Abstract
Our findings suggested that an abnormal DMN homogeneity could be observed in MDD, which highlight the importance of the DMN in the pathophysiology of MDD.
Introduction
Major Depressive Disorder (MDD) is a leading cause of disability worldwide, yet it lacks a definitive biological marker for diagnosis. Research has consistently pointed toward the Default-Mode Network (DMN)—a set of brain regions active during self-reflection and rumination—as central to the disorder's pathophysiology. However, previous findings using traditional connectivity measures have been inconsistent. This study addresses these gaps by applying a voxel-wise Network Homogeneity (NH) metric to identify unbiased "connectivity signatures" in untreated MDD patients.
Methods and Materials
The study compared 24 first-episode, drug-naive MDD patients with 24 age- and gender-matched healthy controls. By focusing on drug-naive patients, the researchers ensured that observed brain changes were characteristic of the disorder itself rather than the effects of medication.
The researchers first defined a DMN "mask" using Independent Component Analysis (ICA) on the control group. They then calculated the NH for every voxel within this mask by measuring the correlation between its activity and all other voxels in the network. Statistical maps were generated to compare NH values between groups, and Receiver Operating Characteristic (ROC) curves were used to evaluate the diagnostic potential of these findings.
Results: A Fragmented Network
The analysis revealed a distinct "dissociation" pattern within the DMN of depressed patients. Specifically, patients exhibited increased NH in the left dorsal medial prefrontal cortex (MPFC) and decreased NH in the right inferior temporal gyrus (ITG).
The increased synchrony in the MPFC ($T = 3.62$) suggests a hyper-synchronized state in regions responsible for self-referential thought, potentially mirroring the clinical symptom of rumination. Conversely, the desynchrony in the ITG ($T = -3.11$) may reflect impaired integration of social and emotional information.
Biomarker Potential
One of the most significant findings was the ability of NH values to distinguish patients from controls. The left dorsal MPFC NH values achieved an Area Under the Curve (AUC) of 0.828, demonstrating high sensitivity (75%) and specificity (87.5%).
Interestingly, these NH abnormalities did not correlate with the severity of depressive symptoms (measured by the Hamilton Rating Scale for Depression), suggesting that these changes might represent a stable "trait" marker of MDD rather than a state-dependent symptom.
Conclusion and Future Directions
This study highlights abnormal DMN homogeneity as a key feature of MDD. While the sample size is relatively small, the high diagnostic accuracy of the MPFC NH values suggests it could serve as a candidate biomarker. Future research should explore whether these connectivity patterns normalize following treatment and investigate NH across other brain networks to provide a more comprehensive neurobiological model of depression.
Summary updated