Link of gray matter volume to cognitive and motor function in elderly patients with mild cognitive impairment

Abstract

BACKGROUND

Mild cognitive impairment (MCI) is a transitional state between normal aging and Alzheimer's disease (AD), characterized by subtle cognitive decline. Amnestic MCI (aMCI), in particular, is a critical precursor often progressing to AD. There is growing interest in understanding the neuroanatomical correlates of aMCI, especially the role of gray matter volume (GMV) in cognitive and motor function decline. This study hypothesized that aMCI patients will exhibit reduced GMV, particularly in brain regions associated with cognition and motor control, impacting both cognitive performance and motor abilities.

AIM

To investigate the association of GMV with cognitive and motor functions in aMCI.

METHODS

In this cross-sectional study conducted from March 2022 to March 2024, 45 aMCI patients and 45 normal controls from our Department of Geratology were enrolled. Voxel-based morphometry was used to compare GMV between groups. Correlation of differential GMV with cognitive scores and gait parameters was assessed via partial correlation analysis. Linear regression was used to assess associations between whole-brain GMV and gait measures.

RESULTS

GMV of aMCI region of interest (ROI) 1 and ROI2 was negatively correlated with Activities of Daily Living (ADL) score. GMV of ROI6 was positively correlated with the total scores of Mini-Mental State Examination and Cambridge Cognitive Examination-Chinese Version (CAMCOG-C) and negatively correlated with ADL score. In the partial correlation analysis of cognitive and motor function parameters, age, gender, educational level, height, and weight were controlled, and the results showed that CAMCOG-C was negatively correlated with Dual Task of Time Up and Go Test (TUG) duration in the aMCI group. The volume of the left occipital gray matter in the aMCI group was negatively correlated with TUG. GMV of the bilateral frontal gyrus, right orbitofrontal gyrus, right occipital cleft, right supraoccipital gyrus, and left anterior central gyrus was positively correlated with walking speed.

CONCLUSION

GMV reduction in aMCI correlates with impaired cognition and motor function, emphasizing key roles for prefrontal, occipital, and central regions in gait disorders.

Key Words: Amnestic mild cognitive impairment; Gray matter volume; Voxel-based morphometry; Alzheimer's disease; Motor function

Core Tip: This study investigated the association between gray matter volume (GMV) reduction in amnestic mild cognitive impairment (aMCI) and decline in cognitive and motor functions. We utilized voxel-based morphometry to identify brain regions with significant GMV differences between aMCI patients and healthy controls. The findings highlight the potential of GMV as a biomarker for cognitive and motor dysfunction in aMCI, emphasizing the importance of the prefrontal, occipital, and central regions in gait disorders.

INTRODUCTION

Alzheimer's disease (AD) is a common age-related neurodegenerative disease. With the increasing aging of the world, AD has become an urgent global public health problem. The early clinical manifestations of AD mainly include cognitive decline and non-cognitive neuropsychiatric symptoms. With the progression of the disease, patients will gradually develop language disorders, disorientation, thinking decline, and behavioral changes, which eventually lead to a complete loss of cognitive function and daily living ability. The American AD Association estimates that by 2050, the number of people aged 65 and older with AD will surge to 13.8 million[1]. In addition, the treatment and nursing costs of AD are quite large, which may bring a huge economic burden to families and society[2]. Therefore, it is necessary to take active prevention and treatment measures for AD.

Mild cognitive impairment (MCI) is considered a pre-dementia state of AD. People with MCI have cognitive dysfunction, including episodic memory and executive dysfunction. Amnestic MCI (aMCI), characterized by episodic memory loss, is considered to be the precursor state of AD[3]. And 40% of patients with aMCI progress to AD four years later. The transition from normal cognition to MCI and then to AD is a gradual process, and early intervention may improve or even reverse cognitive function in some patients with MCI. Therefore, early recognition and diagnosis of MCI is of great importance for early intervention and prevention of further deterioration of cognitive function.

