Greater White Matter Hyperintensities and More Severe Cognitive Dysfunction in Mild Cognitive Impairment With Motoric Cognitive Risk Syndrome

Abstract

Background

The motoric cognitive risk syndrome (MCRS) is characterized by slow gait and cognitive complaints. A motor-based approach to MCRS provides a clinical strategy for identifying individuals at high risk for dementia.

Methods

This study included 81 outpatients with mild cognitive impairment (MCI). All participants underwent clinical evaluations, including volumetric brain magnetic resonance imaging (MRI) and neuropsychological testing with the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet. White matter hyperintensities (WMH) volume was calculated using automated segmentation analysis from three-dimensional MRI images. MCRS was defined by the presence of cognitive complaints and slow gait defined as gait speed at least one standard deviation below age- and sex-specific means.

Results

A total of 31 subjects with MCRS and 50 subjects without slow gait participated in this study. The linear regression analysis revealed a significant negative relationship between WMH volume and gait speed in both the MCI group with MCRS (β = −1.010, P < 0.001) and the MCI group without slow gait (β = −0.427, P = 0.016). Both age (odds ratio [OR], 1.17; 95% confidence interval [CI], 1.02–1.34) and WMH volume (OR, 1.11; 95% CI, 1.01–1.22) were significantly associated with MCRS, even after adjusting for confounding factors. After applying Bonferroni correction (P < 0.0036), the MCRS group exhibited significantly worse performance on word list memory, word list recall, and MMSE-KC compared to the MCI group without slow gait.

Conclusion

MCRS represents a distinct and more severe clinical entity within the MCI population, characterized by greater cognitive impairment and increased WMH burden.

Graphical Abstract

INTRODUCTION

The motoric cognitive risk syndrome (MCRS) is a clinical condition characterized by the coexistence of slow gait and cognitive complaints, serving as a potential marker for individuals at high risk of developing dementia.1,2,3 MCRS has attracted considerable interest in recent years as a means of identifying older adults who may be on the pathway to cognitive decline, providing an accessible approach to early dementia risk assessment.

Mild cognitive impairment (MCI) is a recognized intermediary stage between normal aging and dementia, with a subset of individuals progressing to Alzheimer’s disease (AD) and other types of dementia.4,5 However, the risk of progression within the MCI population is highly variable, and identifying subgroups with heightened risk remains a challenge.4 MCRS may represent such a high-risk subgroup within the MCI population, with slow gait potentially indicating underlying neuropathology that affects both motor and cognitive functions.6,7,8,9 The dual presence of motor and cognitive symptoms in MCRS suggests a complex interaction between cognitive decline and motor impairment, potentially mediated by vascular and neurodegenerative processes.

There are growing evidence that white matter hyperintensities (WMHs) is associated with slow gait.10,11,12 Accumulating evidence indicates that WMH plays a crucial role in lowering the threshold for the clinical expression of dementia in AD-related pathologies.13,14 Moreover, greater WMH volume is associated with accelerated cognitive decline15 and concurrent cognitive dysfunction, in particular, executive functions.3,16,17,18,19 The core symptoms of MCRS, namely slow gait and cognitive dysfunction, are closely related to WMH. However, in the context of MCRS, the relationship between MCRS and WMH is still being investigated, with evidence both supporting3,20,21,22,23 and questioning24,25 their association. Although more recent studies tend to support this association,3,20,21,22,23 standardized studies with advanced methodologies are needed to further clarify it.

Thus, we hypothesized that WMH might contribute to the development of MCRS and could also be associated with more severe cognitive dysfunction in MCRS within the MCI population. Building on this hypothesis, the aim of this study was to investigate the clinical, cognitive, and neuroimaging characteristics of MCRS in an MCI population. Specifically, we sought to determine whether MCI patients with MCRS exhibit greater cognitive impairments, a higher WMH burden, and distinct neuropsychological profiles compared to MCI patients without slow gait.

