Exploration of the Mechanism Underlying the Association of Incident Microinfarct and Motor Deficit: A Preliminary Functional MRI Study

Abstract

Background:

Cerebral microinfarcts (CMIs) might cause measurable disruption to brain connections and are associated with cognitive decline, but the association between CMIs and motor impairment is still unclear.

Objective:

To assess the CMIs effect on motor function in vivo and explore the potential neuropathological mechanism based on graph-based network method.

Methods:

We identified 133 non-demented middle-aged and elderly participants who underwent MRI scanning, cognitive, and motor assessment. The short physical performance battery (SPPB) assessed motor function, including balance, walking speed, and chair stand. We grouped participants into 34 incident CMIs carriers and 99 non-CMIs carriers as controls, depending on diffusion-weighted imaging. Then we assessed the independent CMIs effects on motor function and explored neural mechanisms of CMIs on motor impairment via mapping of degree centrality (DC) and eigenvector centrality (EC).

Results:

CMIs carriers had worse motor function than non-carriers. Linear regression analyses showed that CMIs independently contributed to motor function. CMIs carriers had decreased EC in the precuneus, while increased DC and EC in the middle temporal gyrus and increased DC in the inferior frontal gyrus compared to controls (p < 0.05, corrected). Correlation analyses showed that EC of precuneus was related to SPPB (r = 0.25) and balance (r = 0.27); however, DC (r = –0.25) and EC (r = –0.25) of middle temporal gyrus was related with SPPB in all participants (p < 0.05, corrected).

Conclusion:

CMIs represent an independent risk factor for motor dysfunction. The relationship between CMIs and motor function may be attributed to suppression of functional hub region and compensatory activation of motor-related regions.

Keywords

Cerebral microinfarcts cerebral small vascular disease degree centrality eigenvector centrality motor function resting-state functional MRI

INTRODUCTION

Cerebral microinfarcts (CMIs) are well-described neuropathological findings, mechanically consistent with known ischemic infarctions. The reported sizes of CMIs vary between 100μm to a few millimeters with the neuropathological examination. These mini-stroke lesions are considered to be invisible lesions on conventional MRI. However, invisible lesions have a high incidence in the elderly population (about 24%), and the incidence is higher in cerebral small vascular disease (CSVD) patients [1, 2]. While previous studies demonstrated that CMIs might cause measurable disruption to brain connections and were associated with cognitive decline, the association between CMIs and motor impairment was still unclear due to the difficulty of CMIs in vivo detection [3, 4]. Few neuropathological studies reported the relationship between CMIs and motor impairment, but antemortem motor evaluation might inevitably introduce bias and hinder the investigation on this subject [5]. The feasibility of diffusion-weighted imaging (DWI) to assess acute CMIs has been established recently [6 –8], prompting some scholars to begin to list CMIs as new CSVD imaging markers [9]. Evaluation of CMIs in vivo also makes us possible to investigate the effects of CMIs on motor function [10, 11].

Mathematical models demonstrate that the number of CMIs is usually underestimated; in other words, CMIs may represent hundreds or thousands of potential lesions [1, 2]. Animal studies demonstrated that the negative effect of CMIs extends beyond the lesion core via perilesional and remote effects [11, 12]. The presence of CMIs might cause measurable disruption on the whole brain. Based on these considerations, it was more reasonable to regard the presence of CMIs as a pathological state rather than a single or independent lesion [13, 14]. In recent years, resting-state functional MRI (rsfMRI) provided the chance to explore the CMIs-related neuropathological mechanism in function connectedness [15]. In particular, graph-based brain network analysis has been demonstrated as a reliable method to describe network characteristics. The graph-based network method is data-driven and unbiased without selecting seed points or dividing brain regions; the graph-based network method is data-driven and unbiased.

The core concept of graph theory is centrality, representing the importance of nodes in the brain network. Unlike seed-based correlation analysis or independent component analysis, centrality measure regards the whole brain as a large network instead of dividing it into several sub-networks in advance. Centrality mapping is a data-driven, prior knowledge free, and objective network analysis method. There is a variety of metrics for network centrality, each emphasizing different aspects of connectivity. For example, centrality measures consist of degree centrality (DC), counting the edge number direct connected to the node, and eigenvector centrality (EC), the sum of centralities of the direct neighbors of the node. Previously, EC and DC were frequently used in combination because they reflect brain function at local and global levels, respectively [16, 17]. The combined application of eigenvector centrality mapping (ECM) and degree centrality mapping (DCM) have proven valuable in analyzing various Neurodegenerative diseases and neuropsychiatric disorders [18, 19].

This study aimed to assess the CMIs effect on motor function among middle-aged participants in vivo and explore the potential neuropathological mechanism. According to whether there were acute CMIs or not, we grouped these participants into CMIs carriers and controls. We hypothesized that CMIs were an independent contributor to motor dysfunction in the middle-aged and elderly populations. Frontoparietal circuit disconnection was the primary mechanism underlying the relationship between CMIs and motor deficits [20, 21].

