This longitudinal observational study consecutively enrolled patients admitted to the Department of Neurology at The Second People’s Hospital of Hefei between September 1, 2020, and July 30, 2024. The study protocol was approved by the Ethics Review Committee of The Second People’s Hospital of Hefei and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants or their legal guardians.

Inclusion criteria were as follows: (1) underwent assessment of carotid IPN using AngioPLUS ultrasound technology; and (2) completed baseline and follow-up cranial magnetic resonance imaging (MRI) with assessable WMH burden.

Exclusion criteria included: (1) severe carotid artery stenosis (defined as ≥70% stenosis); (2) presence of extensive calcified plaques precluding IPN evaluation; (3) history of intracerebral hemorrhage or large-territory ischemic infarction; (4) acute ischemic stroke occurring at baseline or during the follow-up period were excluded; (5) history of or concurrent malignancy; (6) severe dementia or behavioral disorders preventing cooperation; (7) atrial fibrillation or other cardioembolic sources; and (8) incomplete clinical/imaging data or suboptimal image quality.

Comprehensive demographic data (age, sex) and vascular risk profiles were collected upon admission. Risk factors included current smoking, alcohol consumption, hypertension, diabetes mellitus, and history of transient ischemic attack (TIA) or ischemic stroke. Body mass index (BMI) was calculated using standard protocols (see Figures 1, 2).

Fasting venous blood samples were obtained from all participants the morning after admission. A complete blood count was performed using a Sysmex XN-10 automated hematology analyzer. Biochemical parameters, including triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), fasting blood glucose (FBG), and homocysteine (Hcy), were quantified using an automated biochemical analyzer (HITACHI 7600–020, Japan).

Carotid arteries were initially evaluated using conventional B-mode ultrasonography. Upon detection of a plaque, contrast-enhanced ultrasound or microflow imaging (AngioPLUS) was performed using a SuperSonic Imagine AixPlorer system (Aix-en-Provence, France) equipped with an SL10-2 linear array transducer. This modality allows for the visualization of microvascular flow signals indicative of IPN (Zhang et al., 2017). In patients with multiple plaques, the thickest plaque was selected as the target for analysis.

IPN severity was graded based on the visual characteristics of microflow signals within a 2-min video clip, as previously described (Chen et al., 2022; Zamani et al., 2019):

For statistical analysis, IPN status was dichotomized into Low IPN (Grades 0–1) and High IPN (Grade 2). Two experienced sonographers, blinded to the patients’ clinical and MRI data, independently performed the evaluations. Discrepancies were resolved by consensus with a third senior sonologist. The inter-rater agreement, expressed by Cohen’s kappa (ĸ), was substantial for total progression at follow-up (0.72).

MR images were obtained with a1.5 Tesla (T). For this study standard axial T2-weighted fast spin echo and fluid attenuated inversion recovery sequences were used. Scanning parameters were unchanged between baseline and follow-up. All patients underwent both the carotid IPN ultrasound assessment and the baseline cranial MRI within the same hospitalization or outpatient evaluation period, with an interval of no more than 3 days between the two examinations.

Data analysis was performed using R software (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were assessed for normality using the Shapiro–Wilk test and are presented as mean ± standard deviation (SD) or median [interquartile range (IQR)], as appropriate. Between-group comparisons were conducted using the Student’s t-test or Mann–Whitney U test. Categorical variables are expressed as frequencies (percentages) and were compared using the Chi-square test or Fisher’s exact test.

The relationship between carotid IPN and the progression of WMH was assessed using binary logistic regression analysis. We performed three separate analyses: PWMH progression, DWMH progression and progression of WMH in any (periventricular and/or deep) region, respectively. All models were adjusted for cardiovascular risk factors—including age, smoking, alcohol consumption, hypertension, diabetes mellitus, history of stroke, HDL, LDL, and BMI—as well as follow-up duration. Odds ratios (OR) are given with 95% confidence interval (CI). Statistical significance was considered at p < 0.05.

