Whole-brain structural covariance network abnormality in first-episode and drug-naïve major depressive disorder

Explainable multimodal prediction of treatment-resistance in patients with depression leveraging brain morphometry and natural language processing

Although 20 % of patients with depression receiving treatment do not achieve remission, predicting treatment-resistant depression (TRD) remains challenging. In this study, we aimed to develop an explainable multimodal prediction model for TRD using structured electronic medical record data, brain morphometry, and natural language processing. In total, 247 patients with a new depressive episode were included. TRD-predictive models were developed based on the combination of following parameters: selected tabular dataset features, independent components-map weightings from brain T1-weighted magnetic resonance imaging (MRI), and topic probabilities from clinical notes. All models applied the extreme gradient boosting (XGBoost) algorithm via five-fold cross-validation. The model using all data sources showed the highest area under the receiver operating characteristic of 0.794, followed by models that used combined brain MRI and structured data, brain MRI and clinical notes, clinical notes and structured data, brain MRI only, structured data only, and clinical notes only (0.770, 0.762, 0.728, 0.703, 0.684, and 0.569, respectively). Classifications of TRD were driven by several predictors, such as previous exposure to antidepressants and antihypertensive medications, sensorimotor network, default mode network, and somatic symptoms. Our findings suggest that a combination of clinical data with neuroimaging and natural language processing variables improves the prediction of TRD.

Fronto-striato network function is reduced in major depressive disorder

Introduction

Major depressive disorder (MDD) is a major cause of poor quality of life and disability and is highly prevalent worldwide. Various pathological mechanisms are implicated in MDD, including the reward system. The human brain is equipped with a reward system that is involved in aspects such as motivation, pleasure, and learning. Several studies including a meta-analysis have been reported on the reward system network and MDD. However, to our knowledge, no studies have examined the relationship between the reward system network of drug-naïve, first-episode MDD patients and the detailed symptoms of MDD or age. The fronto-striato network (FSN) is closely related to the reward system network. The present study primarily aimed to elucidate this point.

Methods

A total of 89 drug-naïve first-episode MDD patients and 82 healthy controls (HCs) patients were enrolled in the study. The correlation between the FSN and age and the interaction between age and illness in the FSN were investigated in 75 patients in the MDD group and 79 patients in the HC group with available information on the FSN and age. In addition, the association between the FSN and the total scores on the 17-item Hamilton Rating Scale for Depression (HAMD-17) and scores in each symptom item was analyzed in 76 MDD subjects with information on the FSN and HAMD-17. The significance of each result was evaluated according to a p-value of <0.05.

Results

Age was inversely correlated with the FSN (p=2.14e-11) in the HC group but not in the MDD group (p=0.79). FSN varied with the presence of MDD and with age, particularly showing an interaction with MDD and age (p=1.04e-08). Specifically, age and the presence or absence of MDD each affected FSN, but the effect of age on FSN changed in the presence of depression. FSN did not correlate with total HAMD-17 scores or scores in each item.

Discussion

The reward system may be dysfunctional in patients with MDD. In addition, the effect could be greater in younger patients. Meanwhile, there is no correlation between the function of the reward system and the severity of MDD or the severity of each symptom. Thus, the reward system network may be an important biological marker of MDD, although careful consideration should be given to age and its association with the severity of the disorder.

Conclusion

The reward system function is decreased in MDD patients, and this decrease may be more pronounced in younger patients, although further research is still needed.

Relationship between brain structural network integrity and emotional symptoms in youth with perinatally-acquired HIV

