Association of magnetic resonance imaging glymphatic function with gray matter volume loss and cognitive impairment in Alzheimer’s disease: a diffusion tensor image analysis along the perivascular space (DTI-ALPS) study

Abstract

Background

Diffusion tensor image analysis along the perivascular space (DTI-ALPS) has been used for diagnosing Alzheimer’s disease (AD); however, few studies have examined the relationship between the DTI-ALPS index and cortical metrics, and the differentiation between AD severity levels remains unclear. This study aimed to explore the differences in DTI-ALPS index and cortex among AD patients with varying severities and to analyze the interactions between DTI-ALPS index, cortical metrics, and cognitive function.

Methods

A total of 19 individuals with mild cognitive impairment (MCI), 17 individuals exhibiting mild AD, 25 individuals with moderate AD, and 28 healthy controls (HC) who were matched for age, sex, and education level were recruited. All the participants underwent diffusion tensor imaging (DTI) magnetic resonance imaging (MRI), followed by the calculation of the DTI-ALPS index to assess lymphatic system function. FreeSurfer (v7.4.1) was used to calculate thickness, volume, local gyre index, and area. One-way analysis of variance (ANOVA) was performed to compare the differences among HC, MCI, mild AD, and moderate AD groups. Pearson correlation analysis was employed to investigate the connection between the DTI-ALPS index and cognitive function, along with cortical metrics.

Results

The HC, MCI, mild AD, and moderate AD groups exhibited significant differences in the DTI-ALPS index of the left hemisphere (P=0.008), whereas 13 cortical metrics revealed a statistical significance between groups (P<0.05). In the left hemisphere, the DTI-ALPS index showed a positive trend with the Montreal Cognitive Assessment (MoCA) score (r=0.397, P<0.001). Higher DTI-ALPS was also associated with an increase in 10 cortical metrics after controlling for age, sex, and education.

Conclusions

There is a significant relationship between the DTI-ALPS index, cortical metrics, and cognitive function. This result may suggest that lymphatic dysfunction indicated by the DTI-ALPS index could mirror cortical structural degeneration and cognitive decline within the pathological process of AD. DTI-ALPS can be used as an indicator of structural degeneration and decline in cognitive function in AD.

Introduction

Alzheimer’s disease (AD) is a latent neurodegenerative disease, exhibiting features such as progressive memory loss, cognitive decline, and altered behavior. AD is characterized by the pathological accumulation of misfolded proteins, such as amyloid β-protein (Aβ), in the brain parenchyma (1). Slowed cerebrospinal fluid (CSF) flow plays an important role in the pathogenesis of AD (2). CSF circulatory function in the human brain declines with age (3) and leads to Aβ clearance of brain tissue. In 2012, a study (4) verified in a mouse model that there was a fluid transport pathway similar to the peripheral lymphatic system in the central nervous system of vertebrates, which was named the glymphatic system (GS). The GS is a well-structured fluid transportation network within the human brain that primarily functions to transport waste products from the brain. The GS is the key to Aβ clearance (4).

Dynamic visualization of the whole brain’s GS fluid images can be obtained by two methods including dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and positron emission tomography (PET) (5,6). However, both DCE-MRI and PET are invasive methods, among which DCE-MRI requires multiple imaging within a few hours to a few weeks after injection of the contrast agent, and it is difficult to carry out extensive imaging. PET has limited anatomical resolution and can only provide macroscopic CSF fluid data.

In recent years, diffusion tensor image analysis along the perivascular space (DTI-ALPS) has rapidly emerged as a non-invasive technique for evaluating lymphatic function by analyzing the diffusivity in the X, Y, and Z directions of the horizontal section of the lateral ventricle body. Subsequently, a DTI-ALPS index was defined to assess the diffusion rate along the perivascular space (PVS) of the lateral white matter relative to the main white matter fibers’ direction (7). A value nearing 1 suggests limited water movement along the PVS, whereas a higher number indicates more significant water diffusion.