The pathogenesis of AD involves a series of interacting pathophysiological cascades. The main pathological changes include the accumulation of amyloid beta (Aβ) of 42 amino acids into the brain parenchyma to form amyloid plaques[4]. Aβ may cause neuronal and synaptic degeneration by interfering with the connection between synapses and neurons, and then lead to neuronal injury and apoptosis, especially in the entorhinal cortex and hippocampus, resulting in structural changes and loss of functional connections in the corresponding brain regions. Other pathologic mechanisms, such as excessive oxidative stress, mitochondrial dysfunction leading to neuroinflammation, and neuroplasticity changes, are also involved in the pathogenesis of AD[5]. Aggregation of Aβ and Tau proteins in relevant brain regions can be observed at the MCI stage.

Brain imaging studies have shown that the higher control centers in the brain responsible for planning, executing, and controlling gait include areas such as the executive control network (ECN), salience network, and somato-motor network (SMN)[6], as well as other brain regions responsible for attention, executive function, and visuospatial function[7]. In addition, regions related to the formation and control of motor tasks, such as the cerebellum and motor cortex, are also included. Therefore, there is an overlap between the areas that control walking and the areas that control cognitive function, which also explains that the simultaneous occurrence of cognitive and gait decline tends to be a "shared theory" mechanism[8]. A longitudinal study found that both cognitive function and gait speed degrade with age[9]. In all the scores of specific domain cognitive function and overall cognitive function, the faster the baseline walking speed, the slower the cognitive decline. However, baseline cognition was not associated with changes in gait speed, so slow gait may be associated with future cognitive decline. Abnormal gait may be an early biomarker of progression to AD or vascular dementia in patients with cognitive impairment. There is growing evidence that a slower pace occurs early in dementia and may precede a decline in cognitive function and that gait abnormalities may be more pronounced in older adults with cognitive impairment compared to older adults with normal cognition[10]. In addition, older adults who have no motor system-related diseases and whose baseline cognitive function is normal but who walk at a slower pace may be indicative of subsequent cognitive decline[11].

Current evidence suggests that gait disorders and falls are associated not only with age-related decline in muscle strength and endurance, but also with cognitive decline and poor ability to perform dual tasks[12]. Cognitive functions such as executive function and decline in executive function and attention are significantly associated with limited mobility, falls, and an increased risk of developing dementia[13]. In older adults, an abnormal gait is associated with an increased risk of falls and can negatively impact their independence in daily living[14]. Fall-related injuries, such as fractures, muscle injuries, and related complications, may be an important reason for the rapid progression of dementia and increased mortality in the elderly, and the medical costs of treating related complications may also increase[15]. These costs are expected to increase significantly in the future as the world's aging population becomes more serious. Therefore, effective prevention of falls and related complications is essential.

To sum up, the brain regions controlling gait and cognition are intersecting; cognitive and motor functions may influence each other, and the cognitive and motor functions may be changed correspondingly in the early stage of AD. Studies have been conducted to explore the relationship between MCI and AD, but research on the correlation between gray matter volume (GMV) and cognitive and motor function in patients with aMCI is still relatively limited, especially in terms of exploring how these changes affect daily living abilities and gait impairment. The present study aimed to fill this research gap by analyzing the correlation of GMV changes with cognitive and gait deficits in patients with aMCI, with a view to providing new perspectives on the early diagnosis and intervention of aMCI. Our study not only contributes to the understanding of the neurobiological basis of aMCI but may also have important implications for clinical practice, especially in the development of new diagnostic tools and intervention strategies. By identifying specific brain regions associated with cognitive and motor function decline, we may provide new targets for preventing the development of AD. Although previous studies have revealed a link between aMCI and AD, our study provides a more precise measure of GMV by employing voxel morphometry (VBM), which may reveal new regions that have been under-explored in previous studies. In addition, our study considered the interaction between cognitive and motor functions, which is a key factor in understanding the progression of aMCI.