Perinatally acquired HIV infection (PHIV) currently affects approximately 1.7 million children worldwide. Youth with PHIV (YPHIV) are at increased risk for emotional and behavioral symptoms, yet few studies have examined relationships between these symptoms and brain structure. Previous neuroimaging studies in YPHIV report alterations within the salience network (SN), cognitive control network (CCN), and default mode network (DMN). These areas have been associated with social and emotional processing, emotion regulation, and executive function. We examined structural brain network integrity from MRI using morphometric similarity networks and graph theoretical measures of segregation (transitivity), resilience (assortativity), and integration (global efficiency). We examined brain network integrity of 40 YPHIV compared to 214 youths without HIV exposure or infection. Amongst YPHIV, we related structural brain network metrics to the Emotional Symptoms Index of the Behavioral Assessment System for Children, 2nd edition. We also examined the relationship of inflammatory biomarkers in YPHIV to brain network integrity. YPHIV had significantly lower global efficiency in the SN, DMN, and the whole brain network compared to controls. YPHIV also demonstrated lower assortativity or resilience (i.e., network robustness) compared to controls in the DMN and whole brain network. Further, higher emotional symptom score was associated with higher global efficiency in the SN and lower global efficiency in the DMN, signaling more emotional challenges. A significant association was also found between several inflammatory and cardiac markers with structural network integrity. These findings suggest an impact of HIV on developing brain networks, and potential dysfunction of the SN and DMN in relation to network efficiency.

Triple network connectivity changes in patients with major depressive disorder versus healthy controls via structural network imaging after electroconvulsive therapy treatment

OBJECTIVE

To investigate the effect of electroconvulsive treatment (ECT) on dynamic structural network connectivity in major depressive disorder (MDD), based on the triple-network model.

METHODS

Twenty-one first-episode, drug-naïve patients with MDD and 21 age- and sex-matched healthy subjects were recruited. Bilateral electrical stimulation was performed thrice a week for a total of 4-5 weeks in the MDD group. MRI data were obtained, and triple-network structural connectivity was evaluated using source-based morphometry (SBM) analysis. A paired t-test was used to analyze structural connectivity differences between pre- and post-ECT MDD groups, one-way analysis was used to calculate three intrinsic network differences between HCs, pre- and post-ECT groups, and partial least squares structural equation modelling was used to investigate dynamic structural network connectivity (dSNC) across groups.

RESULTS

Pre-ECT patients with MDD exhibited significantly lower salience network (SN) structural connectivity (p = 0.010) than the healthy control (HC) group and after ECT therapy SN structural connectivity was significantly elevated (p = 0.002) in post-ECT group compared with pre-ECT. PLS-SEM analysis conducted on inter-network connectivity in the triple-network model indicated a significant difference between SN and central executive network (CEN) in all three groups. The HC and post-ECT MDD groups showed notable direct connectivity between the SN and default mode network (DMN), while the pre-ECT MDD group showed consequential pathological connectivity between the CEN and DMN. A mediation analysis revealed a significant indirect effect of the SN on the DMN through the CEN (β = 0.363, p = 0.008) only in the pre-ECT MDD group.

CONCLUSIONS

ECT may be an effective and minimally invasive treatment for addressing structural changes in the SN and direct communication abnormalities between the three core brain networks in patients with MDD, with possible beneficial correction of indirect connections.

Altered topology of individual brain structural covariance networks in major depressive disorder

BACKGROUND

The neurobiological pathogenesis of major depression disorder (MDD) remains largely controversial. Previous literatures with limited sample size utilizing group-level structural covariance networks (SCN) commonly generated mixed findings regarding the topology of brain networks.

METHODS

We analyzed T1 images from a high-powered multisite sample including 1173 patients with MDD and 1019 healthy controls (HCs). We used regional gray matter volume to construct individual SCN by utilizing a novel approach based on the interregional effect size difference. We further investigated MDD-related structural connectivity alterations using topological metrics.

RESULTS

Compared to HCs, the MDD patients showed a shift toward randomization characterized by increased integration. Further subgroup analysis of patients in different stages revealed this randomization pattern was also observed in patients with recurrent MDD, while the first-episode drug naïve patients exhibited decreased segregation. Altered nodal properties in several brain regions which have a key role in both emotion regulation and executive control were also found in MDD patients compared with HCs. The abnormalities in inferior temporal gyrus were not influenced by any specific site. Moreover, antidepressants increased nodal efficiency in the anterior ventromedial prefrontal cortex.

