Multivariate age-related variations in quantitative MRI maps: widespread age-related differences revisited

Area of Science:

• Neuroimaging
• Gerontology
• Biophysics

Background:

• Aging is associated with significant brain microstructural changes.
• Previous studies often used univariate analyses, potentially missing complex alterations.
• Quantitative MRI offers insights into tissue properties like myelin, iron, and water content.

Purpose of the Study:

• To apply multivariate ANOVA for investigating age-related brain microstructural changes.
• To compare the sensitivity of multivariate versus univariate analyses in detecting these changes.
• To examine coordinated alterations in myelin, iron, and water content across brain regions.

Main Methods:

• Multivariate ANOVA applied to quantitative MRI data (R1, R2*, MTsat, PD maps).
• Voxel-wise and region of interest (ROI) analyses performed on gray matter (GM) and white matter (WM).
• Data re-analysis from a previous univariate study, with cross-validation by data subset splitting.

Main Results:

• Multivariate approach identified coordinated alterations in multiple tissue properties across various brain regions (e.g., caudate nucleus, hippocampus, cerebellum).
• Demonstrated bidirectional correlations between age and MRI modalities.
• Showed increased sensitivity compared to univariate analyses in detecting age-related changes in specific regions, though cross-validation reduced sensitivity.

Conclusions:

• Multivariate analysis is more sensitive for detecting subtle, coordinated age-related microstructural brain changes.
• Concurrent age-dependent alterations in myelin, iron, and water content are confirmed.
• Highlights the importance of multivariate approaches in aging research using multi-parameter MRI.