Voxel-based morphometry reveals gray matter atrophy in cerebral small vessel disease: correlation with cognitive impairment

Introduction

Cerebral small vessel disease (CSVD) is a group of clinical, imaging, and pathological syndromes caused by damage to small cerebral arteries and their distal branches, traditional small arteries, microarterioles, capillaries, microvessels, and small veins due to a variety of etiological factors. It is characterized by an insidious onset, complexity of symptoms, heterogeneity, dynamic changes, and preventability (1). Approximately 45% of dementia patients also have CSVD, which is an important subtype causing vascular cognitive impairment (VCI) and often interpenetrating with Alzheimer’s disease (AD) (2-4). However, as the pathogenesis of CSVD is not fully understood, there are no clinical targeted interventional treatment options currently available. Therefore, the systematic elaboration of the possible pathogenic mechanisms of CSVD and the screening of biological markers with diagnostic value are of great clinical value and social significance for the early identification and treatment of CSVD-associated VCI.

The Neuroimaging Vascular Alterations Reporting Criteria were updated with eight imaging markers of CSVD, namely, recent small subcortical infarcts, lacunae, cerebral white matter high signal, perivascular gap enlargement, microhemorrhages, iron deposition on the surface of the cerebral cortex, intracranial hemorrhage and other hemorrhagic signs, and cerebral atrophy (5,6). A study found cerebral gray matter volume (GMV) changes to be an independent predictor of cognitive impairment (CI) in CSVD (7). Brain atrophy in total CSVD imaging load is less well studied and still in disagreement, with research reporting that brain atrophy and white matter hyperintensity (WMH) are better predictors than age for assessing changes in cognitive function (8). A growing number of cognitive studies are focusing on brain atrophy, in addition to conventional magnetic resonance imaging (MRI) examination to assess CSVD and its severity. At present, there is a tendency to use automated methods to assess brain atrophy, hoping to more accurately assess the degree of brain atrophy (9,10). Since brain gray matter activity is related to neuronal activity, exploring the characteristics of structural brain changes in patients with CSVD is important for understanding the brain mechanisms underlying the complications of CI.

The core objective of this study was to quantify GMV alterations and their correlation with neuropsychological scale scores in patients with CSVD with or without CI using voxel-based morphometry (VBM) and to compare their diagnostic performance. The focus was on resolving intergroup differences in CSVD patients with CI (CSVD-CI) on the dimensions of brain atrophy patterns, and on revealing in-depth neuroanatomical associations between impaired cognitive function and structural alterations in specific brain regions. The study will provide multimodal imaging evidence for exploring the neuropathological mechanisms of CSVD-CI and is expected to build a system for assessing disease progression based on structural imaging markers. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-783/rc).

Methods

Participants

This cross-sectional study involved patients with CSVD attending the neurology outpatient clinics or inpatient wards of the First Affiliated Hospital of Jinzhou Medical University between October 2023 and August 2024. A total of 100 cases were included as participants according to the inclusion and exclusion criteria, including 52 patients (23 males and 29 females) with CSVD-CI, aged 51–80 years, and 48 patients (22 males and 26 females) with CSVD with normal cognitive (NC) function, aged 51–77 years. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The First Affiliated Hospital of Jinzhou Medical University Medical Research Ethics Committee approved the study (approval No. KYLL 2023150). Each participant signed an informed consent form.

The inclusion criteria were as follows: age range of 50–80 years; suspected CSVD on cranial MRI, based on current MRI consensus criteria (6) (presence of lacunae, white matter high signal, perivascular gap enlargement, microhemorrhages) and total imaging load score ≥1; daily living assessment showing no functional impairment; and ability to cooperate in completing neuropsychological scale assessment. The exclusion criteria were as follows: CI due to neurodegenerative diseases other than cerebrovascular disease, such as AD and Parkinson’s disease, and acquired brain injury (including inflammatory lesions of the central nervous system, such as multiple sclerosis, infectious lesions, space-occupying lesions, and traumatic brain injury); neurological impairment related to substance dependence or intoxication (alcohol abuse and psychotropic substance dependence); and combined important organ dysfunctions such as heart disease, liver dysfunction, chronic kidney disease, tumor or other systemic diseases; severe acute severe cardiovascular and cerebrovascular disease, anxiety, and depression; and contraindications to cranial magnetic resonance (MR) examination or similar, who are unable to cooperate with the completion of the examination.

Cognitive function and neuropsychological assessment

Every participant received a thorough evaluation that included neuropsychological, neurological, and physical tests. The Mini-Mental State Examination (MMSE) and the Beijing version of the Montreal Cognitive Assessment (MoCA) were part of the neuropsychological evaluation (11,12). To classify cognitive educational status, individuals were categorized based on their scores on the cognitive assessments: those who scored ≤13 points in the illiterate group, ≤19 points in the primary school group, and ≤24 points in the junior high school and above group were all deemed to have objective CI. Overall cognitive function, which includes several cognitive domains including attention and focus, executive function, memory, language, visuospatial skills, abstract reasoning, computation, and orientation, was mainly assessed using the MoCA score. All participants’ demographic and health-related data, including age, gender, years of education, body mass index (BMI), and medical history concerning hypertension, diabetes, hyperlipidemia, as well as smoking and alcohol use, were gathered as baseline data.

