Brain age gap as a potential biomarker for schizophrenia: A multi-site structural MRI study

Area of Science:

• Neuroimaging
• Psychiatry
• Machine Learning

Background:

• Gray matter atrophy is recognized in schizophrenia, but its neurodegenerative nature is debated.
• Brain age gap (BAG) shows promise as an early neurodegeneration biomarker.
• The diagnostic value and implications of BAG in schizophrenia are not well understood.

Purpose of the Study:

• To investigate the utility of BAG as a biomarker for schizophrenia diagnosis.
• To explore the relationship between BAG and gray matter volume in schizophrenia patients.
• To assess the generalizability of BAG findings across multiple independent sites.

Main Methods:

• Utilized structural MRI data from 501 schizophrenia patients and 512 healthy controls across 8 independent sites.
• Trained a support vector regression (SVR) model on healthy controls' gray matter volume (GMV) to predict brain age.
• Applied the trained model to predict ages of schizophrenia patients and performed meta-analysis and regression analyses.

Main Results:

• Schizophrenia patients exhibited a significantly higher BAG compared to healthy controls (Cohen's d = 0.38), consistently across all sites.
• BAG significantly discriminated schizophrenia patients from healthy controls (OR = 1.07, P = 7.14 × 10^-8).
• A significant negative correlation was found between BAG and GMV in both groups, particularly in subcortical regions and the prefrontal cortex.

Conclusions:

• The brain age gap, derived from machine learning, is a potential imaging biomarker for schizophrenia.
• Elevated BAG in schizophrenia is associated with gray matter atrophy, suggesting accelerated brain aging.
• This multi-site study validates the clinical relevance of BAG as a potential diagnostic marker in schizophrenia.