Exploring brain connectivity changes in major depressive disorder using functional-structural data fusion: A CAN-BIND-1 study

Classification of Major Depressive Disorder Using Vertex-Wise Brain Sulcal Depth, Curvature, and Thickness with a Deep and a Shallow Learning Model

Major depressive disorder (MDD) is a complex psychiatric disorder that affects the lives of hundreds of millions of individuals around the globe. Even today, researchers debate if morphological alterations in the brain are linked to MDD, likely due to the heterogeneity of this disorder. The application of deep learning tools to neuroimaging data, capable of capturing complex non-linear patterns, has the potential to provide diagnostic and predictive biomarkers for MDD. However, previous attempts to demarcate MDD patients and healthy controls (HC) based on segmented cortical features via linear machine learning approaches have reported low accuracies. In this study, we used globally representative data from the ENIGMA-MDD working group containing 7,012 participants from 30 sites (N=2,772 MDD and N=4,240 HC), which allows a comprehensive analysis with generalizable results. Based on the hypothesis that integration of vertex-wise cortical features can improve classification performance, we evaluated the classification of a DenseNet and a Support Vector Machine (SVM), with the expectation that the former would outperform the latter. As we analyzed a multi-site sample, we additionally applied the ComBat harmonization tool to remove potential nuisance effects of site. We found that both classifiers exhibited close to chance performance (balanced accuracy DenseNet: 51%; SVM: 53%), when estimated on unseen sites. Slightly higher classification performance (balanced accuracy DenseNet: 58%; SVM: 55%) was found when the cross-validation folds contained subjects from all sites, indicating site effect. In conclusion, the integration of vertex-wise morphometric features and the use of the non-linear classifier did not lead to the differentiability between MDD and HC. Our results support the notion that MDD classification on this combination of features and classifiers is unfeasible. Future studies are needed to determine whether more sophisticated integration of information from other MRI modalities such as fMRI and DWI will lead to a higher performance in this diagnostic task.

Assessing remission in major depressive disorder using a functional-structural data fusion pipeline: A CAN-BIND-1 study

Neural network-level changes underlying symptom remission in major depressive disorder (MDD) are often studied from a single perspective. Multimodal approaches to assess neuropsychiatric disorders are evolving, as they offer richer information about brain networks. A FATCAT-awFC pipeline was developed to integrate a computationally intense data fusion method with a toolbox, to produce a faster and more intuitive pipeline for combining functional connectivity with structural connectivity (denoted as anatomically weighted functional connectivity (awFC)). Ninety-three participants from the Canadian Biomarker Integration Network for Depression study (CAN-BIND-1) were included. Patients with MDD were treated with 8 weeks of escitalopram and adjunctive aripiprazole for another 8 weeks. Between-group connectivity (SC, FC, awFC) comparisons contrasted remitters (REM) with non-remitters (NREM) at baseline and 8 weeks. Additionally, a longitudinal study analysis was performed to compare connectivity changes across time for REM, from baseline to week-8. Association between cognitive variables and connectivity were also assessed. REM were distinguished from NREM by lower awFC within the default mode, frontoparietal, and ventral attention networks. Compared to REM at baseline, REM at week-8 revealed increased awFC within the dorsal attention network and decreased awFC within the frontoparietal network. A medium effect size was observed for most results. AwFC in the frontoparietal network was associated with neurocognitive index and cognitive flexibility for the NREM group at week-8. In conclusion, the FATCAT-awFC pipeline has the benefit of providing insight on the 'full picture' of connectivity changes for REMs and NREMs while making for an easy intuitive approach.

Characterizing major depressive disorder and substance use disorder using heatmaps and variable interactions: The utility of operant behavior and brain structure relationships

BACKGROUND

Rates of depression and addiction have risen drastically over the past decade, but the lack of integrative techniques remains a barrier to accurate diagnoses of these mental illnesses. Changes in reward/aversion behavior and corresponding brain structures have been identified in those with major depressive disorder (MDD) and cocaine-dependence polysubstance abuse disorder (CD). Assessment of statistical interactions between computational behavior and brain structure may quantitatively segregate MDD and CD.

METHODS

Here, 111 participants [40 controls (CTRL), 25 MDD, 46 CD] underwent structural brain MRI and completed an operant keypress task to produce computational judgment metrics. Three analyses were performed: (1) linear regression to evaluate groupwise (CTRL v. MDD v. CD) differences in structure-behavior associations, (2) qualitative and quantitative heatmap assessment of structure-behavior association patterns, and (3) the k-nearest neighbor machine learning approach using brain structure and keypress variable inputs to discriminate groups.

