Quantifying Deviations of Brain Structure and Function in Major Depressive Disorder Across Neuroimaging Modalities

Area of Science:

• Clinical Neuroscience
• Neuroimaging
• Psychiatry

Background:

• Decades of research sought neurobiological differences in major depressive disorder (MDD).
• Recent meta-analyses question the replicability and clinical relevance of identified brain alterations in depression.

Purpose of the Study:

• To quantify effect sizes, predictive utility, and distributional dissimilarity across neuroimaging modalities in MDD.
• To compare neuroimaging findings with MDD polygenic risk scores (PRS) and environmental variables.

Main Methods:

• Cross-sectional, case-control study utilizing structural MRI, diffusion-tensor imaging, and functional MRI (task-based and resting-state).
• Included 1809 participants (861 MDD patients, 948 controls), aged 18-65.
• Analyses controlled for age, sex, and modality-specific confounders; secondary analyses examined acute vs. chronic MDD subgroups.

Main Results:

• Maximum univariate effect sizes (partial η²) ranged from 0.004 to 0.017, indicating minimal group differences.
• Distributional overlap between 87-95% and classification accuracies of 54-56% across modalities.
• Findings were comparable to polygenic risk scores but smaller than environmental variables; consistent across acute/chronic MDD.

Conclusions:

• Deviations between MDD patients and controls are remarkably small, even at maximum univariate effect sizes.
• Single-participant prediction is not feasible, with significant overlap between groups.
• Biological psychiatry requires more meaningful outcome measures and predictive approaches for personalized clinical practice.