Structural Analysis of Brain Hub Region Volume and Cortical Thickness in Patients with Mild Cognitive Impairment and Dementia

Background and Objectives: A complex network of axonal pathways interlinks the human brain cortex. Brain networks are not distributed evenly, and brain regions making more connections with other parts are defined as brain hubs. Our objective was to analyze brain hub region volume and cortical thickness and determine the association with cognitive assessment scores in patients with mild cognitive impairment (MCI) and dementia. Materials and Methods: In this cross-sectional study, we included 11 patients (5 mild cognitive impairment; 6 dementia). All patients underwent neurological examination, and Montreal Cognitive Assessment (MoCA) test scores were recorded. Scans with a 3T MRI scanner were done, and cortical thickness and volumetric data were acquired using Freesurfer 7.1.0 software. Results: By analyzing differences between the MCI and dementia groups, MCI patients had higher hippocampal volumes (p < 0.05) and left entorhinal cortex thickness (p < 0.05). There was a significant positive correlation between MoCA test scores and left hippocampus volume (r = 0.767, p < 0.01), right hippocampus volume (r = 0.785, p < 0.01), right precuneus cortical thickness (r = 0.648, p < 0.05), left entorhinal cortex thickness (r = 0.767, p < 0.01), and right entorhinal cortex thickness (r = 0.612, p < 0.05). Conclusions: In our study, hippocampal volume and entorhinal cortex showed significant differences in the MCI and dementia patient groups. Additionally, we found a statistically significant positive correlation between MoCA scores, hippocampal volume, entorhinal cortex thickness, and right precuneus. Although other brain hub regions did not show statistically significant differences, there should be additional research to evaluate the brain hub region association with MCI and dementia.


Introduction
The human brain cortex is interlinked by a complex network of axonal pathways that range from smaller local circuits and broader long-range fiber pathways [1,2]. There are several structural features of cortical networks that can be utilized as a quantitative variable through graph theory, i.e., nodal degree, strength, eccentricity, path length, clustering coefficient, transitivity, centrality, etc. [3][4][5] Brain networks are not distributed evenly. Brain hubs are the parts of the brain that are making many connections with other parts of the brain. [6,7] Brain network hub functionality is essential for neuronal communication and integration [6,8]. These hubs are located in the bilateral putamen, thalamus, superior parietal, superior frontal, precuneus, hippocampus, insula, right pallidum, and left lingual gyrus [7].
Several studies found a brain network hub connectivity disruption association with mild cognitive impairment and Alzheimer's disease [9][10][11].
We analyzed brain network hub volume and cortical thickness and determined whether there are distinct differences in these regions for patients with mild cognitive impairment (MCI) and dementia. Additionally, we analyzed entorhinal cortex cortical thickness that is considered as a reliable Alzheimer's disease brain biomarker [12,13].

Materials and Methods
Participants were admitted to the neurological outpatient clinic due to suspected cognitive impairment. Patients underwent neurological examination, and the Montreal Cognitive Assessment (MoCA) test was performed, and scores were recorded. Patients were divided into two groups-mild cognitive impairment and dementia. The average and median age and MoCA scores are shown in Table 1. Exclusion criteria for patients were other clinically significant neurological diseases, and drug or alcohol abuse. Patients did not have any other significant MR abnormalities (e.g., tumors, malformations, large vessel stroke).
MRI was performed on a single-site scanner to avoid scanner differences. All patients were scanned on a 3T scanner. We used the GE MP-RAGE sequence protocol with 1 mm 3 resolution and 1 mm slice thickness with the appropriate gray-white matter contrast that was evaluated for every patient. All scans were converted from DICOM format to NIFTI format to perform further analysis.
Data were analyzed with statistical analysis software JASP Version 0.13. Descriptive statistics for volumetric data and cortical thickness data were determined in the MCI and dementia groups. The Mann-Whitney U test was used to analyze brain hub volume and cortical thickness differences in the MCI and dementia groups. Spearman's correlation was calculated, and statistical significance was determined in both groups by correlating MoCA scores and hub volume and cortical thickness data.
Ethics committee (3 October 2019) and institutional review board (21 October 2019) approvals were obtained (ethical approval number: AP-144/19). Written informed consent for participation in a study and use of anonymous data was obtained for every patient.