MATERIALS AND METHODS

Basic information

Patients admitted to the outpatients in the Geriatrics Department of our hospital from March 2022 to March 2024 were included, including aMCI patients aged 60-85 years (aMCI group, n = 45) and controls with normal cognitive function (NC group, n = 45) during the same period. The study was approved by the Ethics Committee of our hospital, and all subjects understood the purpose of the study in advance and proved written informed consent.

Inclusion and exclusion criteria

Inclusion criteria: The aMCI diagnosis was revised according to the Petersen diagnostic criteria[16] as follows: (1) Subjective cognitive impairment reported by participants or their caregivers; (2) Objective cognitive impairment that does not meet the criteria for dementia in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; (3) Clinical Dementia Rating = 0.5; (4) Mini-Mental State Examination (MMSE) scores were 17-19 for illiterates, 20-22 for primary schools, and 23-26 for junior high schools and above; (5) Memory function score was 1.5 standard deviations lower than that of the NC group; and (6) Hachinski score ≤ 4 points.

Exclusion criteria: (1) Mental disorders, such as schizophrenia or depression; (2) High white matter signal of cerebral infarction, malacia, or Fazakes grade ≥ II; (3) Parkinson's disease, epilepsy, multiple system atrophy, and other diseases that may affect cognitive function; (4) Visual or hearing abnormalities, severe aphasia or paralysis, and being unable to cooperate with the evaluation; (5) Any drug that may affect cognitive function; (6) Bone and joint diseases, rheumatism, rheumatoid disease, fractures within the past two months, trauma, and other diseases affecting gait; and (7) People with metal foreign bodies, such as cochlear implants or heart stents or other contraindications related to magnetic resonance imaging (MRI) scans.

Methods

Neuropsychological test: Within one week of the MRI scan, a neuropsychological scale assessment was conducted by two neurology professionals trained in relevant scales, including MMSE, Montreal Cognitive Assessment Scale, and Cambridge Cognitive Examination-Chinese Version (CAMCOG-C; sub-items included: Memory, attention, executive, visuospatial, computation, language, thinking, and perception), and Activities of Daily Living (ADL).

Motor function assessment: All subjects wore shoes with heels no larger than 2 cm and completed the following tests on a 3-meter and 30-meter walkway: (1) Time Up and Go Test (TUG): This is a classic and simple test that measures how long it takes a person to get up from a standard chair without the use of armrests. In this experiment, all participants underwent three TUG tests to obtain the average value; (2) Dual Task of TUG (D-TUG): Based on TUG, the subject says all the numbers from 1 to 100 that contain 7 (e.g., 7, 17, and 27) while walking. In the process of walking, the total time and the correct number were calculated; (3) Berg Balance Scale (BBS): This scale measures subjects' balance ability. It contains 14 balance tests ranging from simple tasks to complex tasks (such as upright bending forward and standing on one leg). Each item is rated on a 5-point scale (0 to 4), where 0 indicates an inability to act and 4 indicates that the individual has completed the assigned task, for a total of 56 points; and (4) Evaluation of step length, step speed, step frequency, and other parameters: Intelligent energy consumption and daily activity recorder was used to collect step length, step speed, and step frequency. When using Intelligent Device for Energy Expenditure and Physical Activity (IDEEA), subjects wore the device to walk 30 meters on flat ground at a comfortable pace. To measure steady-state walking, IDEEA software was used to analyze the spatio-temporal parameters of only the middle 20 steps. The recorded data were analyzed using the IDEEA software package GaitView to calculate step length, step speed, and step frequency.

MRI data parameters: T1WI, T2WI, FLAIR, and 3D-T1 images were captured using a 3.0 T MRI scanner. 3D-T1 high resolution structural imaging parameters were: Repetition time = 9.5 m/s; echo time = 3.9 m/s; reversal time = 450 m/s; turning angle = 20°; field of view = 256 mm; matrix size = 512 × 512.