MRI scanning

A Siemens 3.0T MAGNETOM Vida (Siemens, Erlangen, Germany) high field strength MR scan was performed using a standard 64-channel cranial matrix coil, and the anterior conjugate-posterior conjugate line was used as the scanning baseline. During the scanning, the participants wore earplugs and were placed in a supine position, and were asked to close their eyes during the examination to avoid the effects of eye movements. In addition, in order to minimize the adverse effects of artefacts generated by head movements on the quality of the data, participants were instructed to keep their whole body still and had their head fixed with foam on both sides.

A cranial MRI was performed including sequences such as T2-fluid-attenuated inversion recovery (T2-FLAIR), three-dimensional-T1-weighted imaging (3D-T1WI) sequence, and susceptibility-weighted imaging (SWI). A 3D magnetization-prepared rapid acquisition gradient echo (MPRAGE) was employed to obtain T1WI. Repetition time (TR) =2,000 ms, inversion time (TI) =712 ms, echo time (TE) =2.63 ms, flip angle (FA) =8°, and section thickness =1 mm were the parameters set for this sequence. In addition, T1WI, diffusion-weighted imaging (DWI), and SWI sequences were obtained to assess intracerebral structural abnormalities, CSVD severity, and total load scores.

CSVD load score

The films were read independently by two 5-year neurological imaging physicians (each with ≥5 years of practice) in a double-blind method, and the decision was made by consensus in the case of disagreement. The Fazekas score (0–3) using Fazekas periventricular WMH (PVWMH) and deep WMH (DWMH) impairments and the recently adopted CSVD total load score (0–4), which is recommended to predict cognitive decline, were used to assess the severity of CSVD (12,13). To calculate the CSVD score, one point was awarded for each of the following outcomes: ≥1 lacunar, ≥1 cerebral microbleed (CMB), high WMH (Fazekas score =3 in PVWMH or ≥2 in DWMH), and moderate to severe enlarged perivascular spaces (EPVS) (basal ganglia area >10) were reported. The improved CSVD total load score was 0–6, with 1 additional point for each of the following: ≥5 CMB lesions; and WMH (total score of 5 or 6) (9). Medial temporal lobe atrophy (MTA) visual score (14): in T1-weighted (T1W) coronal hippocampus, the hippocampus was scored from 0 to 4 according to hippocampal height, choroidal width, and temporal horn width: 0: no atrophy; 1: mild widening of the choroidal fissure; 2: concomitant widening of the temporal horn of the lateral ventricle; 3: moderate reduction in hippocampal volume (decreased height); and 4: severe reduction in hippocampal volume.

MRI data preprocessing and analysis

All raw data were ranked to exclude images affected by artefacts, occupancy, or other intracerebral lesions, and so on, and to control the quality of the raw data. VBM data processing: 3D-T1-weighted images were preprocessed using the VBM method in the MATLAB 2017a (MathWorks, Natick, MA, USA) environment using the Computational Anatomy Toolbox (CAT12; Structural Brain Mapping Group, Jena University Hospital, Jena, Germany; http://www.neuro.uni-jena.de/cat/) software fragment plug-in in the Statistical Parametric Mapping 12 (SPM12; The Wellcome Centre for Human Neuroimaging, London, UK; http://www.fil.ion.ucl.ac.uk/spm/software/spm12/) software toolbox (15).

The steps are as follows: (I) format conversion: 3D-T1WI raw image data in Digital Imaging and Communications in Medicine (DICOM) format were converted into neuroimaging informatics technology initiative (NIfTI) format data, and the image quality was checked again after conversion; (II) origin correction: all participants’ images were origin corrected to reduce alignment errors in the later stages; (III) spatial normalization: participants’ transformed structural images were aligned to Montreal Neurological Institute (MNI) to eliminate spatial location differences between different samples (16,17); (IV) tissue image segmentation: the standardized structural brain images were effectively segmented to accurately differentiate white matter, gray matter, and cerebrospinal fluid; (V) quality inspection: the segmented gray matter images were quality inspected on CAT12 software; (VI) homogeneity detection: based on the CAT12 software, the voxel-based cortical signal heterogeneity detection was performed using the voxel-based cortical signal heterogeneity test method to assess the quality homogeneity of the gray matter images obtained from segmentation, to ensure that the spatial distribution of the data meets the requirements of the subsequent analysis; (VII) smoothing: set the smoothing kernel size [full-width at half-maximum (FWHM) =8 mm] (18), and optimize the smoothing of the segmented gray matter images by reducing high-frequency noise interference to improve the signal-to-noise ratio at the gray matter-white matter boundary for subsequent statistical analysis; and (VIII) calculation of whole cranial brain volume: the whole brain volume parameter was estimated using the Get TIV function of the CAT12 software on the segmented 3D-T1WI images and used as a covariate for the subsequent VBM analysis in order to control for the effect of the intracranial volume differences of the participants on the results.