CONCLUSIONS

The MDD patients at different stages exhibit distinct patterns of randomization in their brain networks, with increased integration during illness progression. These findings provide valuable insights into the disruption in structural brain networks that occurs in patients with MDD and might be useful to guide future therapeutic interventions.

Resolving heterogeneity in depression using individualized structural covariance network analysis

BACKGROUND

Elucidating individual aberrance is a critical first step toward precision medicine for heterogeneous disorders such as depression. The neuropathology of depression is related to abnormal inter-regional structural covariance indicating a brain maturational disruption. However, most studies focus on group-level structural covariance aberrance and ignore the interindividual heterogeneity. For that reason, we aimed to identify individualized structural covariance aberrance with the help of individualized differential structural covariance network (IDSCN) analysis.

METHODS

T1-weighted anatomical images of 195 first-episode untreated patients with depression and matched healthy controls (n = 78) were acquired. We obtained IDSCN for each patient and identified subtypes of depression based on shared differential edges.

RESULTS

As a result, patients with depression demonstrated tremendous heterogeneity in the distribution of differential structural covariance edges. Despite this heterogeneity, altered edges within subcortical-cerebellum network were often shared by most of the patients. Two robust neuroanatomical subtypes were identified. Specifically, patients in subtype 1 often shared decreased motor network-related edges. Patients in subtype 2 often shared decreased subcortical-cerebellum network-related edges. Functional annotation further revealed that differential edges in subtype 2 were mainly implicated in reward/motivation-related functional terms.

CONCLUSIONS

In conclusion, we investigated individualized differential structural covariance and identified that decreased edges within subcortical-cerebellum network are often shared by patients with depression. The identified two subtypes provide new insights into taxonomy and facilitate potential clues to precision diagnosis and treatment of depression.

Brain structural abnormalities in adult major depressive disorder revealed by voxel- and source-based morphometry: evidence from the REST-meta-MDD Consortium

BACKGROUND

Neuroimaging studies on major depressive disorder (MDD) have identified an extensive range of brain structural abnormalities, but the exact neural mechanisms associated with MDD remain elusive. Most previous studies were performed with voxel- or surface-based morphometry which were univariate methods without considering spatial information across voxels/vertices.

METHODS

Brain morphology was investigated using voxel-based morphometry (VBM) and source-based morphometry (SBM) in 1082 MDD patients and 990 healthy controls (HCs) from the REST-meta-MDD Consortium. We first examined group differences in regional grey matter (GM) volumes and structural covariance networks between patients and HCs. We then compared first-episode, drug-naïve (FEDN) patients, and recurrent patients. Additionally, we assessed the effects of symptom severity and illness duration on brain alterations.

RESULTS

VBM showed decreased GM volume in various regions in MDD patients including the superior temporal cortex, anterior and middle cingulate cortex, inferior frontal cortex, and precuneus. SBM returned differences only in the prefrontal network. Comparisons between FEDN and recurrent MDD patients showed no significant differences by VBM, but SBM showed greater decreases in prefrontal, basal ganglia, visual, and cerebellar networks in the recurrent group. Moreover, depression severity was associated with volumes in the inferior frontal gyrus and precuneus, as well as the prefrontal network.

CONCLUSIONS

Simultaneous application of VBM and SBM methods revealed brain alterations in MDD patients and specified differences between recurrent and FEDN patients, which tentatively provide an effective multivariate method to identify potential neurobiological markers for depression.

Disturbed hippocampal intra-network in first-episode of drug-naïve major depressive disorder

