Test-retest reliability and predictive utility of a macroscale principal functional connectivity gradient

Area of Science:

• Neuroscience
• Brain Imaging
• Connectomics

Background:

• Mapping individual brain function differences is challenging due to low reliability and interpretability.
• Brain-wide functional connectivity (FC) patterns show promise for understanding individual variations.
• A macroscale principal FC gradient, reflecting a sensory-to-association cortical hierarchy, is a parsimonious measure for individual differences.

Purpose of the Study:

• To evaluate the measurement reliability of global and regional principal FC gradient measures.
• To compare the reliability of principal FC gradients with traditional edge-wise FC measures.
• To assess the utility of principal FC gradient measures in predicting cognition and aging.

Main Methods:

• Utilized test-retest data from the Human Connectome Project (HCP-YA) and the Dunedin Study.
• Assessed reliabilities of global and regional principal FC gradient measures.
• Compared FC measures derived from general FC (GFC) and resting-state FC, analyzing effects of scan length.

Main Results:

• Principal FC gradient measures demonstrated higher reliability than traditional edge-wise FC.
• FC measures derived from GFC exhibited greater reliability than resting-state FC.
• Regional FC gradients and global gradient range were significantly associated with aging and moderately with cognition.

Conclusions:

• Principal FC gradient measures, particularly those derived using GFC, capture reliable and interpretable individual brain variations.
• These gradient measures offer advantages over traditional FC methods for brain-behavior association studies.
• The findings highlight the potential of FC gradients for understanding human behavior, cognition, and aging.