Statistical analysis

Diagnostic agreement between the two physicians assessing the CSVD total load visual score was analyzed using a linearly weighted kappa test. Kappa ≤0.20 was considered poor agreement, 0.20< kappa ≤0.40 indicated fair agreement, 0.40< kappa ≤0.60 was considered moderate agreement, 0.60< kappa ≤0.80 indicated good agreement, and 0.80< kappa ≤1.00 represented excellent consistency. For the measurement data, normality test was performed first, and if the data conformed to normal distribution, it was expressed as mean ± standard deviation (x±SD). Independent samples t-test was used to compare differences between age groups of CSVD patients. Measurement data that were not normally distributed, expressed as median and interquartile spacing, were compared using the two independent samples Mann-Whitney U nonparametric test. For count data expressed as frequencies and percentages, the Chi-squared test was used to analyze the comparison of between-group differences in gender, years of education, cerebrovascular disease risk factors, total CSVD score, and modified CSVD score between the groups of patients with and without CI in CSVD. Subsequently, the presence of CI was used as the dependent variable, and statistically significant variables between the groups were used as independent variables for covariate diagnosis, and binary multifactorial logistic regression analyses were used to determine the independent risk factors in the CI group, and all statistical results were considered statistically different at P<0.05.

VBM statistical analysis

The 3D-T1WI image data of all participants were statistically analyzed using SPM12 software. Statistical analysis of the smoothed images was performed using a general linear model and a two-sample t-test, with age, sex, and whole brain volume as covariates to remove effects, voxel-level P<0.001 as a threshold, and cluster-level family-wise error (FWE) correction to statistically obtain a cluster size threshold of 910, which was corrected to show the brain regions with significant differences in GMV between the two groups. Visualization was performed using the MATLAB-based xjView toolkit and BrainNet toolkit. The locations of brain regions with statistically significant GMV differences, number of voxels, MNI coordinates, and T-values (peak intensity of voxel regions) were recorded and plotted into a table. Finally, GMV data were extracted using CAT12 software region of interest (ROI) extraction on the peak point coordinate regions of the difference brain regions between the CSVD-CI group and those without CI for subsequent analysis.

Brain regions with GMV changes were extracted, and SPSS partial correlation analysis was applied to assess the correlation between local GM volume changes and MoCA and MMSE scores. Further Spearman correlation analysis was performed to ascertain the correlation between MoCA scores in different cognitive domains and GMV in distinct brain regions. A P value <0.05 was considered statistically significant; r>0 was considered a positive correlation; and r<0 was considered a negative correlation.

Area under the curve for the receiver operating characteristic (ROC) curve

Finally, the modified CSVD total load score, hippocampal atrophy score and grey matter atrophy parameters of patients in the CSVD-CI group and the group without CI were analyzed by ROC curve analysis, and a paired comparison was performed. The difference of each index was statistically significant at P<0.05.

Results

Demographic and clinical characteristics of the participants

The test of agreement between the two physicians assessing the total CSVD load score was excellent, kappa =0.930 [95% confidence interval (CI): 0.869–0.991]. The CSVD burden score was as follows: 9 cases (17.3%) in the CSVD-CI group were 1 point, 19 cases (36.5%) were 2 points, 15 cases (28.8%) were 3 points, and 9 cases (17.3%) were 4 points. In the NC group, 23 cases (47.9%) scored 1 point, 14 cases (29.2%) scored 2 points, 8 cases (16.7%) scored 3 points, and 3 cases (6.3%) scored 4 points.

Univariate analysis showed that the CMB, WMH patient composition ratio, CSVD total load score, MTA score, age, and years of education in the cognitively impaired group were statistically different from those in the non-CI group (P<0.05), whereas other CSVD phenotypes, as well as the vascular risk factors, and the gender differences, were not statistically significant (as Table 1 for details). Firstly, the variables with P<0.05 in Table 1 were used as independent variables for collinearity diagnosis, and collinearity factors were excluded. Whether there was CI was used as a dependent variable (cognitive normal =0, CI =1). After adjusting for confounding factors such as age and years of education, binary multivariate logistic regression analysis was performed. MTA [odds ratio (OR) 2.472, 95% confidence interval: 1.029–5.938, P=0.043] was an independent risk factor for CI in patients with CSVD (see Table 2).

VBM analyses to measure GMV

Compared with the NC group, GMV decreased in the left lingual gyrus, parahippocampal gyrus, fusiform gyrus, talar fissure, hippocampus, insula, middle temporal gyrus, and dorsal thalamic nucleus in the CSVD-CI group and decreased in the right temporal pole, fusiform gyrus, parahippocampal gyrus, fusiform gyrus, amygdala, and dorsal thalamic nucleus, with the most significant reduction in the left lingual gyrus, the dorsal thalamic nucleus, and the right lingual gyrus as well as the amygdala, and no volume increase in brain regions (P<0.001 at the voxel level, cluster >910 corrected for cluster level FWE, as shown in Figure 1 and Table 3).

Figure 1 Brain map of GMV difference between the CSVD cognitive impairment group and cognitive normal group. (A) Differential brain regions are shown in cross-sectional layers using the xjView toolkit. The yellow areas represent the brain regions with reduced GMV in the CSVD-CI group. (B) Differential brain regions are shown from top, bottom, left, right, front, and back using the BrainNet toolkit, and the blue areas represent the brain regions with reduced GMV in the CSVD-CI group (P<0.001, corrected by mass level FWE, cluster size ≥910). CSVD, cerebral small vessel disease; CSVD-CI, CSVD patients with cognitive impairment; GMV, gray matter volume; FWE, family-wise error.