Magnetic resonance data processing: Dcm2nii software was used to convert the high-resolution T1 image into ".nii" format. Based on the MATLAB platform, the SPM8 nested VBM8 software package was used to preprocess the data, including tissue segmentation, registration, and spatial standardization. Spatial smoothing was performed using an 8 mm Gaussian kernel. The GMV after smoothing was used for statistical analysis and the results are presented. Total intracranial volume (cm³) was calculated as GMV + white matter volume + cerebrospinal fluid volume.

Statistical analysis

SPSS26.0 software was used for statistical analyses of clinical data. Measurement data with a normal distribution are represented by the mean ± SD, and an independent sample t-test was used for comparisons between groups. Measurement data with a non-normal distribution are represented by the median (quartile) [M (P25, P75)], and the independent sample Mann-Whitney U test was used for comparisons between groups. Categorical data are expressed as frequencies (percentages) and the χ² test was used for comparisons between groups. Partial correlation analysis was used to assess the correlation between GMV and clinical scale scores. P < 0.05 was considered statistically significant. The general linear model was used for VBM analysis between the two groups, and the two-sample t-test was performed for covariables. The GMV and gait parameters of the whole brain were analyzed by linear regression using SPM (GRF correction, voxel-level P < 0.001, mass-level P < 0.05, bilateral test).

RESULTS

Comparison of clinical data

A total of 90 people were included in this study, including 45 in the NC group (21 males and 24 females, with an average age of 66.15 ± 8.06 years). There were 45 cases in the aMCI group, including 23 males and 22 females, with an average age of 65.97 ± 8.45 years. The basic information of the two groups of patients is shown in Table 1. The comparison of diencephalic structural changes between the two groups revealed significant reductions in gray matter volume (GMV) in the aMCI group compared to the NC group. The specific regions of interest (ROIs) showing these reductions included: ROI1: Right hippocampus, right superior temporal gyrus, right parahippocampal gyrus, right amygdala, right entorhinal cortex, and right fusiform gyrus; ROI2: Right middle temporal gyrus; ROI3: Right inferior parietal marginal angular gyrus; ROI4: Right occipital lobe; ROI5: Bilateral orbital frontal lobe; ROI6: Left middle frontal gyrus and straight gyrus; and ROI7: Left fusiform gyrus and left parahippocampal gyrus (Table 2).

Table 1 Comparison of clinical characteristics, neuropsychology, and gait parameters between normal control group and amnestic mild cognitive impairment group.

Item	NC group (n = 45)	aMCI group (n = 45)	Z/t/χ2	P value
Gender [male (%)]	21 (46.67)	23 (51.11)	0.178	0.673
Age (years)	66.15 ± 8.06	65.97 ± 8.45	0.103	0.919
BMI index (kg/m2)	23.84 ± 2.68	23.54 ± 3.05	0.501	0.618
Smoke, n (%)	5 (11.11)	9 (20.00)	1.353	0.245
Heart disease, n (%)	8 (17.78)	6 (13.33)	0.338	0.561
Diabetes mellitus, n (%)	6 (13.33)	8 (17.78)	0.338	0.561
MMSE, [M (P25, P75)]	32 (30, 33)	25 (23, 27)	-5.124	0.000
MoCA, [M (P25, P75)]	29 (28, 31)	21 (19, 33)	-6.051	0.000
CAMCOG-C, [M (P25, P75)]	93 (90, 97)	76 (72, 81)	-6.315	0.000
GDS, [M (P25, P75)]	6 (3, 8)	8 (5, 10)	-2.012	0.046
ADL, [M (P25, P75)]	20 (19, 21)	20 (20, 21)	-1.548	0.140
TUG (s)	11.35 ± 2.35	13.94 ± 2.64	-4.916	0.000
D-TUG (s)	12.02 ± 2.64	15.06 ± 2.10	-6.054	0.000
BBS, [M (P25, P75)]	56 (54, 57)	54 (53, 56)	-2.458	0.006
Step length (m)	0.57 ± 0.04	0.50 ± 0.08	5.582	0.000
Leg speed (m/s)	1.08 ± 0.13	0.94 ± 0.18	4.249	0.000
Step frequency (steps/min)	108.15 ± 7.84	103.11 ± 12.05	2.350	0.021