MATERIALS AND METHODS

Participant recruitment, cognitive and motor function assessment

The Medical Ethics Committee approved the study of the Second Affiliated Hospital, Zhejiang University School of Medicine. We obtained written informed consent from all participants. We retrospectively reviewed middle-aged and elderly outpatients who subjectively complained of mild dizziness, headache, dysphagia, and fatigue in the Neurology Department between July 2018 and December 2020. Neurologists first performed a preliminary physical examination and inquired about medical history. Each participant was 1) muscle strength grading 5 (muscle activation against maximum resistance of examiner, full range of motion); 2) activities of daily living (ADL) function degree (eating, dressing, indoor mobility, bathing, using the toilet, and continence) fully independent and no assistance required [22]. After the outpatient screening, those participants who had written informed consent underwent head MRI scanning and behavioral assessment (motor function and cognitive function, the interval between examinations did not exceed one week.). Specifically, inclusion criteria are as follows: 1) age > 55 (middle-aged and elderly); 2) visible white matter hyperintensities on fluid attenuated inversion recovery (FLAIR, i.e., Fazekas score 1–3 in periventricular or deep white matter); 3) non-demented, no subjective complaint of cognitive decline, and Mini-Mental State Examination (MMSE) > 24; 4) normal vision and hearing. We excluded participants with: 1) MRI-DWI showed hyperintensity > 5 mm, including lacunar stroke; 2) incomplete demographics, motor assessment, and imaging data (severe head motion, artifacts during MRI scanning; 3) secondary causes of white matter lesions (e.g., demyelinating, metabolic, immunological, toxic, infectious, and other causes); 4) abnormal MRI findings (e.g., head trauma, hemorrhage, non-lacunar infarction, and other space-occupying lesions); 5) definitive peripheral neuropathy and spinal cord disease; 6) psychiatric illness (e.g., major depression); 7) use of medication known to influence cerebral function; 8) alcohol or drug abuse; 9) left-handedness. We finally included 133 participants (silent CSVD patients) who had no noticeable motor or cognitive symptoms.

We used the short physical performance battery protocol (SPPB) [23, 24]. The battery comprises three simple parts involves gait and balance. Each participant had one trial per task condition. Amongst, gait speed was measured based on 10 meters walk; balance is tested by scoring stand, semi-tandem stand, tandem stand, and repetitive chair-stand. Each set of performance measures were assigned a score of 0 to 4, and the total score of SPPB is 12. A higher SPPB score means better global motor function. The cognitive assessment included Trail Making Test Part A and B (TMT-A/TMT-B), symbol digit modalities test (SDMT) [25 –27], forward and backward digital span (DSF/DSB).

MRI data acquisition

MRI data were acquired from the 3T (Discovery 750, GE Healthcare) scanner using an 8-channel brain phased array coil [28]. We acquired axial DWI with TR = 5000 ms; TE = 86 ms; voxel size =0.94×0.94×4 mm³; b = 1000 s/mm², and gradient recalled echo-planar imaging sequence for rsfMRI with time point = 180 volumes, TR = 2000; TE = 30 ms; slice thickness = 4mm; More MRI data acquisition detail in Supplementary Material 1.

Acute cerebral microinfarct detection

Based on MRIcron (https://www.nitrc.org/projects/mricron/), acute CMIs were rated by visual inspection according to previously proposed criteria by two experienced radiologists without knowledge of clinical information [8 , 30]. CMIs was defined as: 1) small hyperintense lesions on DWI at any brain parenchymal location, with the most significant dimension≤5 mm; 2) lesions should be isointense or hypointense in apparent diffusion coefficient to exclude T2 shine-through of high T2 signal; 3) lesions did not show hypointensities on susceptibility-weighted imaging (Fig. 1). We divide participants into the 34 CMIs carriers and 99 controls according to the presence of CMIs or not.

Fig. 1

Illustrates four examples of diffusion-weighted imaging-based CMIs diagnosis from our research sample. Case 1:79-year-old male with left basal ganglia cerebral microinfarcts (CMIs), short physical performance battery (SPPB, 10); Case 2:83-year-old male with right basal ganglia CMIs (SPPB 12, left picture), after a 1-year follow-up, the lesion disappears (right picture); Case 3:70-year-old female with lobar CMIs (SPPB 9); Case 4:62-year-old male with lobar CMIs (SPPB 11).

White matter hyperintensity assessment

The white matter hyperintensity (WMH) lesion map was automatically created for each subject based on the 3D T1-weighted and FLAIR images using the Lesion Segmentation Toolbox (Lesion prediction algorithm) [31]. Lesion masks were manually edited by one experienced radiologist (MMZ). We then co-registered the 3D T1W, FLAIR, and corrected masks to the standard atlas (UNC adult brain atlas template, http://www.nitrc.org). We finally calculated each WMH volume of the individual.

MRI preprocessing

Functional data pre-processing

We pre-processed rsfMRI data using the DPARSF, based on the SPM 12 and MATLAB R2012b [32]. In brief, we discarded the first ten time points, corrected timing differences and head motion, and co-registered and normalized the rsfMRI image to standard space; we then normalized the resultant functional images and performed the linear trends and temporal filtering (0.01 < f<0.08 Hz). More MRI pre-processing detail in Supplementary Material 2.