In various neurological diseases, such as Parkinson’s disease (8,9), multiple sclerosis (10), epilepsy (11,12), and depressive disorder (13), the DTI-ALPS index decreases. A few studies have indicated a decrease in the DTI-ALPS index among AD patients (7,14,15). There are few studies (16,17) on the relationship between DTI-ALPS index and cortical index, and there is no difference in the severity of disease in AD patients. In this study, we aimed to investigate the differences of the DTI-ALPS index and cortex in AD patients with different severities of the disease, and to analyze the interaction between DTI-ALPS index, cortical metrics, and cognitive function. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-96/rc).

Methods

Participants

Between November 2022 and December 2023, 19 patients with mild cognitive impairment (MCI), 17 patients with mild AD (AD1 group), 25 patients with moderate AD (AD2 group), and 28 gender- and age-matched healthy controls (HC) were recruited from the local community at the Memory Disorders Clinic of the Neurology Department of The Second Affiliated Hospital of Soochow University. The inclusion criteria were as follows: (I) aged 60–85 years, right-handed; (II) meeting the core diagnostic criteria for probable AD dementia jointly developed by the National Institute on Aging (NIA) and Alzheimer’s Association (AA) in 2011 (18); (III) able to read and understand the content of the research scale. The exclusion criteria were as follows: (I) other cognitive disorders such as cerebrovascular disease, brain tumor, Hachinski ischemic score ≥4 points; (II) severe hypertension or diabetes; history of stroke or head trauma; (III) contraindications for MRI; (IV) poorly educated (<6 years of formal education). The patient enrollment process is illustrated in Figure 1. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Second Affiliated Hospital of Soochow University (No. JD-LK-2021-049-01) and informed consent was provided by all individual participants.

Figure 1.

Patient enrollment flow diagram. AD, Alzheimer’s disease; MCI, mild cognitive impairment; MRI, magnetic resonance imaging.

MRI acquisition

High-resolution three-dimensional (3D) T1-weighted and diffusion-weighted imaging (DWI) MRI images were collected from a 3.0 T MR scanner (Prisma, Siemens Healthcare, Erlangen, Germany) with a 64-channel phased-array head coil. The participants were instructed to lie flat and stay still. To minimize motion artefacts and scanner noise, foam pads and earplugs were used. DTI was acquired utilizing spin-echo single-shot echo-planar pulse sequences, encompassing a total of 32 distinct diffusion gradient directions [repetition time (TR)/echo time (TE) =8,620/85 ms, flip angle (FA) =90°, slice thickness =2.3 mm, acquisition matrix =122×122, field of view (FOV) =300×300 mm², and b-value =1,000 s/mm²], receiver bandwidth =1,532 Hz/Px, number of signal averages =1, and acquisition time =49 s. T1 images were obtained using 3D magnetic preparation rapid gradient echo (3D-MPRAGE) sequence, and the parameters were as follows: TR =2,300 ms, TE =2.98 ms, inversion time (TI) =900 ms, FA =9°, slice thickness =1.1 mm, slice number =176, and FOV =256×248 mm². The total scan time was 5 minutes 21 seconds.

DTI processing and DTI-ALPS index calculation

Firstly, dcm2nii (https://people.cas.sc.edu/rorden/mricron/dcm2nii.html) was used to convert the diffuse magnetic resonance image of the original data into a four-dimensional (4D) NIfTI format file, b0 image was extracted to remove the scalp skull to obtain brain tissue mask, and then mrtrix software (https://www.mrtrix.org/) was used to perform principal component analysis (PCA) denoising, Gibbs ring artifact removal, eddy correction based on fsl, and N4 paranoid field correction based on ants. Then, the dtifit command of fsl was used to conduct tensor reconstruction of DWI image, obtain a fractional anisotropy (FA) diagram and diffusion coefficients Dxx, Dyy, and Dzz in x, y, and z directions, respectively. Finally, fsleyes was used to draw a 3 mm radius pellet region of interest (ROI) on the projection fiber and the joint fiber, respectively (Figure 2). The mean values of Dxx, Dyy, and Dzz were extracted from within the ROI. According to the mean values of Dxx, Dyy, and Dzz extracted from the two ROIs, DTI-ALPS indexes were calculated using the following formula:

Figure 2.