METHODS

Subjects

Subjects with MCI above 65 years old were consecutively recruited from the dementia clinic of Jeju National University Hospital (Jeju-do, Korea; JNUH) between February 2022 and January 2023. MCI diagnosis was based on clinical criteria, including subjective cognitive complaints corroborated by an informant, objective cognitive impairment determined by neuropsychological testing, and preserved activities of daily living.4 We excluded those with history of stroke, focal neurological signs, and major psychiatric illnesses such as schizophrenia, bipolar disorder, major depressive disorder, substance abuse and dependence, and all types of dementia to exclude individuals with disease-related cognitive dysfunction. Participants were further categorized into two groups: those with MCRS and those with only MCI without slow gait. MCRS was defined according to established criteria, which include cognitive complaints and a gait speed at least one standard deviation below age- and sex-specific norm. Gait speed was measured using a timed 4-meter walk. Participants were instructed to walk at their usual pace, and the time taken to complete the 4-meter distance was recorded. Gait speed (meters/second) was calculated by dividing the distance by the time. Slow gait was defined as a gait speed one standard deviation below age- and sex-specific norms.3

Clinical and demographic assessments

All participants in this study underwent a standardized clinical evaluation, which included a clinical interview, physical and neurological examinations, and laboratory tests, including magnetic resonance imaging (MRI). This assessment was conducted in accordance with the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet (CERAD-K) Clinical Assessment Battery (CERAD-K-C).26 Cognitive function was assessed using the CERAD-K Neuropsychological Assessment Battery (CERAD-K-N),26 administered by two experienced neuropsychologists. This battery comprises ten neuropsychological tests: Verbal Fluency Test, 15-item Boston Naming Test, Korean version of the Mini-Mental State Examination (MMSE-KC), Word List Memory Test, Constructional Praxis Test, Word List Recall Test, Word List Recognition Test, Constructional Recall Test, Trail Making Test, and Stroop test. To assess frontal lobe function, the Frontal Assessment Battery (FAB)27 and Clock Drawing Task (CLOX)28 were used. Depressive symptoms were evaluated using the Korean version of the Geriatric Depression Scale (KGDS),29 and comorbidities were assessed with the Charlson Comorbidity Index (CCI).30 Following a thorough review of all clinical data, two neuropsychiatrists with expertise in dementia research determined the Clinical Dementia Rating Scale Sum of Boxes (CDR-SOB)31,32 scores.

MRI acquisition and analysis

MRI data were collected at JNUH using a 3.0 Tesla Philips Intera scanner. High-resolution three-dimensional (3D) T1-weighted anatomical images were obtained with an acquisition voxel size of 1.0 × 0.5 × 0.5 mm, a sagittal slice thickness of 1.0 mm without inter-slice gaps, a repetition time of 4.61 ms, an echo time of 8.15 ms, one excitation, a flip angle of 8°, a field of view of 240 × 240 mm, and an acquisition matrix of 175 × 256 × 256 mm across the x-, y-, and z-axes. Additionally, 3D fluid-attenuated inversion recovery (FLAIR) images were acquired with a voxel size of 1 × 1 × 3 mm³, a repetition time of 9,900 ms, an echo time of 125 ms, an inversion time of 2,800 ms, one excitation, a flip angle of 90°, a field of view of 240 mm, an axial plane matrix of 256 × 256 mm, a slice thickness of 3 mm, and no inter-slice gap. To minimize motion artifacts that could degrade image quality and diagnostic utility, participants were thoroughly instructed on proper imaging procedures and posture by a trained imaging technician. Furthermore, all scans were reviewed by an experienced radiology specialist to ensure accurate interpretation, and cases with significant analysis limitations were excluded from the study.

We used FreeSurfer software (version 7.0.0) to calculate both estimated intracranial volume (eICV) and total hippocampal volume (HV) from the acquired 3D T1-weighted MRI scans. To calculate WMH volume, we first coregistered the 3D FLAIR images to their corresponding T1-weighted images. This coregistration was achieved using an affine transformation implemented within the Statistical Parametric Mapping software package (SPM12, version 12; Wellcome Institute of Neurology, University College London, UK; http://www.fil.ion.ucl.ac.uk/spm/doc/), running within the Matlab environment (The MathWorks, Inc., Natick, MA, USA). Following coregistration, we segmented WMH regions from the 3D FLAIR images using the lesion prediction algorithm provided by the lesion segmentation toolbox (http://www.statistical-modeling.de/1st/html) designed for SPM12. The WMH volume was subsequently measured in milliliters (mL) through automated segmentation analysis of the 3D T1-weighted MRI images.