Mapping of degree centrality and eigenvector centrality

We computed the Pearson correlations between the time series of all pairs of brain voxels to produce a whole-brain functional connectivity matrix. The procedure is constrained by the grey matter mask generated by setting a threshold of 0.3 on the mean grey matter probability map. Mapping of DC and EC was analyzed in a voxel-wise manner to quantify the brain network’s local and global functional integrity. We then smoothed these images with a Gaussian kernel of 6×6×6 mm³ full widths at half maximum to decrease spatial noise.

In detail, DCM was computed with DPARSFA by counting each voxel the number of voxels correlated to at a threshold of r≥0.25 [33, 34]. The node with higher DC represents more direct connections with the other nodes. On the other hand, ECM was calculated via the Fast ECM toolbox [17], yielding a voxel-wise measure of relevance to the functional brain network. Each voxel was weighted based on its connection with the whole brain [35, 36]. The node with higher EC is usually connected with more nodes with a higher weighting. Finally, we Z-transformed centrality metrics so that DCM and ECM could be closer to the normal distribution and comparable.

Statistical analysis

All analyses in our study were performed using SPSS (version 26.0). We described quantitative variables as the mean and standard deviation, while categorical variables are absolute and relative frequencies. We used two-sample t-tests to examine differences in demographical information, motor and cognitive performance, and WMH burden. The Chi-square test was used to examine sex and a series of vascular risk factors (hypertension, diabetes, hyperlipidemia, smoking, excessing alcohol drinking). Then, linear regression models with entering method (p < 0.05 for significance) were used to determine the association between motor function score (dependent variable, 0–12) and CMIs (yes/no), as well as a series of independent variables. Specifically, WMH burden (ml) was entered into the regression model to test whether CMIs have an independent effect on motor function because WMH is the most commonly used imaging marker of CSVD and is related to motor function [37 –39]. Other subjects included in the regression model included cognitive function (MMSE, 0–30), demographics (age, years; sex, female/male; education, years), and common vascular risk factors (hypertension, diabetes, hyperlipidemia, smoking, excessing alcohol drinking; yes/no). Further, each motor subfield score (tests of balance, gait speed, and chair stand) was tested as a dependent variable (0–4) in the regression mentioned above model. The standardized coefficients (β) with 95% confidence intervals (CI) were reported when statistical analyses were significant (p < 0.05).

The statistical analysis of the imaging is based on the DPABI [32]. We performed a two-sample T-test to explore differences between the CMIs carriers and controls in a voxel-wise manner regarding DCM and ECM. Further, we adjusted for the effects of age, gender, education, and WMH burden in subsequent analyses. Multiple comparisons correction was performed using Gaussian random field (GRF), p < 0.05 at height (i.e., minimum Z value > 2.3), p < 0.05 at the cluster level, two tails [40]. Further, Pearson correlations between the motor function (SPPB, balance, gait speed, chair stands) and functional connectivity metrics (EC and DC) were performed (p < 0.05, Bonferroni corrected) in regions with a between-group difference.

RESULTS

CMIs carriers had worse motor function than non-carriers

In this study, we identified 34 CMIs carriers (66.2±7.9 years) and 99 non-carriers (64.1±8.5 years) as controls, based on MRI-based DWI in a middle-aged and elderly population (64.6±8.3 years) with CSVD. Among 34 CMIs carriers, there were 19 lobar CMIs carriers (67.8±4.8 years) and 15 BG CMIs carriers (64.1±10.5 years), and all CMIs were supratentorial (Supplementary Material 3). Between the CMIs carriers and non-carriers, no significant differences existed in terms of demographic characteristics (age, sex, education), the incidence of hypertension (26.5% versus 39.4%), diabetes (5.9% versus. 6.1%), hyperlipidemia (5.9% versus 6.1%), smoking (20.6% versus 35.4%), and alcoholism (17.6% versus 28.3%).

Regarding cognitive function, CMIs carriers matched well with controls in MMSE and multiple cognitive domains. CMIs carriers had worse motor function (SPPB score), heel and toe (p < 0.001), and chair stand (p < 0.05) than controls (Table 1). Further, CMIs carriers had more WMH burden than controls (p < 0.05).

Table 1

Demographic characteristics and CSVD severity of the study population

Variable	Controls	CMIs carriers	T/χ²	p
Number	99	34	/	/
Age, y	64.1 (8.5)	66.2 (7.9)	–1.2	0.2
Education, y	6.5 (4.2)	6.8 (4.2)	–0.4	0.7
Sex, F/M	38/61	12/22	0.1	0.5
MMSE	25.3 (3.9)	26.5 (3.2)	–1.7	0.1
SPPB	11.3 (1.1)	10.4 (1.7)	3.8	< 0.001^*
Heal and Toe	3.8 (0.6)	3.2 (0.8)	4.2	< 0.001^*
Chair stands	3.6 (0.8)	3.2 (1.0)	2.3	0.02^*
Gait speed	4.0 (0.1)	4.0 (0.2)	0.3	0.8
Cognitive
SDMT	27.9 (11.0)	25.9 (10.2)	0.9	0.9
DSF	6.6 (1.4)	6.9 (1.4)	–0.7	0.5
DSB	3.6 (1.4)	3.6 (1.5)	–0.4	0.7
TMT-A	91.0 (37.4)	84.5 (39.3)	0.9	0.4
TMT-B	211.1 (78.4)	197.5 (87.5)	0.8	0.4
WMH (ml)	8.5 (10.6)	15.3 (15.9)	–2.8	0.03^*

Data are mean±standard deviations unless other indicated. ^*p < 0.05. CSVD, cerebral small vascular disease; CMIs, cerebral microinfarcts; MMSE, Mini-Mental State Examination; SPPB, short physical performance battery; TMT, Trail-Making Test; WMH, white matter hyperintensities; SDMT, symbol digit modalities test.