Analysis of the correlation between GMV changes and clinical scores

Controlling for the effects of confounding factors such as age, sex, and years of education, brain regions with significant positive correlations between ROI brain GMV and total MoCA score in patients with CSVD with or without CI included: lingual gyrus left (r=0.423, P<0.001), lingual gyrus right (r=0.367, P<0.001), amygdala right (r=0.424, P<0.001), and mid-dorsal nucleus of the thalamus left (r=0.439, P<0.001). Brain regions where ROI brain GMV was significantly and positively correlated with total MMSE score included: lingual gyrus left (r=0.286, P=0.005), lingual gyrus right (r=0.277, P=0.006), amygdala right (r=0.492, P<0.001), and thalamus mid-dorsal nucleus left (r=0.381, P<0.001) (see Figure 2).

Figure 2 Analysis of biased correlation between GMV and MoCA and MMSE scores in different brain regions in the CSVD cognitive impairment group. Biased correlation analysis of GMV of the left lingual gyrus, right lingual gyrus, right amygdala, and left mid-dorsal thalamic nucleus with MoCA and MMSE scores, controlling for the effects of confounders such as age, sex, and years of education. CSVD, cerebral small vessel disease; GMV, gray matter volume; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment Score.

The correlation between MoCA scores in different cognitive domains and GMV in different brain regions was further assessed, as demonstrated in Figure 3. Spearman’s analysis revealed a correlation between left lingual gyrus gray matter atrophy and visuospatial and executive function, attention, language, and delayed recall scores. The correlation was most significant in the domain of attention (r=0.428, P<0.001). Right lingual gyrus gray matter atrophy correlated with visuospatial and executive functioning, naming, attention, language, and delayed recall, with the greatest correlation being in the area of attention (r=0.383, P<0.001). Right amygdala volume changes correlated with all cognitive subdomains, mainly in delayed memory (r=0.411, P<0.001) and attention (r=0.409, P<0.001). Volume changes in the mid-dorsal nucleus of the thalamus correlated with attention, language, delayed recall, and orientation, focusing mainly on orientation (r=0.323, P=0.001) and language (r=0.319, P=0.001).

Figure 3 Correlations between the GMV in different brain regions and scores in different cognitive domains in MoCA. Heatmap of correlation coefficient between significantly different GMV and scores in cognitive subdomains in the CSVD-CI and CSVD-NC groups. *, P<0.05; **, P<0.01; ***, P<0.001. CSVD, cerebral small vessel disease; CSVD-CI, CSVD patients with cognitive impairment; CSVD-NC, CSVD patients with normal cognition; GMV, gray matter volume; MoCA, Montreal Cognitive Assessment Score.

ROC curves to analyze CSVD patients with and without CI

In general, ROC analysis showed that the area under the ROC curve (AUC) areas of the modified CSVD total load score, MTA, right amygdala GMV, and the combined model of the three were 0.699, 0.674, 0.727, and 0.801, respectively. The AUC area of the difference brain region value of the three combined models was the best, with an AUC of 0.801, sensitivity of 0.713, and specificity of 0.889. There was no statistical difference in the AUC of the three, and the AUC of the three combined was better than that of any single factor alone (details are shown in Figure 4 and Tables 4,5).

Figure 4 ROC curve of imaging model in predicting cognitive impairment. ROC analysis showed that the AUC of the modified CSVD total load score, MTA, right amygdala GMV, and the combined model of the three were 0.699, 0.674, 0.727 and 0.801, respectively. AUC, area under the ROC curve; CSVD, cerebral small vessel disease; GMV, gray matter volume; MTA, medial temporal lobe atrophy; ROC, receiver operating characteristic.

Discussion

Chronic CSVD has been observed to exhibit clinical insidiousness, with some researchers proposing the classification of CSVD as a whole-brain disease. This assertion is underpinned by the progressive nature of the condition, which initiates with the formation of an initial lesion in a specific brain region and subsequently disseminates to distant regions, resulting in a cascading effect that encompasses the entire brain tissue and clinical junctions. This process under discussion has been demonstrated to exert an influence on the manifestation of impairments in a variety of domains, including cognition, gait balance, emotion, and the development of diabetic complications (19,20). A number of studies have investigated the correlation between imaging changes, such as small subcortical infarcts, white matter high signal, microbleeds, and perivascular gap enlargement, and CI (21,22). However, given that the majority of elderly individuals also exhibit lacunar cerebral infarcts, white matter high signal, and other changes, there exists heterogeneity in the association of cognitive decline with aging (23). Furthermore, cognitive status is influenced by additional factors, including social experience and educational level. Consequently, determining the precise impact of severity of CSVD on CI remains a highly challenging endeavor.

In this study, statistical analysis revealed that although there were intergroup disparities in CSVD imaging loading scores and modified CSVD scores within the CI group, these disparities did not attain statistical significance following the incorporation of binary logistic regression analysis to adjust for confounding variables. This may be attributable to the uneven distribution of participant age, the overall low years of education, and the inadequate sample size. MTA emerged as a stronger independent predictor, highlighting the role of neurodegenerative changes superimposed on vascular pathology. This paper focuses on the changes of GMV in CSVD-CI, and further analyzes the correlation with cognitive subfields to compare the diagnostic efficacy, so as to seek more imaging evidence related to CSVD-CI.