Statistical analysis

Data analysis was performed using statistical software, with significance set at P < 0.05, except where Bonferroni correction was applied. Between-group comparisons (MCI with MCRS vs. MCI without slow gait) for continuous variables, such as age, gait speed, MMSE-KC, and WMH volume, were conducted using independent t-tests. The χ² tests were used for categorical variables, such as the presence of hypertension and diabetes. Linear regression models were employed to investigate the relationship between gait speed and WMH volume, adjusting for potential confounders including age, sex, education, hippocampal volume, eICV, CCI, apolipoprotein E ε4 allele (APOE ε4) status, and grip strength. A scatter plot and partial regression plots were used to visualize and quantify these relationships. Logistic regression analysis was used to identify factors including age and WMH volume, associated with MCRS, adjusting for demographic and clinical covariates. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to determine the strength of association. Analysis of covariance (ANCOVA) was performed to compare cognitive performance between MCI group with MCRS and only MCI group controlling for age, sex, education, hippocampal volume, eICV, CCI, apolipoprotein E ε4 allele status, and grip strength. To account for multiple comparisons in 14 neuropsychological tests, a Bonferroni correction was applied. The corrected p-value threshold was set at P < 0.0036 (0.05/14) for neuropsychological assessments to reduce the likelihood of Type I error. All statistical analyses were performed using STATA version 16.1 (StataCorp., College Station, TX, USA).

Ethics statement

All participants provided informed consent, and the study was approved by the Institutional Review Board of Jeju National University Hospital, Korea (approval No. 2022-03-011).

RESULTS

Clinical characteristics of subjects

Among MCI outpatients, a total of 31 subjects with MCRS and 50 subjects without slow gait participated in this study. Subjects with MCRS were significantly older (77.7 ± 4.7 years) than those without slow gait (73.1 ± 4.4 years, t-test P < 0.001, Student’s t-test). Hypertension was present in 71.9% of the MCRS group and 59.2% of the MCI without slow gait group, though this difference was not statistically significant (P = 0.244, chi-square test). The frequency of diabetes was also similar between two groups (P = 0.750, χ² test). The mean score of MMSE-KC was significantly lower in the MCRS group (23.5 ± 2.6) compared to the MCI group without slow gait (26.4 ± 2.0, P < 0.001, Student’s t-test). Additionally, CDR-SOB scores were higher in the MCRS group (0.7 ± 0.2) than in the MCI group without slow gait (0.5 ± 0.2, P = 0.001, Student’s t-test). The KGDS scores were also significantly higher in the MCRS group (12.8 ± 2.6) compared to the MCI group without slow gait (10.4 ± 4.4, P = 0.008, Student’s t-test). WMH volume was significantly greater in the MCRS group (13.9 ± 13.8 mL) than in the MCI without slow gait (6.7 ± 4.3 mL, P = 0.001, Student’s t-test). Gait speed was significantly lower in the MCRS group (0.75 ± 0.09 m/s) compared to the MCI group without slow gait (1.14 ± 0.17 m/s, P < 0.001, Student’s t-test) (Table 1).

Table 1. Demographic and clinical characteristics of all participants.

Variables		MCI with motoric cognitive risk syndrome (n = 31)	MCI without slow gait (n = 50)	P value
Age, yr		77.7 ± 4.7	73.1 ± 4.4	< 0.001
Female, %		14 (43.8)	28 (57.1)	0.739
Education, yr		9.8 ± 5.2	9.2 ± 4.6	0.568
Hypertension, %		23 (71.9)	29 (59.2)	0.244
Diabetes, %		5 (15.6)	9 (18.4)	0.750
APOE ε4		9 (29.0)	17 (34.0)	0.642
MMSE-KC		23.5 ± 2.6	26.4 ± 2.0	< 0.001
CDR-SOB		0.7 ± 0.2	0.5 ± 0.2	0.001
KGDS		12.8 ± 2.6	10.4 ± 4.4	0.008
CCI		3.6 ± 1.0	3.4 ± 1.1	0.396
Brain volumes, mL
	eICV	1,496.0 ± 148.1	1,530.6 ± 140.3	0.295
	Hippocampus volume	6.1 ± 0.6	6.4 ± 0.7	0.090
	WMH volume	13.9 ± 13.8	6.7 ± 4.3	0.001
Physical functions
	Body mass index, kg/m2	24.5 ± 2.6	25.0 ± 3.0	0.465
	Gait speed, m/s	0.75 ± 0.09	1.14 ± 0.17	< 0.001
	Grip strength, kg	23.3 ± 7.1	25.3 ± 10.5	0.366

Values are presented mean ± standard deviation or number (%).

MCI = mild cognitive impairment, APOE ε4 = apolipoprotein E ε4 allele, MMSE-KC = Korean version of mini-mental state examination, CDR-SOB = clinical dementia rating scale sum of boxes score, KGDS = Korean version of geriatric depression scale, CCI = Charlson comorbidity index, eICV = estimated intracranial volume, WMH = white matter hyperintensities.