Complex networks inside the hippocampus could provide new insights into hippocampal abnormalities in various psychiatric disorders and dementia. However, evaluating intra-networks in the hippocampus using MRI is challenging. Here, we employed a high spatial resolution of conventional structural imaging and independent component analysis to investigate intra-networks structural covariance in the hippocampus. We extracted the intra-networks based on the intrinsic connectivity of each 0.9 mm isotropic voxel to every other voxel using a data-driven approach. With a total volume of 3 cc, the hippocampus contains 4115 voxels for a 0.9 mm isotropic voxel size or 375 voxels for a 2 mm isotropic voxel of high-resolution functional or diffusion tensor imaging. Therefore, the novel method presented in the current study could evaluate the hippocampal intra-networks in detail. Furthermore, we investigated the abnormality of the intra-networks in major depressive disorders. A total of 77 patients with first-episode drug-naïve major depressive disorder and 79 healthy subjects were recruited. The independent component analysis extracted seven intra-networks from hippocampal structural images, which were divided into four bilateral networks and three networks along the longitudinal axis. A significant difference was observed in the bilateral hippocampal tail network between patients with major depressive disorder and healthy subjects. In the logistic regression analysis, two bilateral networks were significant predictors of major depressive disorder, with an accuracy of 78.1%. In conclusion, we present a novel method for evaluating intra-networks in the hippocampus. One advantage of this method is that a detailed network can be estimated using conventional structural imaging. In addition, we found novel bilateral networks in the hippocampus that were disturbed in patients with major depressive disorders, and these bilateral networks could predict major depressive disorders.

Structural covariance network of the hippocampus-amygdala complex in medication-naïve patients with first-episode major depressive disorder

Background

The hippocampus and amygdala are densely interconnected structures that work together in multiple affective and cognitive processes that are important to the etiology of major depressive disorder (MDD). Each of these structures consists of several heterogeneous subfields. We aim to explore the topologic properties of the volume-based intrinsic network within the hippocampus-amygdala complex in medication-naïve patients with first-episode MDD.

Methods

High-resolution T1-weighted magnetic resonance imaging scans were acquired from 123 first-episode, medication-naïve, and noncomorbid MDD patients and 81 age-, sex-, and education level-matched healthy control participants (HCs). The structural covariance network (SCN) was constructed for each group using the volumes of the hippocampal subfields and amygdala subregions; the weights of the edges were defined by the partial correlation coefficients between each pair of subfields/subregions, controlled for age, sex, education level, and intracranial volume. The global and nodal graph metrics were calculated and compared between groups.

Results

Compared with HCs, the SCN within the hippocampus-amygdala complex in patients with MDD showed a shortened mean characteristic path length, reduced modularity, and reduced small-worldness index. At the nodal level, the left hippocampal tail showed increased measures of centrality, segregation, and integration, while nodes in the left amygdala showed decreased measures of centrality, segregation, and integration in patients with MDD compared with HCs.

Conclusion

Our results provide the first evidence of atypical topologic characteristics within the hippocampus-amygdala complex in patients with MDD using structure network analysis. It provides more delineate mechanism of those two structures that underlying neuropathologic process in MDD.

Large-scale structural network change correlates with clinical response to rTMS in depression

Response to repetitive transcranial magnetic stimulation (rTMS) among individuals with major depressive disorder (MDD) varies widely. The neural mechanisms underlying rTMS are thought to involve changes in large-scale networks. Whether structural network integrity and plasticity are associated with response to rTMS therapy is unclear. Structural MRIs were acquired from a series of 70 adult healthy controls and 268 persons with MDD who participated in two arms of a large randomized, non-inferiority trial, THREE-D, comparing intermittent theta-burst stimulation to high-frequency rTMS of the left dorsolateral prefrontal cortex (DLPFC). Patients were grouped according to percentage improvement on the 17-item Hamilton Depression Rating Score at treatment completion. For the entire sample and then for each treatment arm, multivariate analyses were used to characterize structural covariance networks (SCN) from cortical gray matter thickness, volume, and surface area maps from T1-weighted MRI. The association between SCNs and clinical improvement was assessed. For both study arms, cortical thickness and volume SCNs distinguished healthy controls from MDD (p = 0.005); however, post-hoc analyses did not reveal a significant association between pre-treatment SCN expression and clinical improvement. We also isolated an anticorrelated SCN between the left DLPFC rTMS target site and the subgenual anterior cingulate cortex across cortical measures (p = 0.0004). Post-treatment change in cortical thickness SCN architecture was associated with clinical improvement in treatment responders (p = 0.001), but not in non-responders. Structural network changes may underpin clinical response to rTMS, and SCNs are useful for understanding the pathophysiology of depression and neural mechanisms of plasticity and response to circuit-based treatments.