The VBM technique enables voxel-level quantitative analysis of gray and white matter volumes by means of morphological segmentation algorithms with high-resolution T1WI to localize gray matter atrophy that is significantly correlated with executive function, memory encoding, and processing speed. Brain atrophy and CI can also be detected early in the onset of CSVD. The VBM analysis in this study showed that gray matter atrophy in patients with CI in CSVD mainly ranged from the bilateral temporal lobe, occipital lobe, and thalamus, with significant atrophy in the lingual gyrus, amygdala, and thalamus mid-dorsal nucleus of the cortex, as compared with that in the group with no CI in CSVD. It has been shown that patients with CSVD may be accompanied by extensive volumetric atrophy of cerebral gray matter (24,25). It may occur independently of white matter, but the underlying cellular mechanisms that allow structural changes in specific brain gyrus regions have not been fully investigated. It has been shown that extensive degeneration of GM structures in CSVD patients, especially bilateral temporal lobe and hippocampal volume atrophy, is closely associated with CI (26,27). In addition, brain regions with reduced GMV in VCI patients with varying degrees of CI were also found in the frontal, parietal, and limbic brain regions. The results of the present study are not entirely consistent with those of previous studies. This study suggests that cognitive deficits in patients with CI in CSVD may be accompanied by temporal lobe and thalamic atrophy, which may produce cascading damage and consequently abnormal coupling of deep gray matter and functional structures.

Brain atrophy in patients with CSVD is influenced by a variety of factors, which may be the first microvascular alterations, which in turn trigger a series of cascading reactions (28). The pathological mechanism of CSVD-associated gray matter atrophy is hypothesized to be related to retrograde degeneration through axonal–neuronal degeneration. The pathogenesis of microbleeds-associated brain atrophy is due to the excessive deposition of amyloid white matter in the cerebral vasculature, leading to insufficient cerebral perfusion and abnormal connectivity among neurons. Consequently, white matter ischemic changes occur, resulting in the destruction of the brain’s structural network and eventually leading into a reduction in brain volume (29,30). The imaging study has shown that WMH severity promotes the onset of brain atrophy, which in turn leads to further cognitive decline, a process that may accelerate neurodegeneration through disruption of white matter fiber integrity, which in turn exacerbates cognitive decline (31).

Consistent with previous studies assessing changes in brain volume, these regions of brain atrophy are associated with a variety of cognitive functions, such as attention, executive function, visuospatial ability, and language ability (32,33). Lambert et al. found that volume atrophy in the hippocampus and medial temporal lobe was associated with CSVD and was not a specific brain region for the onset of AD (34). This is consistent with the distributional features of gray matter atrophy in the temporal lobe hippocampus and amygdala in the present study, and found that the related CI fields focus on delayed memory and attention. This suggests that CSVD may cause neurodegeneration through ischemic injury, which in turn affects limbic system microstructure. It has also been claimed that reduced amygdala size may be associated with the development of anxiety and depression (35). The bilateral hippocampus is a hub for brain memory network interaction and is vulnerable to ischemic damage (36,37). Atrophy of the parahippocampal gyrus and the fusiform gyrus may mediate temporal, spatial, and other situational memory deficits and face recognition difficulties in patients (38,39). In this study, we found bilateral temporo-occipital and thalamic atrophy in the CI group; cortical atrophy in the lingual gyrus of the occipitotemporal lobe was closely related to impaired attention, delayed memory, and visuospatial and executive functions, and cortical atrophy in the mid-dorsal nucleus of the thalamus was related to orientation, language, and attention. Analyzed from a neural loop perspective, this may suggest a compensatory role for cortico-striatal-thalamic connectivity loops (40). The inferotemporal gyrus of the temporal lobe receives visual pathways from the occipital lobe via white matter fiber projections and connects to the prefrontal cortex, which is involved in higher-order visual processing, fluency of verbal expression, and gray matter atrophy in the area of the inferotemporal gyrus is found in patients with VCI and AD (41). Taken together, these findings suggest that there may be unique neuroanatomical markers of CSVD-related CI and that bilateral temporo-occipital and thalamocortical atrophy, with particular limbic system alterations, may provide some insights into understanding the multilevel pathological process from vascular injury to cognitive decline in patients with CSVD, which warrants further exploration.

CSVD is an important cause of CI. Evidence suggests that elevated levels of CSVD are linked to a higher probability of cognitive dysfunction, mild cognitive impairment (MCI), and dementia (42,43). The importance of monitoring key factors of CSVD phenotype and severity in the assessment of cognitive function is highlighted. Notably, WMH and CMBs were independently associated with CI after adjusting for confounders, whereas lacunes and EPVS were not. This suggests that ischemic white matter damage and microhemorrhages may disproportionately contribute to gray matter atrophy and cognitive dysfunction in CSVD, possibly due to their widespread effects on neuronal connectivity and perfusion. Future studies should explore whether targeted management of these specific markers could mitigate cognitive decline and explore mechanistic pathways underlying these lesion-specific effects. MTA as well as hippocampal changes suggestive of CI in memory or amnestic VCI, and MTA scores contribute to the detection of subjective and MCI and may be a sensitive biomarker of functional situational memory deficits associated with MCI (44,45). The diagnostic efficacy of the total CSVD score is low compared to microstructural changes. Therefore, gray matter brain atrophy has an advantage in detecting CI in CSVD compared to CSVD total score, and the gray matter brain atrophy combined with the CSVD total load score model showed the best discriminatory ability, which was superior to gray matter atrophy and the CSVD load score. This suggests that gray matter brain atrophy can be used as a potential biomarker of cognitive dysfunction, contributing to CSVD disease progression monitoring and early detection and diagnosis.

