Abnormal Structural Covariance Network in Major Depressive Disorder

622 

Abstract Submission 

Changmin Chen¹, Zhao Qing¹

¹Southeast University, Nanjing, JiangSu

Changmin Chen  
Southeast University
Nanjing, JiangSu

Zhao Qing  
Southeast University
Nanjing, JiangSu

Major depressive disorder (MDD) is a prevalent mental disorder that can lead to disability. Structural magnetic resonance imaging (sMRI) has been widely employed to investigate structural changes in the brains of individuals with MDD. Structural covariance networks (SCN) can provide information on structural changes in the brain at the network level in addition to localized brain morphology changes. However, SCN rely on the correlation among individuals, requiring a larger sample size to obtain more reliable results. This study will use a recently established large-sample multi-center brain imaging database to explore SCN alterations in MDD.

A total of 798 T1-weighted MRI images from patients with MDD and 974 T1-weighted MRI images from health controls (HCs) from 24 sites of the REST-meta-MDD consortium were utilized after quality control. In the data analysis, voxel-based morphometry was first performed for all the images to generate a voxel-wise gray matter (GM) volume map for each subject. Then with the preprocessed GM images, we carried out the source-based morphometry processing using the GIFT toolbox. We extracted 20 components and labeled them as components A to component T. The scores for each component across subjects represent individual differences in the volume of this component. Therefore, two-sample t-tests with age, gender, education, and sites as covariates were used to examine whether there were significant group differences in each component between the MDD and NC groups. Results were corrected with the false discovery rate (FDR) method at p < 0.05. Subsequently, we calculated the SCN based on the SBM components. For each of the MDD and NC groups, Pearson' correlation analyses were performed on the individual scores between each pair of the 20 components, using age, gender, education, and sites as covariates. Interaction analyses were then utilized to assess whether the correlations were significantly different between the MDD and NC groups, the significance level was set at p < 0.01. Finally, we used the Network-based statistic (NBS) of correction to select the SCN networks that were significant between the MDD and NC groups, the significance level was set at p < 0.05.

As shown in Figure 1, SBM decomposed all the variance of the GM volumes across subjects in our data into 20 components (components A-T). The results of two-sample t-tests after FDR correction (p < 0.05) revealed that 3 of the 20 components showed significant group differences between the MDD and NC groups: J (p = 0.003), R (p = 0.003), T (p < 0.0001). As illustrated in Figure 2, the SCNs of 11 pairs of components showed significant differences between the two groups. The red lines represent the structural covariance in MDD greater than that in NC, while the blue lines indicate the opposite.

We identified 20 covariant brain components, with three components exhibiting significant differences between the MDD and HC groups. Our findings also suggest that the structural covariance network is altered in patients with MDD, and that there are both enhancements and attenuations in this alteration. Additionally, the prefrontal lobe is a key brain region in this alteration, which is consistent with existing studies.

Psychiatric (eg. Depression, Anxiety, Schizophrenia) ¹

Connectivity (eg. functional, effective, structural)

Anatomical MRI ²

Affective Disorders

Morphometrics

STRUCTURAL MRI

Other - Structural Covariance Network