Altered brain network in first-episode, drug-naive patients with major depressive disorder

BACKGROUND

Resting-state functional magnetic resonance imaging has been widely used for the assessment of brain functional network, yet with inconsistent results. The present study aimed to investigate intranetwork and internetwork connectivity differences between patients with major depressive disorder (MDD) and healthy controls at the integrity, network and edge levels of 8 well-defined resting state networks.

METHODS

Thirty patients with MDD and sixty-three healthy control subjects were recruited in this study.

RESULTS

Compared with healthy controls, patients with MDD showed increased node degree in the right amygdala and putamen, increased connectivity strength in the deep gray matter network (DGN) and increased functional connectivity in intranetwork and internetwork. Meanwhile, MDD showed decreased connectivity strength in visual network-DGN pair.

LIMITATIONS

The sample size was small, and all patients in this study were of Asian ethnicity, especially Han individuals.

CONCLUSIONS

These findings demonstrate that MDD cases and healthy controls may have divergent intranetwork and internetwork connectivity at an early stage without confounding influence of medication. These differences may underlie cognitive and behavioral alterations in patients with MDD. And these differences may help with the discrimination of patients and healthy people at an early stage of MDD. More studies in the future are warranted to assist in the diagnosis of this burdensome disease.

Volume of Amygdala Subregions and Clinical Manifestations in Patients With First-Episode, Drug-Naïve Major Depression

We examined amygdala subregion volumes in patients with a first episode of major depression (MD) and in healthy subjects. Covariate-adjusted linear regression was performed to compare the MD and healthy groups, and adjustments for age, gender, and total estimated intracranial volume showed no differences in amygdala subregion volumes between the healthy and MD groups. Within the MD group, we examined the association between amygdala subregion volume and the 17-item Hamilton Rating Scale for Depression (HAMD) score and the HAMD subscale score, and found no association in the left amygdala. In the right amygdala, however, there was an inverse linear association between the HAMD total and the HAMD core and lateral nucleus and anterior-amygdaloid-regions. Furthermore, an inverse linear association was seen between the HAMD psychic and the lateral nucleus, anterior-amygdaloid-regions, transition, and whole amygdala. The findings of this study suggest that the severity of MD and some symptoms of MD are associated with right amygdala volume. There have been few reports on the relationship between MD and amygdala subregional volume, and further research is needed to accumulate more data for further validation.

Gout Is Not Just Arthritis: Abnormal Cortical Thickness and Structural Covariance Networks in Gout

Background: Hyperuricemia is the cause of gout. The antioxidant and neuroprotective effects of uric acid seem to benefit some patients with central nervous system injury. However, changes in the brain structure have not been discovered in patients with gout.

Object: Clarify the changes in cortical thickness in patients with gout and the alteration of the structural covariance networks (SCNs) based on cortical thickness.

Methods: We collected structural MRIs of 23 male gout patients and 23 age-matched healthy controls. After calculating and comparing the difference in cortical thickness between the two groups, we constructed and analyzed the cortical thickness covariance networks of the two groups, and we investigated for any changes in SCNs of gout patients.

Results: Gout patients have thicker cortices in the left postcentral, left supramarginal, right medial temporal, and right medial orbitofrontal regions; and thinner cortices were found in the left insula, left superior frontal, right pericalcarine, and right precentral regions. In SCN analysis, between-group differences in global network measures showed that gout patients have a higher global efficiency. In regional network measures, more nodes in gout patients have increased centrality. In network hub analysis, we found that the transfer of the core hub area, rather than the change in number, may be the characteristic of the gout's cortical thickness covariance network.

Conclusion: This is the first study on changes in brain cortical thickness and SCN based on graph theory in patients with gout. The present study found that, compared with healthy controls, gout patients show regional cortical thinning or thickening, and variation in the properties of the cortical thickness covariance network also changed. These alterations may be the combined effect of disease damage and physiological compensation. More research is needed to fully understand the complex underlying mechanisms of gout brain variation.