This study had some limitations. First of all, the sample size was small and heterogeneous, and the distribution of samples in groups was uneven. In the future, it is recommended to establish a community population model to better avoid confounding factors and statistical effectiveness. The absence of amyloid or tau biomarkers may limit the exclusion of comorbid AD pathology in some CSVD patients. Second, we did not include follow-up diffusion MRI in the study. In order to confirm our findings, we plan to recruit more participants and conduct longer longitudinal follow-up studies to better evaluate longitudinal data to verify microstructural changes. Third, the multimodal mechanism is insufficient to explain the structural-functional coupling abnormalities for the integration of functional images, such as functional MRI resting-state network. Therefore, future research may need to address these limitations to achieve a more accurate characterization of brain microstructure.

Conclusions

The VBM technique can sensitively and quantitatively detect changes in brain tissue microstructure in CSVD patients with and without CI. Gray matter in cognitively impaired patients with CSVD atrophies with the progression of the disease, mainly in the temporo-occipital lobe and thalamus, and the associated subdomains of CI are mainly focused on attention and delayed memory. The combined model of modified CSVD score, MTA score, and quantitative brain atrophy imaging is the most discriminative of CI in patients with CSVD.

Acknowledgments

We would like to thank the medical staff for their invaluable assistance in collecting clinical data, and the radiology department staff for their efforts in acquiring imaging data.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-783/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-783/dss