RESULTS

This study yielded three primary findings. First, CTRL, MDD, and CD participants had distinct structure-behavior linear relationships, with only 7.8% of associations overlapping between any two groups. Second, the three groups had statistically distinct slopes and qualitatively distinct association patterns. Third, a machine learning approach could discriminate between CTRL and CD, but not MDD participants.

CONCLUSIONS

These findings demonstrate that variable interactions between computational behavior and brain structure, and the patterns of these interactions, segregate MDD and CD. This work raises the hypothesis that analysis of interactions between operant tasks and structural neuroimaging might aide in the objective classification of MDD, CD and other mental health conditions.

Effects of Cortisol Administration on Resting-State Functional Connectivity in Women with Depression

Previous resting-state functional connectivity (rsFC) research has identified several brain networks impacted by depression and cortisol, including default mode (DMN), frontoparietal (FPN), and salience networks (SN). In the present study, we examined the effects of cortisol administration on rsFC of these networks in individuals varying in depression history and severity. We collected resting-state fMRI scans and self-reported depression symptom severity for 74 women with and without a history of depression after cortisol and placebo administration using a double-blind, crossover design. We conducted seed-based rsFC analyses for DMN, FPN, and SN seeds to examine rsFC changes after cortisol vs. placebo administration in relation to depression history group and severity. Results revealed a main effect of depression group, with lower left amygdala (SN)-middle temporal gyrus connectivity in women with a history of depression. Cortisol administration increased insula (SN)-inferior frontal gyrus and superior temporal gyrus connectivity. We also found that greater depression severity was associated with increased PCC (DMN)-cerebellum connectivity after cortisol. These results did not survive Bonferroni correction for seed ROIs and should be interpreted with caution. Our findings indicate that acute cortisol elevation may normalize aberrant connectivity of DMN and SN regions, which could help inform clinical treatments for depression.

Examining resting-state network connectivity in children exposed to perinatal maternal adversity using anatomically weighted functional connectivity (awFC) analyses; A preliminary report

Introduction

Environmental perturbations during critical periods can have pervasive, organizational effects on neurodevelopment. To date, the literature examining the long-term impact of early life adversity has largely investigated structural and functional imaging data outcomes independently. However, emerging research points to a relationship between functional connectivity and the brain’s underlying structural architecture. For instance, functional connectivity can be mediated by the presence of direct or indirect anatomical pathways. Such evidence warrants the use of structural and functional imaging in tandem to study network maturation. Accordingly, this study examines the impact of poor maternal mental health and socioeconomic context during the perinatal period on network connectivity in middle childhood using an anatomically weighted functional connectivity (awFC) approach. awFC is a statistical model that identifies neural networks by incorporating information from both structural and functional imaging data.

Methods

Resting-state fMRI and DTI scans were acquired from children aged 7–9 years old.

Results

Our results indicate that maternal adversity during the perinatal period can affect offspring’s resting-state network connectivity during middle childhood. Specifically, in comparison to controls, children of mothers who had poor perinatal maternal mental health and/or low socioeconomic status exhibited greater awFC in the ventral attention network.

Discussion

These group differences were discussed in terms of the role this network plays in attention processing and maturational changes that may accompany the consolidation of a more adult-like functional cortical organization. Furthermore, our results suggest that there is value in using an awFC approach as it may be more sensitive in highlighting connectivity differences in developmental networks associated with higher-order cognitive and emotional processing, as compared to stand-alone FC or SC analyses.

Percentage amplitude of fluctuation and structural covariance changes of subjective cognitive decline in patients: A multimodal imaging study

Back ground

Subjective cognitive decline (SCD) may be the first clinical sign of Alzheimer’s disease (AD). The possible neural mechanisms of SCD are not well known. This study aimed to compare percent amplitude of fluctuation (PerAF) and structural covariance patterns in patients with SCD and healthy controls (HCs).

Methods

We enrolled 53 patients with SCD and 65 HCs. Resting-state functional magnetic resonance imaging (MRI) data and T1-weighted anatomical brain 3.0-T MRI scans were collected. The PerAF approach was applied to distinguish altered brain functions between the two groups. A whole-brain voxel-based morphometry analysis was performed, and all significant regions were selected as regions of interest (ROIs) for the structural covariance analysis. Statistical analysis was performed using two-sample t-tests, and multiple regressions were applied to examine the relationships between neuroimaging findings and clinical symptoms.