Four spherical ROIs were placed within the projection fibers and association fibers on a color-coded FA map in bilateral hemispheres. FA, fractional anisotropy; ROIs, regions of interest.

$$ALPS\text{\hspace{0.05em}\hspace{0.17em}index}=\frac{mean\left(Dxxproj,Dxxassoc\right)}{mean\left(Dyyproj,Dzzassoc\right)}$$ [1]

Structural MRI processing and volumetric analysis

The image format was converted from the original DICOM format to NIfTI format, and the image quality was checked. FreeSurfer (v7.4.1; https://surfer.nmr.mgh.harvard.edu/) was used to analyze the structural image data of the included cases. Cortical structures are automatically differentiated by the software. The raw data of each case undergoes preprocessing steps including motion image reconstruction, brain tissue correction, conversion of T1 images to the Talairach template for obtaining precise coordinates, image intensity correction, 3D brain image reconstruction, cortical surface expansion, and spherical brain image set mapping. The preprocessing takes approximately 20 hours. Following preprocessing, formal brain imaging data analysis is conducted. Finally, it calculated four indexes, including thickness, local gyre index (LGI), area, and volume, according to FreeSurfer preprocessed data. The calculated thickness, volume, LGI, and area were smoothed and the smooth core was 10 mm.

Quantitative susceptibility mapping (QSM) measurement of ALPS ROI in DTI space

The ants toolkit was used to register the ALPS ROI of the DTI space linearly to the QSM individual space with the help of images. We used the STISuite_V3.0 toolbox (https://people.eecs.berkeley.edu/~chunlei.liu/software.html) to calculate the QSM metric.

Statistical analysis

SPSS software (version 26; IBM Corp., Armonk, NY, USA) was used for statistical analysis. Categorical variables were expressed in terms of counts and analyzed through Chi-squared tests. Normally distributed measures were expressed as mean ± standard deviation; non-normally distributed data were expressed as median (P25, P75). If there were missing data, we would delete some samples. FreeSurfer (v7.4.1) came with covariates including gender, age, and years of education. Permutation-based Monte Carlo simulation cluster-level correction (P<0.05) was applied at the vertex level, where P<0.001. The Desikan-Killiany (DK) atlas was used for the mapping. One-way analysis of variance (ANOVA) was conducted to assess the overall differences between the HC, MCI, AD1, and AD2 groups. Subsequently, post-hoc tests with Bonferroni correction were implemented to identify specific pairs of groups that exhibited significant differences once an overall difference was established. Statistical analysis was conducted on a total of four cortical metrics for the HC, MCI, AD1, and AD2 groups to investigate the relationship between the DTI-ALPS index and cognitive function, as well as cortical metrics, while considering the impact of age, gender, and education on the DTI-ALPS index, Pearson correlation analysis was conducted after adjusting for age, gender, and education. One-way ANOVA was used to assess QSM differences of ALPS ROI in DTI space between HC, MCI, AD1, and AD2 groups, controlling for the age, gender, and years of education variables.

Results

Demographic and clinical characteristics

The detailed demographic and clinical characteristics of the HC, MCI, AD1, and AD2 participants are displayed in Table 1. There were no significant differences among the HC, MCI, AD1, and AD2 groups in age and gender.

Table 1. Demographic data and cognition among the HC, MCI, AD1, and AD2 groups.