Funding: This study was supported by National Natural Science Foundation of China (grant No. 82471220).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-783/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The First Affiliated Hospital of Jinzhou Medical University Medical research ethics committee approved the study (approval No. KYLL 2023150). Informed consent was obtained from all individual participants included in the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Iadecola C, Duering M, Hachinski V, Joutel A, Pendlebury ST, Schneider JA, Dichgans M. Vascular Cognitive Impairment and Dementia: JACC Scientific Expert Panel. J Am Coll Cardiol 2019;73:3326-44. [Crossref] [PubMed]
2. Cai Z, Wang C, He W, Tu H, Tang Z, Xiao M, Yan LJ. Cerebral small vessel disease and Alzheimer's disease. Clin Interv Aging 2015;10:1695-704. [Crossref] [PubMed]
3. Low A, Su L, Stefaniak JD, Mak E, Dounavi ME, Muniz-Terrera G, Ritchie K, Ritchie CW, Markus HS, O'Brien JT. Inherited risk of dementia and the progression of cerebral small vessel disease and inflammatory markers in cognitively healthy midlife adults: the PREVENT-Dementia study. Neurobiol Aging 2021;98:124-33. [Crossref] [PubMed]
4. Gorelick PB, Scuteri A, Black SE, Decarli C, Greenberg SM, Iadecola C, et al. Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the american heart association/american stroke association. Stroke 2011;42:2672-713. [Crossref] [PubMed]
5. Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, et al. Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 2013;12:822-38. [Crossref] [PubMed]
6. Duering M, Biessels GJ, Brodtmann A, Chen C, Cordonnier C, de Leeuw FE, et al. Neuroimaging standards for research into small vessel disease-advances since 2013. Lancet Neurol 2023;22:602-18. [Crossref] [PubMed]
7. Jokinen H, Koikkalainen J, Laakso HM, Melkas S, Nieminen T, Brander A, et al. Global Burden of Small Vessel Disease-Related Brain Changes on MRI Predicts Cognitive and Functional Decline. Stroke 2020;51:170-8. [Crossref] [PubMed]
8. Habes M, Pomponio R, Shou H, Doshi J, Mamourian E, Erus G, et al. The Brain Chart of Aging: Machine-learning analytics reveals links between brain aging, white matter disease, amyloid burden, and cognition in the iSTAGING consortium of 10,216 harmonized MR scans. Alzheimers Dement 2021;17:89-102. [Crossref] [PubMed]
9. Xu M, Li B, Zhong D, Cheng Y, Wu Q, Zhang S, Zhang S, Wu B, Liu M. Cerebral Small Vessel Disease Load Predicts Functional Outcome and Stroke Recurrence After Intracerebral Hemorrhage: A Median Follow-Up of 5 Years. Front Aging Neurosci 2021;13:628271. [Crossref] [PubMed]
10. Uchida Y, Nishimaki K, Soldan A, Moghekar A, Albert M, Oishi K. Acceleration of Brain Atrophy and Progression From Normal Cognition to Mild Cognitive Impairment. JAMA Netw Open 2024;7:e2441505. [Crossref] [PubMed]
11. Lu J, Li D, Li F, Zhou A, Wang F, Zuo X, Jia XF, Song H, Jia J. Montreal cognitive assessment in detecting cognitive impairment in Chinese elderly individuals: a population-based study. J Geriatr Psychiatry Neurol 2011;24:184-90. [Crossref] [PubMed]
12. Chen KL, Xu Y, Chu AQ, Ding D, Liang XN, Nasreddine ZS, Dong Q, Hong Z, Zhao QH, Guo QH. Validation of the Chinese Version of Montreal Cognitive Assessment Basic for Screening Mild Cognitive Impairment. J Am Geriatr Soc 2016;64:e285-90. [Crossref] [PubMed]
13. Amin Al Olama A, Wason JMS, Tuladhar AM, van Leijsen EMC, Koini M, Hofer E, Morris RG, Schmidt R, de Leeuw FE, Markus HS. Simple MRI score aids prediction of dementia in cerebral small vessel disease. Neurology 2020;94:e1294-302. [Crossref] [PubMed]
14. Scheltens P, Leys D, Barkhof F, Huglo D, Weinstein HC, Vermersch P, Kuiper M, Steinling M, Wolters EC, Valk J. Atrophy of medial temporal lobes on MRI in "probable" Alzheimer's disease and normal ageing: diagnostic value and neuropsychological correlates. J Neurol Neurosurg Psychiatry 1992;55:967-72. [Crossref] [PubMed]
15. Gaser C, Dahnke R, Thompson PM, Kurth F, Luders EThe Alzheimer's Disease Neuroimaging Initiative. CAT: a computational anatomy toolbox for the analysis of structural MRI data. Gigascience 2024;13:giae049. [Crossref] [PubMed]
16. Ashburner J. A fast diffeomorphic image registration algorithm. Neuroimage 2007;38:95-113. [Crossref] [PubMed]
17. Colloby SJ, O'Brien JT, Taylor JP. Patterns of cerebellar volume loss in dementia with Lewy bodies and Alzheimer׳s disease: A VBM-DARTEL study. Psychiatry Res 2014;223:187-91. [Crossref] [PubMed]
18. Abe O, Yamasue H, Kasai K, Yamada H, Aoki S, Inoue H, Takei K, Suga M, Matsuo K, Kato T, Masutani Y, Ohtomo K. Voxel-based analyses of gray/white matter volume and diffusion tensor data in major depression. Psychiatry Res 2010;181:64-70. [Crossref] [PubMed]
19. Li Y, Ying Y, Yao T, Jia X, Liang H, Tang W, Jia X, Song H, Shao X, Wang DJJ, Wang C, Cheng X, Yang Q. Decreased water exchange rate across blood-brain barrier in hereditary cerebral small vessel disease. Brain 2023;146:3079-87. [Crossref] [PubMed]
20. Boa Sorte Silva NC, Dao E, Liang Hsu C, Tam RC, Lam K, Alkeridy W, Laule C, Vavasour IM, Stein RG, Liu-Ambrose T. Myelin and Physical Activity in Older Adults With Cerebral Small Vessel Disease and Mild Cognitive Impairment. J Gerontol A Biol Sci Med Sci 2023;78:545-53. [Crossref] [PubMed]
21. Han X, Wang Y, Chen Y, Qiu Y, Gu X, Dai Y, Xu Q, Sun Y, Zhou Y. Predicting white-matter hyperintensity progression and cognitive decline in patients with cerebral small-vessel disease: a magnetic resonance-based habitat analysis. Quant Imaging Med Surg 2024;14:6621-34. [Crossref] [PubMed]
22. Hong H, Tozer DJ, Markus HS. Relationship of Perivascular Space Markers With Incident Dementia in Cerebral Small Vessel Disease. Stroke 2024;55:1032-40. [Crossref] [PubMed]