Mean Values of Volumes and Cortical Thickness
Mean values of volumes and cortical thickness with standard deviation and standard error were calculated and are shown in Table 2.

The Mann-Whitney U Test
By analyzing differences between MCI and dementia groups, statistically significant results were found in both hemisphere hippocampal volumes and left entorhinal cortex thickness (p < 0.05). Other regions did not show statistically significant differences between the MCI and dementia groups. Multiple comparison correction was not performed when reporting p values. Thus, results serve as exploratory data that must be validated with a larger cohort and further multiple comparison correction. The results are presented in Table 3.  3 28.000 0.017 * Left Pallidum Volume, mm 3 22.000 0.247 Right Pallidum Volume, mm 3 22.000 0.247 Left Putamen Volume, mm 3 25.000 0.082 Right Putamen Volume, mm 3 22.000 0.247 Left Thalamus Volume, mm 3 21.000 0.329 Right Thalamus Volume, mm 3  The median value for left hippocampal volume (Table 4) in the MCI group was 3742.1 mm 3 , and in the dementia group 2938.0 mm 3 , and the distribution in these two groups differed significantly (Mann-Whitney U = 27, p = 0.03). The median value for right hippocampal volume (Table 4) in the MCI group was 4004.0 mm 3 , and in the dementia group 2995.4 mm 3 , and the distribution in these two groups differed significantly (Mann-Whitney U = 28, p = 0.02).
The median value for left entorhinal cortex thickness (Table 4) in the MCI group was 2.896 mm, and in the dementia group 2.226 mm, and the distribution in these two groups differed significantly (Mann-Whitney U = 28, p = 0.02).

Spearman's Correlations
The Spearman's correlations (Table 5) were conducted to determine whether there are associations of the MoCA score and volume or cortical thickness in hub regions. The two-tailed test of significance indicated that there was a significant positive correlation between MoCA score and left hippocampus volume (r = 0.767, p < 0.01), right hippocampus volume (r = 0.785, p < 0.01), right precuneus cortical thickness (r = 0.648, p < 0.05), left entorhinal cortex thickness (r = 0.767, p < 0.01), and right entorhinal cortex thickness (r = 0.612, p < 0.05).

Discussion
In this study, we compared brain hub region structural data and entorhinal cortex thickness in patients with MCI and dementia. We found that out of all brain hub regions, only hippocampal volume had statistically significant differences in both groups. Additionally, we analyzed entorhinal cortex thickness in the MCI and dementia groups and found statistically significant differences in the left entorhinal cortex.
So, to analyze the brain hub region association with MCI and dementia, we performed MoCA score, volume, and cortical thickness associations analysis. We found a significant positive correlation in both hemispheres' hippocampal volume, right precuneus cortical thickness, and both hemispheres' entorhinal cortex thickness with MoCA scores.
Many studies have found smaller hippocampal volumes in patients with dementia than MCI or healthy controls [30,31].
Further, the entorhinal cortex is proposed as a biomarker for early detection of Alzheimer's disease. [10,11].
Regarding putamen structural changes, there are mixed results. One study found that putamen had strongly reduced volumes in Alzheimer's disease [34], but also in another study, putamen structural values did not diverge from normal cognition patients across the entire lifespan [35].
Other brain hub regions have been analyzed separately and associated with cognitive decline, MCI, or dementia [35,36].
By analyzing our data, we did not find significant differences in the MCI and dementia groups apart from the structures mentioned above. This study was exploratory rather than confirmatory and performed on a small cohort. We did not perform multiple comparison correction when reporting p values. We are planning to validate our results and perform multiple comparison corrections with a larger cohort.

Conclusions
In our study, hippocampal volume and entorhinal cortex showed significant differences in the MCI and dementia patient groups. Additionally, we found a statistically significant positive correlation between MoCA scores, hippocampal volume, entorhinal cortex thickness, and right precuneus.
Although other brain hub regions did not show statistically significant differences, there should be additional research to evaluate the brain hub region association with MCI and dementia.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.