Of the 245 patients initially enrolled with carotid plaque and WMH, 32 were excluded due to poor ultrasound image quality (n = 7), loss to follow-up (n = 10), cardioembolic sources (n = 3), malignancies (n = 3), moyamoya disease (n = 3), or incomplete clinical/imaging data (n = 6). Consequently, the final analysis included 213 patients (121 men, 56.8%; mean age 71.0 ± 8.9 years). The mean follow-up duration was 17.8 months (range: 4–52 months).

During the follow-up period, 124 patients (58.2%) exhibited progression of WMH, either in periventricular regions, deep regions, or both. Specifically, progression of PWMH occurred in 90 patients (42.3%), while DWMH progressed in 72 patients (33.8%). As presented in Table 1, compared with the non-progression group, patients with overall WMH progression had a significantly higher prevalence of smoking, alcohol consumption, and history of ischemic stroke, a longer follow-up duration, and elevated LDL levels (p < 0.05). Subgroup analysis revealed that smoking, history of ischemic stroke, and longer follow-up time were associated with PWMH progression (p < 0.05). In contrast, DWMH progression was significantly associated with a history of ischemic stroke, higher BMI, and higher LDL levels (p < 0.05).

To evaluate the presence of multicollinearity among the covariates included in the model, the Variance Inflation Factor (VIF) was calculated for each variable (Table 2). The VIF values ranged from 1.10 to 2.00, which are considerably below the commonly accepted thresholds of 5 or 10. These findings confirm the absence of significant multicollinearity, ensuring the reliability of the regression estimates.

Binary logistic regression analysis demonstrated that, after adjusting for age, smoking, alcohol consumption, hypertension, diabetes, history of stroke, body mass index, low-density lipoprotein cholesterol, and follow-up duration, grade 2 intraplaque neovascularization in the carotid artery was associated with the progression of deep white matter hyperintensities (OR 2.06, 95% CI 1.050–4.050, Table 3). Given the considerable variation in follow-up time among the study population, sensitivity analyses were performed. The association remained significant after excluding participants with a follow-up duration of less than 6 months (OR 2.11, 95% CI 1.040–4.310; Table 4) and further after excluding those with less than 12 months of follow-up (OR 2.89, 95% CI 1.21–6.91; Table 4). However, in the regression analyses, grade 2 intraplaque neovascularization showed no significant association with the progression of either overall white matter hyperintensities or periventricular white matter hyperintensities (see Table 5).

Several limitations should be acknowledged. First, this was a single-center study with a relatively small sample size, which limits generalizability. Second, we focused on WMH and did not assess other imaging markers of cerebral small vessel disease, such as cerebral microbleeds or enlarged perivascular spaces. Future multi-center studies with comprehensive multimodal imaging are warranted to validate these findings. Finally, in this study, both baseline and follow-up cranial MRI examinations were performed based on clinical indications rather than at pre-specified, uniform intervals according to a prospective research protocol. Although this design aligns more closely with real-world clinical practice, it may have introduced heterogeneity in follow-up timing and potentially affected the precise estimation of the rate of white matter hyperintensity progression. Additionally, while the visual rating scales employed in this study ensure clinical comparability and practicality, their sensitivity to early-stage or subtle white matter changes is limited. Currently, machine learning-based automated whole-brain WMH segmentation and volumetric quantification techniques offer higher sensitivity and objectivity, enabling precise measurement of WMH progression rates and spatial distribution patterns. The application of such automated methods may allow earlier detection of WMH evolution and reveal more nuanced associations between carotid IPN and white matter injury. Future multicenter prospective cohort studies incorporating automated quantification and multimodal neuroimaging are needed to further validate the findings of this study and to explore the dose–response relationship as well as the integrated pathological mechanisms linking IPN to WMH progression.

Our findings establish carotid intraplaque neovascularization (IPN) as an independent risk factor for deep white matter hyperintensity progression, carrying direct clinical relevance. Detection of high-grade IPN by ultrasound aids in risk stratification by identifying patients prone to rapid white matter injury, thereby informing personalized surveillance. This supports a bidirectional screening rationale: assessing carotid plaque vulnerability in individuals with progressive deep white matter lesions, and conversely, considering brain MRI screening in atherosclerosis patients with high-risk IPN. Ultimately, targeting IPN emerges as a promising therapeutic strategy for stabilizing vulnerable plaques and mitigating the progression of cerebral small vessel disease.