A Multimodal Neurodemographic Signature for Immunometabolic Depression.

Area of Science:

• Neuroimaging
• Psychiatry
• Metabolic Disorders

Background:

• The neurobiology of immuno-metabolic depression (IMD), a major depressive disorder (MDD) subtype, is not well understood.
• IMD is characterized by inflammation and metabolic issues.

Purpose of the Study:

• To investigate the neurobiological and demographic characteristics of the IMD subtype in MDD.
• To develop a model for distinguishing IMD from non-IMD MDD subgroups.

Main Methods:

• Integrated multimodal neuroimaging (MRI) and demographic data from 145 MDD patients and 68 healthy controls.
• Defined a composite IMD score using C-reactive protein, BMI, triglycerides, and HDL cholesterol.
• Used machine learning to classify IMD and non-IMD subgroups within MDD.

Main Results:

• A multimodal model accurately classified IMD vs. non-IMD subgroups (AUC=0.826).
• The model showed higher accuracy for within-MDD subtyping than for differentiating MDD from controls.
• Key predictors included right amygdala volume and hippocampal/mid-cingulate cortex functional activity.

Conclusions:

• IMD is associated with a unique neuro-demographic signature involving cortico-limbic circuits.
• This signature aids in understanding MDD heterogeneity and supports biologically-informed patient stratification.
• The IMD subtype is clinically linked to anhedonia.