Major depressive disorder on a neuromorphic continuum

Area of Science:

• Neuroscience
• Psychiatry
• Radiology

Background:

• Major depressive disorder (MDD) exhibits significant heterogeneity, complicating clinical translation and the identification of reliable neuromarkers.
• Current approaches like biotyping attempt to address heterogeneity by classifying patients into subgroups, but individual variations suggest a continuum model may be more appropriate.

Purpose of the Study:

• To develop a data-driven approach to deconstruct MDD heterogeneity using structural magnetic resonance imaging (MRI) data.
• To identify latent disease factors and their individual expression patterns in MDD patients.
• To investigate the association of these factors with neurotransmitter systems and their predictive value for treatment response.

Main Methods:

• Utilized a Bayesian model to analyze structural MRI data from a multisite, cross-sectional cohort of MDD patients.
• Decomposed MRI features into three latent spatial disease factors and continuum factor compositions representing individual expression.
• Correlated identified disease factors with neurotransmitter receptor/transporter densities from open PET data and assessed stability and predictive value in longitudinal and transdiagnostic cohorts.

Main Results:

• Identified three distinct disease factors associated with specific patterns of cortical thickness changes and neurotransmitter densities (norepinephrine, 5-HT2A, 5-HTT).
• Factor 1: Sensory/orbitofrontal changes linked to norepinephrine and 5-HT2A.
• Factor 2: Cingulo-opercular/subcortical changes linked to norepinephrine and 5-HTT.
• Factor 3: Social/affective network changes linked to 5-HTT.
• Disease factor patterns predicted symptom improvement in a longitudinal cohort and showed stable individual expression over time.

Conclusions:

• MDD can be conceptualized as a continuum rather than discrete categories, characterized by continuous dimensions affecting distinct brain regions.
• The identified data-driven disease factors offer potential neuromarkers for MDD heterogeneity and treatment response.
• This dimensional approach advances understanding of MDD pathophysiology and may guide personalized treatment strategies.