CMIs were independently contributed to motor deficits in middle-aged and elderly

The linear regression model showed that general motor function (SPPB score) was significantly associated with WMH burden (β=–0.28, p = 0.002) and presence of CMIs (β= –0.21, p = 0.01, Table 2). For sub-item, we found that balance test was related with WMH burden (β= –0.21, p = 0.02) and presence of CMIs (β= –0.27, p = 0.001); age (β= –0.25, p = 0.001, Supplementary Material 4); while chair stand test was associated with WMH burden (β= –0.23, p = 0.02, Supplementary Material 5).

Table 2

Linear regression models to determine the contributing factors to motor function

Variables	Standardized Coefficients β	T	p
(Constant)	/	9.53	< 0.001
CMIs (yes/no)	–0.21	–2.58	0.01^*
WMH burden (ml)	–0.28	–3.16	0.002
Age (y)	–0.15	–1.64	0.10
Education (y)	0.09	0.97	0.33
Gender (female/male)	0.02	0.22	0.83
MMSE	–0.06	–0.62	0.53
Hypertension (yes/no)	–0.06	–0.70	0.48
Diabetes (yes/no)	–0.003	–0.04	0.97
Hyperlipidemia (yes/no)	0.07	0.85	0.39
Smoking (yes/no)	0.12	1.07	0.28
Alcohol (yes/no)	0.03	0.29	0.77

The dependent variable is the motor function, reflected by the short physical performance battery. The independent variables include CMIs, WMH burden (ml), cognitive function, age, sex, education, hypertension, diabetes, hyperlipidemia, smoking, excessive alcohol drinking. CMIs, cerebral microinfarcts; WMH white matter hyperintensities; MMSE, Mini-Mental State Examination.

Table 3

Significant group differences in the mapping of degree centrality and eigenvector centrality between CMIs carriers and non-carriers

Regions	MNI Coordinate			Cluster voxels	Peak intensity
	X	Y	Z
DC
Right MTG	60	–12	–21	870	–4.5
Left IFG	–45	27	12	1046	–5.0
EC
Right MTG	63	–15	–18	741	–4.7
Left Precuneus	–27	–30	69	688	3.8
Adjusting for age, sex, education, and WMH burden
DC
Right MTG	60	–12	–21	686	–4.2
Left IFG	–45	27	12	761	–5.1
EC
Left Precuneus	–3	–45	63	508	3.6

Note: comparisons of DC and EC were corrected by Gaussian random field (GRF), p < 0.05 at height (i.e., minimum Z value > 2.3), p < 0.05 at the cluster level, two tails; Further, we adjusted for the effects of age, gender, education, and white matter hyperintensities, the burden in subsequent analyses. CMIs, cerebral microinfarcts; DC, degree centrality; EC, eigenvector centrality; WMH, white matter hyperintensities; MNI, Montreal Neurological Institute stereotactic space.

Brain functional deficit and compensation co-exist in CMIs carriers

CMIs carriers had increased DC at the local level in the right middle temporal gyrus (MTG) and left inferior frontal gyrus (IFG); CMIs carriers had decreased EC in the left precuneus at the global level while increased EC in the right MTG (p < 0.05, GRF corrected). Results remained primarily unchanged, adjusting for age, sex, education, and WMH burden (p < 0.05, GRF corrected, Fig. 2, Table 1). We also explored the effects of CMIs location on the brain network, and the preliminary results are located in Supplementary Material 3.

Fig. 2

Illustrates regions showing differences of degree centrality (DC) and eigenvector centrality (EC) between cerebral microinfarcts (CMIs) carriers and non-carriers (GRF corrected, p < 0.05 at height, and p < 0.05 at the cluster level, two tails). Hot and cold areas represent DC/EC decrease and increase when CMIs carriers compared to controls, respectively. A) CMIs carriers versus controls, mapping of EC (GRF corrected); B) CMIs carriers versus controls, mapping of DC (GRF corrected); C) CMIs carriers versus controls, ECM (GRF corrected, adjusting for age, sex, education, WMH burden); D) CMIs carriers versus controls, ECM (GRF corrected, adjusting for age, sex, education, WMH burden).

Degree centrality and eigenvector centrality metrics were related to motor function

Pearson correlation results showed that EC of precuneus positively related to SPPB (r = 0.25, p = 0.004) and balance performance (r = 0.27, p = 0.002), while DC (r = –0.25, p = 0.003) and EC (r = –0.25, p = 0.003) of MTG was negatively related with SPPB in all participants (Fig. 3).