Parameter/characteristic	HC	MCI	AD1	AD2	P value
Number of participants	28	19	17	25	–
Age (years)	66.6±8.2	70.3±7.6	70.9±10.5	70.7±9.8	0.287
Gender (male/female), n	14/14	12/7	11/6	20/5	0.163
Education (years)	10 [2, 20]	9 [0, 16]	11 [0, 16]	8 [0, 17]	0.026*
MMSE total score	28.4±1.2	26.8±1.6	23.2±1.7	16.2±3.6	<0.001***
Temporal orientation	5 [4, 5]	4 [2, 5]	3 [1, 5]	2 [0, 5]	<0.001***
Place orientation	5 [4, 5]	5 [4, 5]	4 [3, 5]	3 [1, 5]	<0.001***
Immediate memory	3 [3, 3]	3 [2, 3]	3 [2, 3]	3 [1, 3]	<0.001***
Attention and calculation	5 [2, 5]	5 [3, 5]	5 [3, 5]	2 [0, 5]	<0.001***
Delayed recall	3 [1, 3]	2 [0, 3]	0 [0, 3]	0 [0, 3]	<0.001***
Naming	2 [1, 2]	2 [2, 2]	2 [2, 2]	2 [0, 2]	0.415
Retelling	1 [0, 1]	1 [0, 1]	1 [0, 1]	1 [0, 1]	0.177
Executive	3 [2, 3]	3 [0, 3]	3 [0, 3]	2 [0, 3]	0.001**
Reading	1 [1, 1]	1 [1, 1]	1 [0, 1]	1 [0, 1]	0.005**
Writing	1 [0, 1]	1 [0, 1]	1 [0, 1]	0 [0, 1]	<0.001***
Drawing	1 [0, 1]	1 [0, 1]	1 [0, 1]	0 [0, 1]	0.001**
MoCA total score	27.2±1.2	21.9±2.7	18.6±3.0	12.0±4.0	<0.001***
Visuospatial/executive	5 [2, 5]	4 [1, 5]	3 [0, 5]	2 [0, 4]	<0.001***
Naming	3 [2, 3]	3 [1, 3]	2 [1, 3]	1 [0, 3]	<0.001***
Attention	6 [4, 6]	6 [3, 6]	6 [4, 6]	5 [0, 6]	<0.001***
Language	2 [1, 3]	2 [0, 3]	1 [0, 3]	0 [0, 3]	<0.001***
Abstraction	2 [0, 2]	2 [0, 2]	1 [0, 2]	0 [0, 2]	<0.001***
Delayed recall	4 [2, 5]	0 [0, 4]	0 [0, 3]	0 [0, 3]	<0.001***
Orientation	6 [5, 6]	6 [3, 6]	4 [2, 6]	2 [0, 6]	<0.001***

Values are reported as mean ± SD or median [interquartile range] for the quantitative variables. *, P<0.05; **, P<0.01; ***, P<0.001. AD1, mild Alzheimer’s disease; AD2, moderate Alzheimer’s disease; HC, healthy control; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; SD, standard deviation.

DTI-ALPS index

The significant differences were observed among the HC, MCI, AD1, and AD2 groups in the DTI-ALPS index of the left hemisphere, as depicted in Figure 3 (P=0.008), whereas no significant difference was found in the right hemisphere (P=0.612). Patients with moderate AD showed significantly decreased DTI-ALPS index compared to the HC (P=0.003). Patients with moderate AD had a lower DTI-ALPS index compared to the MCI (P=0.004). The DTI-ALPS index was also significantly decreased in patients with moderate AD compared to patients with mild AD (P=0.023). There was a negative correlation between age and DTI-ALPS index (r=−0.432, P<0.001).

Figure 3.

Differences in the DTI-ALPS index of the left hemisphere among the HC, MCI, AD1, and AD2 groups. *, P<0.05; **, P<0.01. AD1, mild Alzheimer’s disease; AD2, moderate Alzheimer’s disease; ALPS, analysis along the perivascular space; DTI, diffusion tensor image; HC, healthy control; MCI, mild cognitive impairment.

Cortical metrics

There were 13 cortical metrics that displayed a statistical significance between groups (P<0.05). The clusters of brain regions with significant differences among the 10 cortical index groups are shown in Figure 4. Figure 5 and Table 2 depict the significant inter-group differences in 13 cortical metrics.

Figure 4.

Illustration of brain region clusters at which 13 cortical metrics revealed a statistical significance among the HC, MCI, AD1, and AD2 groups, including: (A-D) brain region clusters with significant inter-group differences in cortical thickness; (E) brain region cluster with significant inter-group differences in cortical LGI; (F) brain region cluster with significant inter-group differences in cortical area; (G-M) brain region clusters with significant inter-group differences in cortical volume. The color bar represents −log10CWP; with red and yellow indicating HC group > MCI, AD1, and AD2 groups. AD1, mild Alzheimer’s disease; AD2, moderate Alzheimer’s disease; CWP, composite-weighted P value; HC, healthy control; IPL, inferior parietal lobule; IT, inferior temporal; LG, lingual gyrus; LGI, local gyre index; lOFC, lateral orbitofrontal cortex; MCI, mild cognitive impairment; MT, middle temporal; ST, superior temporal.

Figure 5.

Differences in 13 cortical metrics between HC, MCI, AD1, and AD2 groups. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. AD1, mild Alzheimer’s disease; AD2, moderate Alzheimer’s disease; HC, healthy control; LGI, local gyre index; lh, left hemisphere; MCI, mild cognitive impairment; rh, right hemisphere.

