several patients have both intimal and medial calcifications. However, for statistical convenience and considering the small number of cases in this study, we just divided all patients into two groups: those with predominant medial calcifications and patients with mainly intimal calcifications. Future studies that divide the data into three groups (only medial calcification, only intimal calcification, and both media and intimal calcification) will be more helpful in exploring the relationship between different patterns of IAC and CSVD. Finally, in this study, CrCP was derived from the MCA. Future studies incorporating vessel-specific CrCP measurements may help clarify the regional relationship between IAC and cerebrovascular function.

In summary, more severe IAC was associated with higher WMH scores in patients with CSVD. A substantial portion of the relationship between IAC and CSVD markers may be attributable to impaired CA, as suggested by causal mediation analysis, especially for the patients with predominantly medial calcification.

Highlights

Higher IAC scores may contribute to the severity of WMH. The CrCP serves as a mediator between IAC and WMH. The mediating effect of CrCP was mainly in patients with media calcification.

Acknowledgments

We express our heartfelt gratitude to all the individuals who participated in this research endeavor.

Statement of Ethics

This study protocol was reviewed and approved by the Medical Ethics Review Committee of The Second Affiliated Hospital of Guangzhou Medical University, Approval No. [2021-YJS-KS-04]. All participants provided written informed consent.

Conflict of Interest Statement

The authors have declared no conflict of interest.

Funding Sources

This study was supported by the Clinical Featured Technology Program of China (OF04024), and the Guangzhou Science and Technology Plan Project (202201010960).

Author Contributions

Conceptualization: Xuelongli and Xiangyan Chen; data curation: Gaoxian Zhong; formal analysis: Xianliang li; funding acquisition: Xiangyan Chen; investigation: Qingchun Gao; methodology: Xuelongli, Daniel Bos, and Xiangyan Chen; project administration: Gaoxian Zhong; resources: Junru Chen; software: Qingchun Gao and Gaoxian Zhong; supervision: Xiangyan Chen; validation; Xianliang li and Qingchun Gao; visualization: Gaoxian Zhong and Junru Chen; writing – original draft: Xuelongli; writing – review and editing: Daniel Bos and Xiangyan Chen.

Funding Statement

This study was supported by the Clinical Featured Technology Program of China (OF04024), and the Guangzhou Science and Technology Plan Project (202201010960).

Data Availability Statement

The datasets analyzed during the current study are not publicly available due to patient privacy but are available from the corresponding author upon reasonable request.