Fig. 3

Correlation relationship between imaging metrics and motor function. Across groups, the Pearson correlation results showed that eigenvector centrality (EC) of precuneus positively related to short physical performance battery (SPPB, r = 0.25, p = 0.004, A) and balance performance (r = 0.27, p = 0.002, B). In contrast, DC (r = –0.25, p = 0.003, C) and EC (r = –0.25, p = 0.003, D) of MTG was related with SPPB.

DISCUSSION

We found that CMIs are an independent risk factor for motor function impairment in CSVD patients. Based on graph-based functional connectivity analysis, CMIs carriers had a global decreased connection in the functional hub region (precuneus) during compensatory increased connection in the frontotemporal lobe. CMIs might be an alarm for clinical management because they may affect motor performance by disrupting functional connections.

CMIs independently related to motor function impairment

Apart from skeletal muscle bulk, accumulating studies have demonstrated that CSVD surrogates, such as WMH, contributed to the late-life impaired gait, balance, and postural stability [41, 42]. We also demonstrated that WMH independently contributed to motor dysfunction. Notably, for the first time, our work suggested that CMIs were related to the motor function in vivo, results independent of WMH burden and vascular risk factors. Our results were primarily in line with neuropathological work documenting that CMIs were associated with the motor function before death [5]. However, their results reported that CMIs are related to gait speed rather than balance [5]. The potential interpretations of the discrepancy include the difference in sample age and the ceiling effect of the balance test in the very elderly population (mean age 88.5 years).

Presence of CMIs linking both increase and decrease of centrality connectivity to motor impairment

We found that the CMIs carriers had decreased global centrality in the precuneus. After correcting for age, sex, education, and WMH, our results remained unchanged. Based on the hub vulnerability hypothesis, brain functional hub regions are usually the first to be affected in the presence of pathological factors [43, 44]. As a core hub, the precuneus has a pivotal role in the frontoparietal network [45]. Anatomically, the precuneus and prefrontal cortex are interconnected. This circuit covers different executive control functions, including task-elicited awareness, visuospatial function, and motor imagery, contributing to the cognitive control of posture, such as holding something in mind [39, 46]. We inferred that CMIs might preferentially damage the hub regions of the frontoparietal network via perilesional and remote effects [11, 12]. As a result, the global connectedness of precuneus is disrupted, leading to cognitive control dysfunction and balance impairment. Previous work consistently showed that the executive function mediated the relationship between WMH and the impaired motor [47, 48]. Another possible explanation is that decreased functional connection in precuneus merely represents the most salient damage in the presence of CMIs; balance impairment might be attributed to the widespread invisible CMIs, which affect the corticostriatal or thalamostriatal fiber directly [49 –51].

Interestingly, we found that CMIs carriers had both regional and global function connection increases in MTG compared to controls. We thus interpreted that increased centrality in MTG may reflect compensatory effects on motion function in CMIs carriers. In other words, the MTG of the CMIs carriers generated additional activation to finish the comprehensive motion assessment. Many previous animals or functional MRI studies showed that MTG is related to motion observation, such as memory trace consolidation processes and enhancing memory consolidation of actions [52]. One fMRI study showed that MTG takes part in action-feedback monitoring, which is an essential process that aids in motor learning and distinguishes self-generated from externally generated stimuli [53]. Our assumption supported by our correlation analyses showed that DC and EC of MTG were negatively related to global motor function. This negative correlation further indicates that the increased regional and global centrality in MTG is a negative contribution. Notably, in most previous studies, MTG function is more related to language processing; but tends to occur in the language-related dominant hemisphere (i.e., the left MTG).

Another interesting finding is that CMIs carriers had increased DC in IFG relative to non-carriers. We inferred that this result might reflect a compensatory enhancement of functional connections. IFG is a region that plays a pivotal role in motor control and program [54, 55]. Previous studies have frequently reported that compensatory IFG functional connectivity increases when the task becomes more complex [56, 57]. It became understandable that CMIs carriers compensatively activated the IFG region to complete a series of integrated movements when the functional hub region (i.e., precuneus) was impaired. The correlation analysis, however, did not find the association between IFG and SPPB and its sub-item. The possible reason is that IFG is more involved in fine movement, so a more comprehensive evaluation of motor function is necessary for the future [58].

No differences in cognitive function between CMIs carriers and non-carriers

Although cognitive function was not the main objective of our study, we unexpectedly found no significant differences between CMIs carriers and non-carriers in general cognitive function and each sub-item. Two possible explanations may account for the negative result. First, motor symptoms are common after a stroke (> 5 mm); nearly 30% of stroke patients develop cognitive impairment, such as attention and executive function, within 1 year of stroke onset [59]. CMIs are a special type of infarction (< 5 mm) essential; thus, they may affect motor function prior to cognition abilities. We hypothesized that cognitive decline might occur in CMIs carriers after a period of follow-up. Another interpretation is that the current neuropsychological battery is not wide enough, such as lacking memory-related scale.