Table 2. Comparisons in 13 cortical metrics between HC, MCI, AD1, and AD2 groups.

Cortical metrics	Left/right hemisphere	Brain region	Group	P value	Mean difference
Thickness	Left	Precuneus	HC vs. AD1	0.014*	0.213
			HC vs. AD2	<0.001***	0.268
			MCI vs. AD1	0.003**	0.256
			MCI vs. AD2	<0.0001****	0.311
Thickness	Right	Middle temporal	HC vs. AD1	0.034*	0.169
			HC vs. AD2	<0.0001****	0.296
			MCI vs. AD2	<0.0001****	0.281
Thickness	Right	Inferior parietal	HC vs. AD1	0.023*	0.172
			HC vs. AD2	<0.0001****	0.327
			MCI vs. AD2	<0.0001****	0.287
			AD1 vs. AD2	0.046*	0.155
Thickness	Right	Superior temporal	HC vs. AD2	<0.0001****	0.431
			MCI vs. AD2	<0.001***	0.340
			AD1 vs. AD2	0.002**	0.324
LGI	Right	Precentral	HC vs. AD2	<0.001***	0.297
			MCI vs. AD2	<0.0001****	0.347
Area	Left	Lateral orbitofrontal	HC vs. AD2	<0.0001****	0.124
			MCI vs. AD2	0.001**	0.095
			AD1 vs. AD2	<0.001***	0.115
Volume	Left	Precuneus	HC vs. AD1	0.009**	0.206
			HC vs. AD2	<0.001***	0.255
			MCI vs. AD1	<0.001***	0.266
			MCI vs. AD2	<0.0001****	0.316
Volume	Left	Lingual	HC vs. AD1	0.003**	0.316
			HC vs. AD2	<0.0001****	0.459
			MCI vs. AD2	0.030*	0.242
Volume	Left	Inferior temporal	HC vs. MCI	0.009**	0.444
			HC vs. AD1	0.007**	0.472
			HC vs. AD2	<0.0001****	0.728
Volume	Left	Inferior parietal	HC vs. AD1	0.006**	0.375
			HC vs. AD2	<0.0001****	0.510
			MCI vs. AD2	0.002**	0.389
Volume	Right	Middle temporal [1]	HC vs. MCI	0.004**	0.377
			HC vs. AD1	<0.0001****	0.569
			HC vs. AD2	<0.0001****	0.733
			MCI vs. AD2	0.007**	0.356
Volume	Right	Middle temporal [2]	HC vs. AD1	0.01*	0.377
			HC vs. AD2	<0.0001****	0.604
			MCI vs. AD2	<0.001***	0.485
Volume	Right	Fusiform	HC vs. AD2	<0.001***	0.412
			MCI vs. AD2	0.001**	0.355

*, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. AD1, mild Alzheimer’s disease; AD2, moderate Alzheimer’s disease; HC, healthy control; LGI, local gyre index; MCI, mild cognitive impairment.

Association among DTI-ALPS index, cortical metrics, and cognitive function

In the left hemisphere, the DTI-ALPS index showed a positive trend with the Montreal Cognitive Assessment (MoCA) score and the Mini-Mental State Examination (MMSE) score. Furthermore, the DTI-ALPS index was significantly correlated with performance in cognitive domains, including temporal orientation, spatial orientation, immediate memory, delayed recall, language, attention and calculation abilities, visuospatial and executive functions, reading, drawing, and abstraction, as illustrated in Figure 6.

Figure 6.

Correlation between the DTI-ALPS index in left hemispheres, cortical metrics, and cognitive function. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. ALPS, analysis along the perivascular space; DTI, diffusion tensor image; LGI, local gyre index; lh, left hemisphere; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; rh, right hemisphere.

To investigate the interaction between DTI-ALPS index, cortical metrics, and cognitive function, the correlations between the cortical metrics and the DTI-ALPS index were calculated. Among the 13 cortical metrics with significant differences (including cortical thickness, surface area, volume and gyrification), there were 10 cortical metrics with significant correlation with DTI-ALPS index, as shown in Figure 6. Higher DTI-ALPS was also associated with elevated 10 cortical metrics after adjusting for variables such as age, sex, and education.