aMCI: Amnestic mild cognitive impairment; MMSE: Mini-Mental State Examination; MoCA: Montreal Cognitive Assessment Scale; CAMCOG-C: Cambridge Cognitive Examination-Chinese Version; ADL: Activities of Daily Living; TUG: Time Up and Go Test; D-TUG: Dual Task of Time Up and Go Test; BBS: Berg Balance Scale.

Table 2 Gray matter volume differences between the two groups.

Area	Voxel	MNI coordinate			t value
		X	Y	Z
Right hippocampus	1154	26	-30	-7	5.05
Right parahippocampal gyrus	901	30	-3	-28	5.22
Right amygdala	312	-	-	-	-
Right entorhinal cortex	199	-	-	-	-
Right fusiform gyrus	208	-	-	-	-
Right superior temporal gyrus	256	-	-	-	-
Right middle temporal gyrus	990	31	-49	12	4.58
Right inferior parietal marginal angular gyrus	301	40	-52	39	5.11
Right occipital lobe	418	30	-84	40	4.68
Medial frontal gyrus of both orbitalis	268	6	47	-10	4.61
Left medial frontal gyrus	366	-29	40	-15	5.23
Left straight gyrus	70	-10	38	-15	3.48
Left fusiform gyrus	441	-21	-44	-12	4.40
Left parahippocampal gyrus	250	-	-	-	-

“-” indicates no value. MNI: Montreal Neurological Institute.

Partial correlation analysis of gray matter volume and cognition

Partial correlation analysis was performed to assess the correlation between GMV and cognitive function in the two groups. The results showed that the GMV of ROI1 and ROI2 in aMCI was negatively correlated with the ADL score. The GMV of ROI6 was positively correlated with the total score of MMSE and CAMCOG-C and negatively correlated with ADL score (Table 3).

Table 3 Partial correlation analysis of inter-group gray matter volume differences and cognitive function.

Indicator of cognitive function	ROI1	ROI2	ROI3	ROI4	ROI5	ROI6	ROI7
MMSE	0.175	0.033	0.254	0.042	-0.232	0.374a	-0.003
MoCA	0.221	0.026	0.075	0.051	-0.134	0.244	-0.090
CAMCOG-C	0.341	0.210	0.190	0.225	0.004	0.521a	0.110
ADL	-0.518a	-0.412a	-0.060	-0.180	-0.130	-0.366a	-0.267

^aP < 0.05.

The values in the table are partial correlation analysis coefficient. MMSE: Mini-Mental State Examination; MoCA: Montreal Cognitive Assessment Scale; CAMCOG-C: Cambridge Cognitive Examination-Chinese Version; ADL: Activities of Daily Living; ROI: Region of interest.

Partial correlation analysis of GMV and motor function in the telencephalic area

The results of partial correlation analysis controlled for age, sex, height, and weight showed that the GMV of ROI5 in the aMCI group was positively correlated with walking speed (Table 4).

Table 4 Partial correlation analysis of inter-group gray matter volume differences and gait parameters.

Gait parameter	ROI1	ROI2	ROI3	ROI4	ROI5	ROI6	ROI7
TUG	0.102	-0.051	0.205	0.239	-0.068	-0.102	0.210
D-TUG	-0.104	-0.080	-0.008	-0.003	0.011	-0.206	-0.067
BBS	0.152	0.068	-0.294	0.008	0.201	-0.004	0.131
Step length	-0.035	0.094	-0.235	0.105	0.244	0.121	0.205
Leg speed	0.275	0.116	-0.140	0.026	0.401a	0.265	0.159
Step frequency	-0.131	0.031	0.254	-0.301	0.005	-0.061	0.044

^aP < 0.05.