The study has some limitations. First, future longitudinal studies are still required to understand the causal relationship between CMIs and motor function; the effects of CMIs on cognitive function might also be found in longitudinal cohorts. Second, although we adopted the proposed diagnosis criterion of acute CMIs, it should be noted that the size of DWI-based lesions still depends on field strength and the parameters of the DWI sequence [60]. Our results thus need to be further verified when the scanning consensus is determined. Third, movement evaluations in our study are still insufficient, only involving the lower extremities. On the other side, our study did not quantify arthritis, falls, and frailty conditions, which may affect SPPB score, although the neurologists tried to control these physical factors on motor function at the initial screening stage. Manual dexterity and strength, and quantitative physical factors should also be included in future research. Finally, vascular risk factors are not enough in our study; previous work has shown that abnormalities of cardiac function (e.g., atrial fibrillation, ischemic heart diseases, and congestive heart failure) may be an important factor in the occurrence of CMIs [61]. Future studies need to consider more vascular risk factors.

In conclusion, our study demonstrated that CMIs are an independent risk factor for motor function impairment, especially balance. Further, CMIs are related to brain network hub region functional connectivity decrease and compensatory connectivity increase in the motor-related region. CMIs might affect motor function via damaging motor-related cognitive control or fiber tracts directly.

Footnotes

ACKNOWLEDGMENTS

We thank the patients, researchers, and clinicians involved in the database. Our work was supported by the National Natural Science Foundation of China (Grant No. 81901707 and 82001766), the Zhejiang Medicine Health Science and Technology Program (2018KY418), Zhejiang TCM Science and Technology Plan (2022ZQ057).

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Westover

, Bianchi

, Yang

, Schneider

, Greenberg

(2013) Estimating cerebral microinfarct burden from autopsy samples. Neurology 80, 1365–1369.

Auriel

, Westover

, Bianchi

, Reijmer

, Martinez-Ramirez

, Ni

, Van Etten

, Frosch

, Fotiadis

, Schwab

(2015) Estimating total cerebral microinfarct burden from diffusion-weighted imaging. Stroke 46, 2129–2135.

Louis

, Bennett

(2007) Mild Parkinsonian signs: An overview of an emerging concept. Mov Disord 22, 1681–1688.

van der Holst

, van Uden

, Tuladhar

, de Laat

, van Norden

, Norris

, van Dijk

, Rutten-Jacobs

, de Leeuw

F-E

(2016) Factors associated with 8-year mortality in older patients with cerebral small vessel disease: The Radboud University Nijmegen Diffusion Tensor and Magnetic Resonance Cohort (RUN DMC) study. JAMA Neurol 73, 402–409.

Buchman

, Yu

, Boyle

, Levine

, Nag

, Schneider

, Bennett

(2013) Microvascular brain pathology and late-life motor impairment. Neurology 80, 712–718.

Van Veluw

, Jolink

, Hendrikse

, Geerlings

, Luijten

, Biessels

, Klijn

(2014) Cortical microinfarcts on 7T MRI in patients with spontaneous intracerebral hemorrhage. J Cereb Blood Flow Metab 34, 1104–1106.

Hilal

, Sikking

, Shaik

, Chan

, Van Veluw

, Vrooman

, Cheng

C-Y

, Sabanayagam

, Cheung

, Wong

(2016) Cortical cerebral microinfarcts on 3T MRI: A novel marker of cerebrovascular disease. Neurology 87, 1583–1590.

ter Telgte

, Scherlek

, Reijmer

, van der Kouwe

, van Harten

, Duering

, Bacskai

, de Leeuw

, Frosch

, Greenberg

, van Veluw

(2020) Histopathology of diffusion-weighted imaging-positive lesions in cerebral amyloid angiopathy. Acta Neuropathol 139, 799–812.

Wardlaw

, Smith

, Dichgans

(2019) Small vessel disease: Mechanisms and clinical implications. Lancet Neurol 18, 684–696.

10.

Smith

, Schneider

, Wardlaw

, Greenberg

(2012) Cerebral microinfarcts: The invisible lesions. Lancet Neurol 11, 272–282.

11.

Auriel

, Edlow

, Reijmer

, Fotiadis

, Ramirez-Martinez

, Ni

, Reed

, Vashkevich

, Schwab

, Rosand

(2014) Microinfarct disruption of white matter structure: A longitudinal diffusion tensor analysis. Neurology 83, 182–188.

12.

Summers

, Hartmann

, Hui

, Nie

, Deardorff

, McKinnon

, Helpern

, Jensen

, Shih

(2017) Functional deficits induced by cortical microinfarcts. J Cereb Blood Flow Metab 37, 3599–3614.

13.

Venkat

, Chopp

, Zacharek

, Cui

, Zhang

, Li

, Lu

, Zhang

, Liu

, Chen

(2017) White matter damage and glymphatic dysfunction in a model of vascular dementia in rats with no prior vascular pathologies. Neurobiol Aging 50, 96–106.

14.

Wang

, Ding

, Deng

, Guo

, Wang

, Iliff

, Nedergaard

(2017) Focal solute trapping and global glymphatic pathway impairment in a murine model of multiple microinfarcts. J Neurosci 37, 2870–2877.

15.

Taubert

, Lohmann

, Margulies

, Villringer

, Ragert

(2011) Long-term effects of motor training on resting-state networks and underlying brain structure. Neuroimage 57, 1492–1498.

16.