Comparison of magnetic susceptibility values of ROIs

There was no significant difference in QSM between the four groups (P>0.05), as shown in Table 3.

Table 3. Comparison of magnetic susceptibility values of ROIs.

Group	Statistics	P value
		Left association fibers ROI	Right association fibers ROI	Left projection fibers ROI	Right projection fibers ROI
HC vs. MCI	T-test	0.545	0.457	0.653	0.998
HC vs. AD1		0.106	0.093	0.661	0.258
HC vs. AD2		0.799	0.898	0.822	0.371
MCI vs. AD1		0.157	0.260	0.806	0.556
MCI vs. AD2		0.519	0.715	0.884	0.492
AD1 vs. AD2		0.138	0.337	0.936	0.910
All	AVONA	0.295	0.609	0.917	0.699

AD1, mild Alzheimer’s disease; AD2, moderate Alzheimer’s disease; ANOVA, analysis of variance; HC, healthy control; MCI, mild cognitive impairment; ROI, region of interest.

Discussion

In this study, the DTI-ALPS index showed a significant correlation with the cortical index and a positive correlation with MoCA and MMSE scores. Meanwhile, patients with more severe AD had lower cortical metrics. We observed a correlation between glymphatic dysfunction and impaired performance across various cognitive domains. We demonstrated that there was a significant relationship among the DTI-ALPS index, cortical metrics, and cognitive function. These findings suggest that ALPS index alterations may reflect cortical degeneration and cognitive decline in AD progression (19). Meanwhile, the severity and distribution of cortical changes may progress with the decline of clinical cognitive ability.

The DTI-ALPS index is utilized to evaluate the diffusivity along the PVS direction, in comparison to the diffusivity of projection fibers and association fibers on axial slices at the level of the lateral ventricles’ body (20). This serves as an indication of the state of the lymphatic system’s function, where a lower DTI-ALPS index suggests poorer lymphatic function. In our study, significant inter-group differences in the DTI-ALPS index were observed between AD patients with varying degrees of severity and the normal control group, reflecting that the progression of AD may lead to lymphatic damage. When comparing the moderate AD group with other groups, the DTI-ALPS index in the left hemisphere decreased significantly. Therefore, our result is consistent with previous research (21), indicating that the DTI-ALPS index is less sensitive in detecting early-stage AD due to the relatively focal, rather than global, deposition of amyloid and tau proteins in the early stages of AD. A recent study (22) showed that aging leads to a decline in glymphatic function. Our results are consistent with this observation; age was negatively correlated with the DTI-ALPS index. We also observed the differences in the DTI-ALPS index between the left and right hemispheres of AD patients. The changes in the DTI-ALPS index in the left hemisphere were more obvious, whereas there was no significant difference in the DTI-ALPS index between groups in the right hemisphere. This may indicate that lymphatic dysfunction originates in the left brain and propagates to the right as AD progresses. Previous research had also shown that metabolic dysfunction in the left hemisphere of AD patients is more pronounced than that in the right hemisphere (23).

This is a study on cortical differences between patients with varying severities of AD and a HC group. In our study, 13 cortical metrics exhibited significant intergroup differences, involving the temporal lobe, frontal lobe, parietal lobe, and occipital lobe, indicating that AD-related changes are infiltrating the entire cortex. We also observed that most of the 13 cortical metrics decreased with the increase of severity of AD patients, which indicated the correlation between the brain atrophy pattern and the decline of cognitive ability in AD.