The values in the table are partial correlation analysis coefficient. ROI: Region of interest; TUG: Time Up and Go Test; D-TUG: Dual Task of Time Up and Go Test; BBS: Berg Balance Scale.

In the partial correlation analysis of cognition and gait parameters, age, gender, education level, height, weight, etc. were controlled, and the results showed that CAMCOG-C was negatively correlated with D-TUG duration in the aMCI group (Table 5).

Table 5 Partial correlation analysis of gait parameters and cognitive function in the amnestic mild cognitive impairment group.

Indicator of cognitive function	TUG (s)	D-TUG (s)	BBS	Step length (m)	Leg speed (m/s)	Step frequency (steps/min)
MMSE	0.205	-0.223	-0.205	-0.203	0.168	0.025
MoCA	0.154	-0.234	-0.084	-0.155	0.164	-0.135
CAMCOG-C	0.034	-0.388 a	-0.194	-0.188	0.206	-0.118
ADL	-0.222	0.138	-0.002	0.151	-0.064	0.139

^aP < 0.05.

The values in the table are partial correlation analysis coefficient. MMSE: Mini-Mental State Examination; MoCA: Montreal Cognitive Assessment Scale; CAMCOG-C: Cambridge Cognitive Examination-Chinese Version; ADL: Activities of Daily Living; ROI: Region of interest; ROI: Region of interest; TUG: Time Up and Go Test; D-TUG: Dual Task of Time Up and Go Test; BBS: Berg Balance Scale.

Linear regression analysis of gray matter volume and motor function in whole brain

SPM was used for linear regression analysis of gray matter volume and exercise level in the whole brain, and no statistical difference was found in the NC group. In the aMCI group, the volume of the left supraoccipital gray matter was negatively correlated with TUG. The GMV of the bilateral frontal gyrus, right orbital middle frontal gyrus, right tabloid cleft, right supraoccipital gyrus, and left anterior central gyrus were positively correlated with walking speed (Table 6).

Table 6 Results of linear regression between gait parameters and whole brain gray matter volume in the amnestic mild cognitive impairment group.

Item	Area	voxel	MNI coordinate			t value
			X	Y	Z
TUG (s)	Left superior occipital gyrus	140	-20	-81	41	-4.50
Leg speed (m/s)	Left frontal straight gyrus	764	-9	26	-23	4.90
			-10	23	-16	-
	Right frontal straight gyrus	221	8	30	-10	4.30
	Right orbital medial frontal gyrus	155	-	-	-	-
	Right tabloid fissure	186	16	-71	20	5.01
			18	-55	14	-
	Right supraoccipital gyrus	37	20	-76	29	4.25
	Left anterior central gyrus	251	-41	-7	50	4.61

“-” indicates no value. MNI: Montreal Neurological Institute.

DISCUSSION

The present study reveals the pattern of progression of cortical atrophy during the evolution of aMCI to AD by analyzing the GMV in patients with aMCI. Our findings are consistent with those mentioned in the literature[16], namely, atrophy occurs first in the medial temporal lobe, then spreads to the rest of the temporal lobe, and eventually affects the frontal lobe. These changes in GMV are thought to be a common intrinsic mechanism of cognitive and physical decline with age. Our study extends previous studies focusing on the bilateral hippocampus, bilateral parahippampal gyrus, and bilateral inferior parietal marginal angular gyrus in patients with aMCI[17], and with the VBM technique, we also observed changes in GMV in the right occipital lobe, bilateral prefrontal lobes, and left fusiform gyrus. Some studies have shown that white matter fiber bundles in patients with MCI and AD also have early changes. Researchers evaluated the micro-structural differences in white matter fiber density and the macro-structural differences in fiber bundle morphology and carried out an ROI analysis[18]. The results showed that AD patients showed significant white matter loss in both microstructure and macrostructure, especially in DMN-related white matter fiber pathways, which was consistent with the study of DMN as an early AD-involved region. GMV in the hippocampus, superior temporal gyrus, parahippocampal gyrus, middle temporal gyrus, and occipital lobe of aMCI patients not only showed structural atrophy but also changed in function. An analysis based on local consistency (ReHo) showed that patients with aMCI had significant changes in brain functional activity, mainly in DMN, the visual network (VN), and SMN[19]. The structural and functional changes in the hippocampus, parahippocampal gyrus, and occipital lobes contribute to understanding the mechanisms of changes in memory, executive function, and motor function in patients with MCI.