Results

Functional MRI results revealed significantly increased PerAF including the right hippocampus (HIPP) and right thalamus (THA) in patients with SCD relative to HCs. Gray matter volume (GMV) results demonstrated decreased GMV in the bilateral ventrolateral prefrontal cortex (vlPFC) and right insula in patients with SCD relative to HCs. Taking these three areas including the bilateral vlPFC and right insula as ROIs, differences were observed in the structural covariance of the ROIs with several regions between the two groups. Additionally, significant correlations were observed between neuroimaging findings and clinical symptoms.

Conclusion

Our study investigated the abnormal PerAF and structural covariance patterns in patients with SCD, which might provide new insights into the pathological mechanisms of SCD.

Altered Neural Network Connectivity Predicts Depression in de novo Parkinson’s Disease

Background

Depression, one of the most frequent non-motor symptoms in Parkinson’s disease (PD), was proposed to be related to neural network dysfunction in advanced PD patients. However, the underlying mechanisms in the early stage remain unclear. The study was aimed to explore the alterations of large-scale neural networks in de novo PD patients with depression.

Methods

We performed independent component analysis (ICA) on the data of resting-state functional magnetic resonance imaging from 21 de novo PD patients with depression (dPD), 34 de novo PD patients without depression (ndPD), and 43 healthy controls (HCs) to extract functional networks. Intranetwork and internetwork connectivity was calculated for comparison between groups, correlation analysis, and predicting the occurrence of depression in PD.

Results

We observed an ordered decrease of connectivity among groups within the ventral attention network (VAN) (dPD < ndPD < HCs), mainly located in the left middle temporal cortex. Besides, dPD patients exhibited hypoconnectivity between the auditory network (AUD) and default mode network (DMN) or VAN compared to ndPD patients or healthy controls. Correlation analysis revealed that depression severity was negatively correlated with connectivity value within VAN and positively correlated with the connectivity value of AUD-VAN in dPD patients, respectively. Further analysis showed that the area under the curve (AUC) for dPD prediction was 0.863 when combining the intranetwork connectivity in VAN and internetwork connectivity in AUD-DMN and AUD-VAN.

Conclusion

Our results demonstrated that early dPD may be associated with abnormality of attention bias and especially auditory attention processing. Altered neural network connectivity is expected to be a potential neuroimaging biomarker to predict depression in PD.

Fluoxetine-induced hepatic lipid accumulation is mediated by prostaglandin endoperoxide synthase 1 and is linked to elevated 15-deoxy-Δ12,14 PGJ2

Major depressive disorder and other neuropsychiatric disorders are often managed with long-term use of antidepressant medication. Fluoxetine, an SSRI antidepressant, is widely used as a first-line treatment for neuropsychiatric disorders. However, fluoxetine has also been shown to increase the risk of metabolic diseases such as non-alcoholic fatty liver disease. Fluoxetine has been shown to increase hepatic lipid accumulation in vivo and in vitro. In addition, fluoxetine has been shown to alter the production of prostaglandins which have also been implicated in the development of non-alcoholic fatty liver disease. The goal of this study was to assess the effect of fluoxetine exposure on the prostaglandin biosynthetic pathway and lipid accumulation in a hepatic cell line (H4-II-E-C3 cells). Fluoxetine treatment increased mRNA expression of prostaglandin biosynthetic enzymes (Ptgs1, Ptgs2, and Ptgds), PPAR gamma (Pparg), and PPAR gamma downstream targets involved in fatty acid uptake (Cd36, Fatp2, and Fatp5) as well as production of 15-deoxy-Δ^12,14 PGJ₂ a PPAR gamma ligand. The effects of fluoxetine to induce lipid accumulation were attenuated with a PTGS1 specific inhibitor (SC-560), whereas inhibition of PTGS2 had no effect. Moreover, SC-560 attenuated 15-deoxy-Δ^12,14 PGJ₂ production and expression of PPAR gamma downstream target genes. Taken together these results suggest that fluoxetine-induced lipid abnormalities appear to be mediated via PTGS1 and its downstream product 15d-PGJ₂ and suggest a novel therapeutic target to prevent some of the adverse effects of fluoxetine treatment.