Homogeneous grey matter patterns in patients with obsessive-compulsive disorder

BACKGROUND

Changes in grey matter volume have frequently been reported in patients with obsessive-compulsive disorder (OCD). Most studies performed whole brain or region-of-interest based analyses whereas grey matter volume based on structural covariance networks has barely been investigated up to now. Therefore, the present study investigated grey matter volume within structural covariance networks in a sample of 228 participants (n = 117 OCD patients, n = 111 healthy controls).

METHODS

First, an independent component analysis (ICA) was performed on all subjects' preprocessed T1 images to derive covariance-dependent morphometric networks. Then, grey matter volume from each of the ICA-derived morphometric networks was extracted and compared between the groups. In addition, we performed logistic regressions and receiver operating characteristic (ROC) analyses to investigate whether network-related grey matter volume could serve as a characteristic that allows to differentiate patients from healthy volunteers. Moreover, we assessed grey matter pattern organization by correlating grey matter volume in all networks across all participants. Finally, we explored a potential association between grey matter volume or whole-brain grey matter pattern organization and clinical characteristics in terms of symptom severity and duration of illness.

RESULTS

There were only subtle group differences in network-related grey matter volume. Network-related grey matter volume had moreover a very poor discrimination performance. We found, however, significant group differences with regard to grey matter pattern organization. When correlating grey matter volume in all networks across all participants, patients showed a significantly higher homogeneity across all networks and a significantly lower heterogeneity, as assessed by the coefficient of variation across all networks as well as in several single networks. There was no association with clinical characteristics.

CONCLUSION

The findings of the present study suggest that the pathological mechanisms of OCD reduce interindividual grey matter variability. We assume that common characteristics associated with the disorder may lead to a more uniform, disorder-specific morphometry.

Cortical Thinning and Abnormal Structural Covariance Network After Three Hours Sleep Restriction

Sleep loss leads to serious health problems, impaired attention, and emotional processing. It has been suggested that the abnormal neurobehavioral performance after sleep deprivation was involved in dysfunction of specific functional connectivity between brain areas. However, to the best of our knowledge, there was no study investigating the structural connectivity mechanisms underlying the dysfunction at network level. Surface morphological analysis and graph theoretical analysis were employed to investigate changes in cortical thickness following 3 h sleep restriction, and test whether the topological properties of structural covariance network was affected by sleep restriction. We found that sleep restriction significantly decreased cortical thickness in the right parieto-occipital cortex (Brodmann area 19). In addition, graph theoretical analysis revealed significantly enhanced global properties of structural covariance network including clustering coefficient and local efficiency, and increased nodal properties of the left insula cortex including nodal efficiency and betweenness, after 3 h sleep restriction. These results provided insights into understanding structural mechanisms of dysfunction of large-scale functional networks after sleep restriction.

Association of Serum Kynurenine Levels and Neural Networks in Patients with First-Episode, Drug-Naïve Major Depression: A Source-Based Morphometry Study

PURPOSE

The kynurenine (KYN) pathway can directly or indirectly influence cerebral volume and neural integrity in patients with major depression (MD). The aim of the present study was to investigate neural network systems and the KYN pathway in patients with first-episode, drug-naïve MD.

PATIENTS AND METHODS

Twenty right-handed drug-naïve patients, with MD diagnosed using the Diagnostic and Statistical Manual for Mental Disorders, Fourth Edition, Text Revision, Research Version, were included in this study. Magnetic resonance imaging scans and scores on the Hamilton Rating Scale for Depression were assessed, and serum sampling was performed prior to the start of pharmacological treatment. Image processing and data analysis were performed according to our recently published procedure. Serum metabolomes were measured in the cation and anion modes of CE-FTMS-based metabolome analysis.

RESULTS

We found that serum KYN levels were positively correlated with the Z-scores of the salience network but not with those of the central executive network or default mode network. No associations were observed between serum glutamate levels and the Z-score of any of the three networks.

CONCLUSION

Our results indicate that serum KYN levels might affect the activity of the salience network in first-episode, drug-naïve patients with MD.