In exploring the significance of these changes, we found that the total scores of the MMSE and CAMCOG-C were strongly correlated with the changes in GMV in the aforementioned regions. The hippocampus, fusiform gyrus, angular gyrus, frontal lobe, and occipital lobe were all located in the important circuits of cognitive function in aMCI patients, involved in the formation of DMN, ECN, and VN, and were closely related to memory and ECN ability. Therefore, the reduction of GMV in these areas will lead to impaired cognitive function in aMCI patients[20], which will affect patients' daily living ability in many ways. It can be seen that the volume of the gray matter in the occipital, fusiform gyrus, and frontal lobes changes in patients with aMCI. Previous studies mostly focused on the changes in the medial temporal lobe. Structurally, the occipital lobes are part of the fronto-occipital fasciculus (a major fiber tract in the brain). This region has been proven to play a crucial role in maintaining visuospatial attention and object recognition functions[21]. Thus, atrophy of these sites in patients with aMCI may be associated with impaired visuospatial executive function.

Detailed exercise level assessment parameters were used to assess the impairment of gait in aMCI . In terms of gait analysis, our study showed that patients with aMCI had prolonged TUG and D-TUG time, shortened step length, reduced step speed, and reduced step frequency. They also performed poorly in BBS, especially prolonged TUG and D-TUG time. This is consistent with previous findings that aMCI patients perform abnormally in terms of D-TUG, stride length, and stride speed[22]. Prolonged D-TUG duration is associated with an increased risk of falls, and a slower pace is thought to be a major factor contributing to the increased risk of falls[23]. Compared with healthy elderly people, the walking speed of elderly people with MCI is significantly slower, and the postural swing is larger[24]. The lower BBS of aMCI patients indicates that their balance ability is impaired and their physical stability deteriorates. Since falls are secondary to loss of balance or postural stability, patients with MCI in the later stage may be more prone to falls with the progression of the disease[25].

Our study further revealed, through linear regression analysis of motor levels and whole-brain GMV, that GMV in the prefrontal, occipital, peripheral gray matter, and precentral gyrus may also be associated with the development of aMCI gait disturbance. In aMCI patients, GMV in the left supraoccipital gyrus is negatively correlated with TUG duration. The GMV of the bilateral prefrontal lobe, right occipital lobe, and peripheral cortex was positively correlated with walking speed. In the partial correlation analysis of differential GMV and gait, we also found that step speed was positively correlated with frontal lobe volume, and step frequency was positively correlated with occipital lobe volume, which further indicated that the frontal lobe, occipital lobe, and anterior central gyrus may be involved in the regulation of gait.

CONCLUSION

In conclusion, the present study observed a significant reduction in GMV in patients with aMCI, a change may have an important impact on cognitive function and motor performance, particularly in the prefrontal, occipital, peripheral gray matter, and precentral gyrus, regions that may play a crucial role in gait impairment among patients with aMCI. Given the potential effects of selection bias and unmeasured confounders due to the retrospective design of the present study, we recommend that future research adopt a prospective design with more extensive follow-up investigations to enhance the validity of the prediction methods and quantify additional metrics. In addition, we believe that a more comprehensive and critical analysis of the existing literature, as well as an extended discussion of potential clinical applications and study limitations, would increase the impact and practical value of this research. Therefore, we suggest that future studies replicate this study in a wider range of samples and different populations to validate the generalizability of our findings, thereby providing a stronger scientific basis for the early diagnosis and intervention of AD.