Di Martino

, Zuo

X-N

, Kelly

, Grzadzinski

, Mennes

, Schvarcz

, Rodman

, Lord

, Castellanos

, Milham

(2013) Shared and distinct intrinsic functional network centrality in autism and attention-deficit/hyperactivity disorder. Biol Psychiatry 74, 623–632.

17.

Wink

, de Munck

, van der Werf

, van den Heuvel

, Barkhof

(2012) Fast eigenvector centrality mapping of voxel-wise connectivity in functional magnetic resonance imaging: Implementation, validation, and interpretation. Brain Connect 2, 265–274.

18.

Zhou

, Li

, Luo

, Zeng

, Jiaerken

, Wang

, Xu

, Liu

, Li

, Zhang

, Fu

, Zhao

, Huang

, Zhang

(2021) Distinct impaired patterns of intrinsic functional network centrality in patients with early- and late-onset Alzheimer’s disease. Brain Imaging Behav 15, 2661–2670.

19.

Kirshenbaum

, Chahal

, Ho

, King

, Gifuni

, Mastrovito

, Coury

, Weisenburger

, Gotlib

(2021) Correlates and predictors of the severity of suicidal ideation in adolescence: An examination of brain connectomics and psychosocial characteristics. J Child Psychol Psychiatry, doi: 10.1111/jcpp.13512.

20.

Schulz

, Koch

, Zimerman

, Wessel

, Bönstrup

, Thomalla

, Cheng

, Gerloff

, Hummel

(2015) Parietofrontal motor pathways and their association with motor function after stroke. Brain 138, 1949–1960.

21.

Morton

, Bastian

(2004) Cerebellar control of balance and locomotion. Neuroscientist 10, 247–259.

22.

Katz

, Ford

, Moskowitz

, Jackson

, Jaffe

(1963) Studies of illness in the aged. The Index of ADL: A standardized measure of biological and psychosocial function. JAMA 185, 914–919.

23.

Guralnik

, Simonsick

, Ferrucci

, Glynn

, Berkman

, Blazer

, Scherr

, Wallace

(1994) A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol 49, M85–M94.

24.

Guralnik

, Ferrucci

, Simonsick

, Salive

, Wallace

(1995) Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med 332, 556–562.

25.

Bowie

, Harvey

(2006) Administration and interpretation of the Trail Making Test. Nat Protoc 1, 2277.

26.

Nasreddine

, Phillips

, Bédirian

, Charbonneau

, Whitehead

, Collin

, Cummings

, Chertkow

(2005) The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53, 695–699.

27.

Smith

(1973) Symbol digit modalities test, Western Psychological Services, Los Angeles.

28.

Wang

, Jiaerken

, Yu

, Shen

, Luo

, Hong

, Sun

, Xu

, Zhang

, Zhou

(2020) Understanding the association between psychomotor processing speed and white matter hyperintensity: A comprehensive multi-modality MR imaging study. Hum Brain Mapp 41, 605–616.

29.

Van Veluw

, Zwanenburg

, Engelen-Lee

, Spliet

, Hendrikse

, Luijten

, Biessels

(2013) In vivo detection of cerebral cortical microinfarcts with high-resolution 7T MRI. J Cereb Blood Flow Metab 33, 322–329.

30.

van Veluw

, Shih

, Smith

, Chen

, Schneider

, Wardlaw

, Greenberg

, Biessels

(2017) Detection, risk factors, and functional consequences of cerebral microinfarcts. Lancet Neurol 16, 730–740.

31.

Schmidt

, Wink

(2017) LST: A lesion segmentation tool for SPM. Manual/Documentation for version 2, 15.

32.

Chao-Gan

, Yu-Feng

(2010) DPARSF: A MATLAB toolbox for “pipeline” data analysis of resting-state fMRI. Front Syst Neurosci 4, 13.

33.

Zuo XN

, Mennes

, Imperati

, Castellanos

, Sporns

, Milham

(2012) Network centrality in the human functional connectome. Cereb Cortex 22, 1862–1875.

34.

Buckner RL

, Talukdar

, Krienen

, Liu

, Hedden

, Andrews-Hanna

, Sperling

, Johnson

(2009) Cortical hubs revealed by intrinsic functional connectivity: Mapping, assessment of stability, and relation to Alzheimer’s disease. J Neurosci 29, 1860–1873.

35.

Lohmann G

, Horstmann

, Pleger

, Lepsien

, Goldhahn

, Schloegl

, Stumvoll

, Villringer

, Turner

(2010) Eigenvector centrality mapping for analyzing connectivity patterns in fMRI data of the human brain. PLoS One 5, e10232.

36.

García-García I

, Garolera

, Marqués-Iturria

, Horstmann

, Segura

, Pueyo

, Sender-Palacios

, Vernet-Vernet

, Villringer

, Junqué

, Margulies

, Neumann

(2015) Functional network centrality in obesity: A resting-state and task fMRI study. Psychiatry Res 233, 331–338.

37.

Ahmad

, Cerchiai

, Mancuso

, Casani

, Bronstein

(2015) Are white matter abnormalities associated with “unexplained dizziness”? J Neurol Sci 358, 428–431.

38.