WMH and gait speed

The partial regression plot for WMH and gait speed was generated after adjusting for age, education, gender, KGDS, hippocampal volume, grip strength, eICV, CCI, and the presence of APOE ε4. The plot illustrates a significant negative relationship between WMH volume and gait speed in both the MCI group with MCRS (β = −1.010, P < 0.001) and the MCI without slow gait (β = −0.427, P = 0.016). Furthermore, the interaction between WMH volume and the presence of MCRS was not significant in the relationship between WMH volume and gait speed (P = 0.095) (Fig. 1).

Fig. 1. A scatter plot and partial regression plots between white matter hyperintensities volume and gait speed. (A) Negative association between WMH volume and gait speed in both groups. (B) Negative significant association (β = −1.010, P < 0.001) between WMH volume and gait speed in MCI group with MCRS after adjusting for age, education, gender, KGDS, hippocampal volume, grip strength, CCI, eICV, and the presence of APOE ε4. (C) Negative significant association (β = −0.427, P = 0.016) between WMH volume and gait speed in MCI group without slow gait after adjusting for age, education, gender, KGDS, hippocampal volume, grip strength, CCI, eICV, and the presence of APOE ε4.

MCRS = motoric cognitive risk syndrome, MCI = mild cognitive impairment, WMH = white matter hyperintensities, eICV = estimated intracranial volume, APOE ε4 = apolipoprotein E ε4 allele, CCI = Charlson comorbidity index, KGDS = Korean version of geriatric depression scale.

WMH and MCRS

Logistic regression analysis revealed that both age (OR, 1.17; 95% CI, 1.02–1.34) and WMH volume (OR, 1.11; 95% CI, 1.01–1.22) were significantly associated with MCRS, even after adjusting for education, gender, KGDS, hippocampal volume, grip strength, eICV, CCI, and the presence of APOE ε4, while other demographic, brain structure, genetic, and physical function factors were not significantly associated (Table 2).

Table 2. Multivariate logistic regression analysis of factors associated with motoric cognitive risk syndrome.

Variables	Odds ratio (95% CI)	P value
Age, yr	1.17 (1.02–1.34)	0.023
Education, yr	1.05 (0.92–1.19)	0.503
Female	0.62 (0.13–2.98)	0.549
WMH volume	1.11 (1.01–1.22)	0.038
Hippocampal volume	1.0001 (0.999–1.001)	0.850
eICV	1.00 (0.99999–1.000003)	0.278
APOE ε4	0.62 (0.17–2.26)	0.471
KGDS	1.12 (0.94–1.33)	0.210
Grip strength	0.97 (0.90–1.03)	0.327
CCI	1.47 (0.81–2.64)	0.203

CI = confidence interval, WMH = white matter hyperintensities, eICV = estimated intracranial volume, APOE ε4 = apolipoprotein E ε4 allele, KGDS = Korean version of geriatric depression scale, CCI = Charlson comorbidity index.

Cognitive dysfunction in MCRS

When comparing cognitive performance between the MCI group with MCRS and the only MCI group after adjusting for age, education, gender, KGDS, hippocampal volume, grip strength, eICV, CCI, and the presence of APOE ε4, significant group differences were observed in word list memory (P = 0.003, ANCOVA), word List Recall (P = 0.001, ANCOVA), and MMSE-KC (P < 0.001, ANCOVA), with all P values remaining below the Bonferroni-corrected threshold of 0.0036. Although the Boston Naming Test (P = 0.025, ANCOVA), trail making test A (P = 0.025, ANCOVA), CLOX 1 (P = 0.032, ANCOVA), and FAB (P = 0.049, ANCOVA) were initially significant at P < 0.05, they did not reach the adjusted significance threshold (P < 0.0036) after Bonferroni correction (Table 3).

Table 3. Comparison of neuropsychological tests between MCI group with MCRS and MCI group without slow gait.

