Brain-predicted age difference score is related to specific cognitive functions: a multi-site replication analysis.

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Gerontology

Background:

• Brain-predicted age difference (BPAD) is calculated by subtracting chronological age from brain age estimated via neuroimaging.
• Positive BPAD scores indicate accelerated brain aging and are linked to increased mortality and poorer physical function.
• The relationship between BPAD and specific cognitive functions remains underexplored.

Purpose of the Study:

• To systematically examine the association between BPAD and various cognitive functions.
• To validate a machine learning model for predicting chronological age from neuroimaging data across multiple independent cohorts.
• To investigate the correlation between BPAD and cognitive performance in domains relevant to aging.

Main Methods:

• Machine learning applied to 1359 T1-weighted MRI scans to predict chronological age from voxel-wise grey matter data.
• The predictive model was validated on three independent datasets (n=175, n=380, n=487) with comprehensive neuropsychological data.
• Statistical analysis focused on correlations between BPAD scores and cognitive measures.

Main Results:

• The machine learning model accurately predicted chronological age in all three independent datasets.
• Significant negative correlations were observed between higher BPAD scores and cognitive performance.
• These correlations were evident in general cognitive status, processing speed, visual attention, cognitive flexibility, and semantic verbal fluency.

Conclusions:

• This study provides robust evidence linking accelerated brain aging (higher BPAD) to diminished cognitive function in key domains.
• The findings highlight the potential of BPAD as a neuroimaging biomarker for cognitive aging.
• Further research is warranted to explore the underlying mechanisms and clinical implications of these associations.