In this study, the most frequently selected important cortical metrics in the classification task of the HC, MCI, AD1, and AD2 groups were extracted from the temporal gyrus, whereas the least selected were extracted from the occipital lobe. This is consistent with many MRI studies published in recent years (24,25), which showed that brain atrophy in patients with AD is mainly in the temporal lobe, whereas the occipital lobe and the sensorimotor cortex are relatively less affected in AD. It is worth noting that the cortical volume of the left inferior temporal gyrus and right middle temporal gyrus significantly decreased in the MCI, and the statistical significance of the difference between the cortical area of the left lateral orbitofrontal gyrus and the cortical thickness of the right superior temporal gyrus was only observed in moderate AD. Furthermore, the mean thickness and volume of the precuneus in the MCI group and the LGI of the precentral gyrus were not lower than those in the HC group, and the area of the orbital frontal gyrus in the AD1 group was not less than that in the MCI group. Several research reports (26-31) have indicated that the atrophy of the inferior temporal gyrus and middle temporal gyrus begins to become evident in MCI. Previous findings (32-35) have also suggested that the lateral orbitofrontal gyrus and the superior temporal gyrus are relatively spared in the early stages of AD but become increasingly affected as the disease progresses. Previous voxel-based morphometry (VBM) studies (32,36,37) have indicated that atrophy in the superior parietal lobule and precentral gyrus is primarily associated with the transition from MCI to AD, rather than with the transition from healthy aging to MCI. The speed of brain atrophy and the specific regions affected can vary among individuals and at different stages of the disease (38).

Previous studies (7,17) have shown a clear and positive correlation between the DTI-ALPS index and MMSE in AD, and our study obtained consistent results using MoCA and MMSE scores. We hypothesized that the decrease in the DTI-ALPS index may lead to cognitive impairment. Among the cortical metrics, 10 cortical metrics were significantly correlated with the DTI-ALPS index. This suggested that the DTI-ALPS index may be correlated with the degree of cortical degeneration, that lymphatic injury could be one of the causes of structural damage in AD, and that cortical atrophy may affect the diffusion of water in brain tissue, thereby influencing the numerical value of the DTI-ALPS index (39). In our research, we conducted various assessments across multiple cognitive domains on patients with AD. We observed the correlation between lymphatic dysfunction and impaired cognitive domains, including temporal orientation, spatial orientation, immediate memory, delayed memory, language, attention and calculation abilities, visuospatial and executive functions, writing, reading, drawing, and abstraction. The cognitive declines have been linked to changes in cerebral gray matter (GM) volume and white matter integrity (40). Furthermore, these 10 cortical metrics also have significant positive correlations with these cognitive domains. Among these 10 cortical metrics, the inferior parietal lobule is associated with spatial cognition (23) and text-reading proficiency (41). The inferior temporal gyrus and middle temporal gyrus are related to memory, whereas the temporal lobe is involved in language processing (42,43). The middle temporal gyrus also aids in processing multimodal information related to abstract concepts (44). The precuneus has been associated with visuospatial functioning (45). The precuneus is also involved in the recursive process of combining complex word combinations into more intricate semantic structures (46). The lingual gyrus is a brain structure closely related to visual processing (47). When individuals engage in the recognition and processing of word shapes, the fusiform gyrus is activated. The fusiform gyrus also participates in the processing of color information (48). The precentral gyrus plays a significant role in attention control and execution (49). In this study, the GM indexes correlated with the DTI-ALPS index reflect vulnerable areas during the disease process. The result indicated that the impact of lymphatic function on cognition may be mediated through its protective role in maintaining GM integrity and execution (50).

Our study has limitations. Firstly, the sample size for subgroup analysis in this study was relatively small, which necessitates further validation in future studies with larger sample sizes. Secondly, as a cross-sectional survey, our study did not include longitudinal data on changes over time in the DTI-ALPS index, the cortical index, and cognitive function, which would help to further determine the diagnostic value of the DTI-ALPS index. Thirdly, the capacity of the DTI-ALPS index to assess human lymphatic function has yet to undergo comprehensive and rigorous pathophysiological validation to confirm its effectiveness. The DTI-ALPS index assesses overall lymphatic activity and cannot reflect regional lymphatic dysfunction. We should interpret the relationship between the DTI-ALPS index and cortical metrics with caution.

Conclusions

We demonstrated that there was a significant relationship between the DTI-ALPS index, cortical metrics, and cognitive function. This result may suggest that lymphatic dysfunction indicated by the DTI-ALPS index could mirror cortical structural degeneration and cognitive decline within the pathological process of AD. DTI-ALPS can be used as an indicator of structural degeneration and decline in cognitive function in AD.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Second Affiliated Hospital of Soochow University (No. JD-LK-2021-049-01) and informed consent was provided by all individual participants.

Data Sharing Statement

Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-96/dss

qims-15-07-6160-dss.pdf ^{(67.1KB, pdf)}

DOI: 10.21037/qims-2025-96