23. Fletcher E, Gavett B, Harvey D, Farias ST, Olichney J, Beckett L, DeCarli C, Mungas D. Brain volume change and cognitive trajectories in aging. Neuropsychology 2018;32:436-49. [Crossref] [PubMed]
24. Zhu W, Huang H, Yang S, Luo X, Zhu W, Xu S, Meng Q, Zuo C, Liu Y, Wang WAlzheimer’s Disease Neuroimaging Initiative. Cortical and Subcortical Grey Matter Abnormalities in White Matter Hyperintensities and Subsequent Cognitive Impairment. Neurosci Bull 2021;37:789-803. [Crossref] [PubMed]
25. da Silva PHR, de Leeuw FE, Zotin MCZ, Neto OMP, Leoni RF, Tuladhar AM. Cortical Thickness and Brain Connectivity Mediate the Relation Between White Matter Hyperintensity and Information Processing Speed in Cerebral Small Vessel Disease. Brain Topogr 2023;36:613-30. [Crossref] [PubMed]
26. Perosa V, Priester A, Ziegler G, Cardenas-Blanco A, Dobisch L, Spallazzi M, Assmann A, Maass A, Speck O, Oltmer J, Heinze HJ, Schreiber S, Düzel E. Hippocampal vascular reserve associated with cognitive performance and hippocampal volume. Brain 2020;143:622-34. [Crossref] [PubMed]
27. Liu C, Zhao L, Yang S, Luo Y, Zhu W, Zhu W, Shi L. Structural changes in the lobar regions of brain in cerebral small-vessel disease patients with and without cognitive impairment: An MRI-based study with automated brain volumetry. Eur J Radiol 2020;126:108967. [Crossref] [PubMed]
28. Hamilton OKL, Backhouse EV, Janssen E, Jochems ACC, Maher C, Ritakari TE, Stevenson AJ, Xia L, Deary IJ, Wardlaw JM. Cognitive impairment in sporadic cerebral small vessel disease: A systematic review and meta-analysis. Alzheimers Dement 2021;17:665-85. [Crossref] [PubMed]
29. Chong JR, Hilal S, Ashton NJ, Karikari TK, Reilhac A, Vrooman H, Schöll M, Zetterberg H, Blennow K, Chen CP, Lai MKP. Brain atrophy and white matter hyperintensities are independently associated with plasma neurofilament light chain in an Asian cohort of cognitively impaired patients with concomitant cerebral small vessel disease. Alzheimers Dement (Amst) 2023;15:e12396. [Crossref] [PubMed]
30. Du J, Xu Q. Neuroimaging studies on cognitive impairment due to cerebral small vessel disease. Stroke Vasc Neurol 2019;4:99-101. [Crossref] [PubMed]
31. Wang Y, Yang Y, Wang T, Nie S, Yin H, Liu J. Correlation between White Matter Hyperintensities Related Gray Matter Volume and Cognition in Cerebral Small Vessel Disease. J Stroke Cerebrovasc Dis 2020;29:105275. [Crossref] [PubMed]
32. Lyu H, Wang J, Xu J, Zheng H, Yang X, Lin S, Chen J, Zhou L, Hu Y, Guo Z. Structural and Functional Disruptions in Subcortical Vascular Mild Cognitive Impairment With and Without Depressive Symptoms. Front Aging Neurosci 2019;11:241. [Crossref] [PubMed]
33. Sun W, Huang L, Cheng Y, Qin R, Xu H, Shao P, Ma J, Yao Z, Shi L, Xu Y. Medial Temporal Atrophy Contributes to Cognitive Impairment in Cerebral Small Vessel Disease. Front Neurol 2022;13:858171. [Crossref] [PubMed]
34. Lambert C, Benjamin P, Zeestraten E, Lawrence AJ, Barrick TR, Markus HS. Longitudinal patterns of leukoaraiosis and brain atrophy in symptomatic small vessel disease. Brain 2016;139:1136-51. [Crossref] [PubMed]
35. Zhang R, Zhang L, Wei S, Wang P, Jiang X, Tang Y, Wang F. Increased Amygdala-Paracentral Lobule/Precuneus Functional Connectivity Associated With Patients With Mood Disorder and Suicidal Behavior. Front Hum Neurosci 2020;14:585664. [Crossref] [PubMed]
36. Yang AC, Vest RT, Kern F, Lee DP, Agam M, Maat CA, et al. A human brain vascular atlas reveals diverse mediators of Alzheimer's risk. Nature 2022;603:885-92. [Crossref] [PubMed]
37. Jaroudi W, Garami J, Garrido S, Hornberger M, Keri S, Moustafa AA. Factors underlying cognitive decline in old age and Alzheimer's disease: the role of the hippocampus. Rev Neurosci 2017;28:705-14. [Crossref] [PubMed]
38. Liu C, Li C, Gui L, Zhao L, Evans AC, Xie B, Zhang J, Wei L, Zhou D, Wang J, Yin X. The pattern of brain gray matter impairments in patients with subcortical vascular dementia. J Neurol Sci 2014;341:110-8. [Crossref] [PubMed]
39. Gluth S, Sommer T, Rieskamp J, Büchel C. Effective Connectivity between Hippocampus and Ventromedial Prefrontal Cortex Controls Preferential Choices from Memory. Neuron 2015;86:1078-90. [Crossref] [PubMed]
40. Homan P, Argyelan M, Fales CL, Barber AD, DeRosse P, Szeszko PR, Robinson DG, Lencz T, Malhotra AK. Striatal volume and functional connectivity correlate with weight gain in early-phase psychosis. Neuropsychopharmacology 2019;44:1948-54. [Crossref] [PubMed]
41. Scheff SW, Price DA, Schmitt FA, Scheff MA, Mufson EJ. Synaptic loss in the inferior temporal gyrus in mild cognitive impairment and Alzheimer's disease. J Alzheimers Dis 2011;24:547-57. [Crossref] [PubMed]
42. Li X, Yuan J, Qin W, Yang L, Yang S, Li Y, Hu W. Cerebral Microbleeds Are Associated with Impairments in Executive Function and Processing Speed. J Alzheimers Dis 2021;81:255-62. [Crossref] [PubMed]
43. Jansma A, de Bresser J, Schoones JW, van Heemst D, Akintola AA. Sporadic cerebral small vessel disease and cognitive decline in healthy older adults: A systematic review and meta-analysis. J Cereb Blood Flow Metab 2024;44:660-79. [Crossref] [PubMed]
44. Flak MM, Hol HR, Hernes SS, Chang L, Ernst T, Engvig A, Bjuland KJ, Madsen BO, Lindland EMS, Knapskog AB, Ulstein ID, Lona TEE, Skranes J, Løhaugen GCC. Cognitive Profiles and Atrophy Ratings on MRI in Senior Patients With Mild Cognitive Impairment. Front Aging Neurosci 2018;10:384. [Crossref] [PubMed]
45. Håkansson C, Torisson G, Londos E, Hansson O, Björkman-Burtscher IM, van Westen D. Reporting frequency of radiology findings increases after introducing visual rating scales in the primary care diagnostic work up of subjective and mild cognitive impairment. Eur Radiol 2021;31:666-73. [Crossref] [PubMed]