Variables	MCI with MCRS	MCI without slow gait	P value
Categorical verbal fluency	13.7 ± 5.6	14.0 ± 4.6	0.852
Boston naming test	9.8 ± 2.4	11.6 ± 1.7	0.025
MMSE-KC	23.5 ± 2.6	26.4 ± 2.0	< 0.001
Word list memory	12.5 ± 2.8	15.8 ± 3.5	0.003
Constructional praxis	9.2 ± 1.4	9.5 ± 1.7	0.481
Word list recall	3.5 ± 1.3	5.0 ± 1.6	0.001
Word list recognition	8.5 ± 1.4	8.6 ± 1.4	0.580
Constructional recall	6.0 ± 2.7	5.9 ± 2.5	0.186
Trail making A	80.8 ± 48.5	77.9 ± 70.4	0.025
Train making B	226.1 ± 111.8	220.2 ± 103.6	0.361
Stroop word and color	28.3 ± 10.5	27.3 ± 10.0	0.369
Clock drawing task 1	9.7 ± 3.0	11.7 ± 2.8	0.032
Clock drawing task 2	13.6 ± 1.7	13.7 ± 1.4	0.898
Frontal assessment battery	10.8 ± 3.2	13.3 ± 2.9	0.049

Values are presented mean ± standard deviation.

MCI = mild cognitive impairment, MCRS = motoric cognitive risk syndrome, MMSE-KC = Korean version of mini-mental state examination.

DISCUSSION

This study provides important insights into the characteristics and underlying mechanisms of MCRS within an MCI population, particularly focusing on its relationship with WMH and cognitive performance. Our results align with previous research linking WMH to both cognitive decline and gait impairment, further strengthening the notion that WMH plays a pivotal role in the pathophysiology of MCRS.3,20,21,22,23

Participants with MCRS were older and displayed lower cognitive scores on the MMSE-KC and CDR-SOB, as well as higher depression scores on the KGDS. These findings suggest a greater degree of cognitive impairment and depressive symptoms in MCRS patients. Our findings demonstrate that the MCI group with MCRS shows significantly poorer performance in global cognition and specific cognitive domains, particularly in memory-related tasks, as evidenced by significant differences in word list memory, word list recall. These results remained significant even after applying a Bonferroni correction, underscoring the robustness of these difference. The impaired recall and preserved recognition in the MCRS group, compared to MCI group without slow gait, suggests a selective deficit in active information retrieval, consistent with the cognitive profile typically seen in WMH-associated cognitive impairment.33,34,35 While executive function tests (trail making test A, CLOX 1, and FAB) showed differences that were initially significant, their failure to survive Bonferroni correction suggests that memory impairment, rather than executive dysfunction, may be the primary cognitive feature distinguishing MCRS in MCI populations. This finding differs somewhat from previous studies,16,33,36,37,38 which may be attributed to the fact that some prior studies did not apply the stringent statistical criteria of Bonferroni correction for multiple comparisons. Further investigations in larger cohorts are warranted to fully elucidate the executive dysfunction in MCRS.

The negative correlation between WMH volume and gait speed in both groups, which persisted after controlling for multiple confounders, suggests a direct relationship between white matter integrity and motor function.10,11,12,39 The stronger association between WMH and gait speed in the MCRS group (β = −1.007) compared to the non-MCRS group (β = −0.409) suggests that WMH may have a more pronounced effect on motor function in MCRS. While the interaction between WMH volume and MCRS presence was not statistically significant in predicting gait speed, the overall trend (P = 0.095) suggests a potential interplay between these factors. This warrants further investigation in larger cohorts to fully elucidate the complex relationship between WMH, gait, and MCRS.

A central finding of our study was the significantly greater WMH volume in the MCRS group, supporting the hypothesized link between white matter pathology and motor-cognitive dysfunction.3,20,21,22,23 The logistic regression analysis identified WMH volume and age as independent predictors of MCRS, even after adjusting for multiple potential confounders. This finding is particularly important as it suggests that WMH burden contributes to MCRS risk independently of other factors, including hippocampal volume and APOE ε4 status. This finding underscores the importance of considering both age-related factors and cerebrovascular health when assessing individuals for MCRS. While age remains a non-modifiable risk factor, identifying and managing vascular risk factors, such as hypertension and diabetes, could potentially mitigate WMH progression and subsequently reduce the risk of developing MCRS.

WMH tend to be more frequently distributed in the frontal lobe compared to other brain regions.40,41 This distribution leads to ischemic vascular injury in the prefrontal cortex and ultimately results in disruptions of the frontal-subcortical circuits.37 These circuits are essential for processes such as executive control, working memory, and psychomotor coordination.33 Our findings of impaired immediate memory and free recall, accompanied by preserved recognition memory, may primarily reflect disruptions in frontal and subcortical connectivity.33,34,35 This mechanistic explanation supports the observed association between a greater WMH burden and more severe cognitive dysfunction in individuals with MCRS.

