Women's reproductive factors predict local brain aging profiles mapped using deep neural networks

Area of Science:

• Neuroscience
• Radiology
• Women's Health

Background:

• Deep neural networks (DNNs) estimate brain age from MRI, indicating neurological disease risk.
• Brain age gap (GBA) quantifies accelerated aging; local GBA (LBAG) offers voxelwise insights.
• LBAG has not been extensively studied in relation to women's health factors.

Purpose of the Study:

• To estimate LBAG in a large cohort of UK Biobank women using a novel DNN architecture.
• To investigate the relationship between LBAG, cognitive function, and women's health variables.
• To explore how reproductive factors influence brain aging trajectories in women.

Main Methods:

• Utilized a novel DNN architecture to estimate LBAG from brain MRIs of 12,284 UK Biobank females (ages 46-82).
• Correlated LBAG with chronological age, cognitive scores, and women's health data (menopause age, reproductive span, births, HT, CU).
• Analyzed regional LBAG effects, particularly in the temporal lobe and subcortical/cortical regions.

Main Results:

• Longer reproductive spans, later menopause onset, and more births were associated with younger brain age (more negative LBAG).
• These reproductive factors collectively predicted delayed regional brain aging in postmenopausal women.
• Cognitive scores showed negative correlations with LBAG, strongest in subcortical and right-hemisphere regions.

Conclusions:

• Women's reproductive health factors significantly influence brain aging and cognitive trajectories.
• Later menopause onset and longer reproductive span are biomarkers of delayed brain aging.
• LBAG analysis reveals localized brain aging patterns linked to women's health and cognition.