Jokinen

, Kalska

, Mäntylä

, Ylikoski

, Hietanen

, Pohjasvaara

, Kaste

, Erkinjuntti

(2005) White matter hyperintensities as a predictor of neuropsychological deficits post-stroke. J Neurol Neurosurg Psychiatry 76, 1229–1233.

39.

Linortner

, Fazekas

, Schmidt

, Ropele

, Pendl

, Petrovic

, Loitfelder

, Neuper

, Enzinger

(2012) White matter hyperintensities alter functional organization of the motor system. Neurobiol Aging 33, 197.e1–9.

40.

Friston KJ

, Frackowiak

RSJ

, Mazziotta

, Evans

(1994) Assessing the significance of focal activations using their spatial extent. Hum Brain Mapp 1, 214–220.

41.

Choi

, Ren

, Phan

, Callisaya

, Ly

, Beare

, Chong

, Srikanth

(2012) Silent infarcts and cerebral microbleeds modify the associations of white matter lesions with gait and postural stability: Population-based study. Stroke 43, 1505–1510.

42.

Pinter

, Ritchie

, Doubal

, Gattringer

, Morris

, Bastin

, Hernández

MdCV

, Royle

, Corley

, Maniega

(2017) Impact of small vessel disease in the brain on gait and balance. Sci Rep 7, 41637.

43.

de Haan

, Mott

, van Straaten

, Scheltens

, Stam

(2012) Activity dependent degeneration explains hub vulnerability in Alzheimer’s disease. PLoS Comput Biol 8, e1002582.

44.

de Haan

, van der Flier

, Koene

, Smits

, Scheltens

, Stam

(2012) Disrupted modular brain dynamics reflect cognitive dysfunction in Alzheimer’s disease. Neuroimage 59, 3085–3093.

45.

Utevsky

, Smith

, Huettel

(2014) Precuneus is a functional core of the default-mode network. J Neurosci 34, 932–940.

46.

Wenderoth

, Debaere

, Sunaert

, Swinnen

(2005) The role of anterior cingulate cortex and precuneus in the coordination of motor behaviour. Eur J Neurosci 22, 235–246.

47.

Inzitari

, Pracucci

, Poggesi

, Carlucci

, Barkhof

, Chabriat

, Erkinjuntti

, Fazekas

, Ferro

, Hennerici

(2009) Changes in white matter as determinant of global functional decline in older independent outpatients: Three year follow-up of LADIS (leukoaraiosis and disability) study cohort. BMJ 339, b2477.

48.

Inzitari

, Simoni

, Pracucci

, Poggesi

, Basile

, Chabriat

, Erkinjuntti

, Fazekas

, Ferro

, Hennerici

(2007) Risk of rapid global functional decline in elderly patients with severe cerebral age-related white matter changes: The LADIS study. Arch Intern Med 167, 81–88.

49.

Cummings

(1993) Frontal-subcortical circuits and human behavior. Arch Neurol 50, 873–880.

50.

Chesselet

M-F

, Delfs

(1996) Basal ganglia and movement disorders: An update. Trends Neurosci 19, 417–422.

51.

Tekin

, Cummings

(2002) Frontal–subcortical neuronal circuits and clinical neuropsychiatry: An update. J Psychosom Res 53, 647–654.

52.

Scoville

, Milner

(1957) Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry 20, 11.

53.

van Kemenade

, Arikan

, Podranski

, Steinsträter

, Kircher

, Straube

(2019) Distinct roles for the cerebellum, angular gyrus, and middle temporal gyrus in action–feedback monitoring. Cereb Cortex 29, 1520–1531.

54.

Parsons

, Sergent

, Hodges

, Fox

(2005) The brain basis of piano performance. Neuropsychologia 43, 199–215.

55.

Liakakis

, Nickel

, Seitz

(2011) Diversity of the inferior frontal gyrus—a meta-analysis of neuroimaging studies. Behav Brain Res 225, 341–347.

56.

Herath

, Klingberg

, Young

, Amunts

, Roland

(2001) Neural correlates of dual task interference can be dissociated from those of divided attention: An fMRI study. Cereb Cortex 11, 796–805.

57.

Kemmotsu

, Villalobos

, Gaffrey

, Courchesne

, Müller

R-A

(2005) Activity and functional connectivity of inferior frontal cortex associated with response conflict. Cogn Brain Res 24, 335–342.

58.

Rao

, Harrington

, Haaland

, Bobholz

, Cox

, Binder

(1997) Distributed neural systems underlying the timing of movements. J Neurosci 17, 5528–5535.

59.

Al-Qazzaz

, Ali

, Ahmad

, Islam

, Mohamad

(2014) Cognitive impairment and memory dysfunction after a stroke diagnosis: A post-stroke memory assessment. Neuropsychiatr Dis Treat 10, 1677.

60.

Brundel

, De Bresser

, Van Dillen

, Kappelle

, Biessels

(2012) Cerebral microinfarcts: A systematic review of neuropathological studies. J Cereb Blood Flow Metab 32, 425–436.

61.

Hilal

, Chai

, Van Veluw

, Shaik

, Ikram

, Venketasubramanian

, Richards

, Biessels

, Chen

(2017) Association between subclinical cardiac biomarkers and clinically manifest cardiac diseases with cortical cerebral microinfarcts. JAMA Neurol 74, 403–410.