These findings have several important clinical implications. Identifying MCRS within the MCI population allows for early recognition of individuals at higher risk of progressing to dementia. This early identification opens a window of opportunity for implementing timely interventions, such as lifestyle modifications, vascular risk factor management, and cognitive rehabilitation, that could potentially delay or even prevent further cognitive decline. The independent contribution of WMH to MCRS risk suggests that neuroimaging might be valuable in risk assessment and monitoring in MCI population. Given the association between MCRS and increased WMH burden, targeted interventions aimed at reducing WMH progression could potentially mitigate the risk of further cognitive decline.

In the context of the clinical implications of WMH, MCRS, vascular cognitive impairment (VCI), and vascular Parkinsonism exhibit overlapping etiologies and clinical manifestations. VCI represents a lifelong process encompassing a broad spectrum of cognitive disorders, ranging from subtle deficits to prodromal states and fully developed dementia, all originating from cerebrovascular lesions, including WMH.42,43 Shared vascular pathology, overlapping clinical manifestations, and evidence of frontal-subcortical circuit dysfunction support the conceptualization of MCRS as a potential subtype of VCI.

Most cases of vascular Parkinsonism typically manifest as an insidious syndrome, occurring in the context of vascular risk factors and radiological evidence of small vessel disease.44 This pathology predominantly affects the basal ganglia neuronal networks, resulting in disruption of motor pathways.44 WMH are considered one of characteristic radiological features of vascular Parkinsonism.45 Our identification of WMH as a potential underlying mechanism for gait impairment in MCRS emphasizes the necessity for longitudinal studies to elucidate whether MCRS may evolve into vascular Parkinsonism as WMH severity progresses and extends to motor pathways.

There are some strengths in this study. First, instead of using visual analog scales, WMH volume was calculated using automated segmentation analysis from 3D MRI images in this study. Most previous studies have utilized the visual analogue scale for WMH measure to examine the relationship between WMH and MCRS,3,20,21,22,23,24 yet the results have been inconsistent. This inconsistency underscores the necessity for more precise and accurate measurements of WMH to further elucidate its association with MCRS. Visual analog scales for WMH measure tend to oversimplify WMH severity, which can hinder precise examination of their clinical effects. For instance, in our previous study, WMHs with the same grade of 2 on the modified Fazekas scale46 varied widely in volume, ranging from 9.0 to 66.5 mL.47 Automated segmentation analysis of WMH from 3D MRI images, as utilized in this study, offers a valuable tool for quantifying WMH burden and monitoring its impact on cognitive and motor functions over time. Second, the study’s strength is its thorough evaluation of cognitive, motor, and neuroimaging parameters within a clearly characterized MCI cohort. Third, we adopted a conservative statistical threshold by adjusting for various demographic, genetic, and brain structure factors, as well as other potential confounders, when evaluating the relationships between MCRS, cognitive function, gait speed, and WMH. Additionally, a Bonferroni correction was applied to the neuropsychological assessments to reduce the likelihood of Type I error.

Several limitations should be considered when interpreting these results. First, the cross-sectional design prevents determination of causal relationships between WMH, cognitive decline, and motor dysfunction. Second, the relatively small sample size may have limited statistical power, particularly for detecting smaller effect sizes and the study population was limited to outpatients, potentially affecting generalizability. Third, a comprehensive etiological work-up for MCI, including amyloid PET, tau PET, and CSF Aβ42, was not conducted in this study. To mitigate this limitation, we excluded individuals with disease-related cognitive dysfunction and controlled for potential confounding variables that could influence cognitive function during the comparative analyses, thereby ensuring the inclusion of a relatively homogenous group of MCI patients. Identifying and describing etiological differences among participants with MCI could provide more valuable insights into the variability within the cohort and enhance the interpretation of findings. Future research should address these limitations through longitudinal studies to establish temporal relationships and progression patterns and larger, more diverse cohorts to confirm generalizability

In conclusion, this study provides strong evidence that MCRS represents a distinct and more severe clinical entity within the MCI population, characterized by greater cognitive impairment and increased WMH burden. The independent correlation between WMH and MCRS risk points to white matter pathology as a potential underlying mechanism in development of MCRS. These findings support the utility of MCRS as a clinical construct and suggest that combined assessment of cognitive, motor, and neuroimaging measures may improve risk stratification in MCI populations. Future research should focus on longitudinal outcomes and potential interventions targeting modifiable risk factors, particularly those related to vascular health.