Fundus-Derived Predicted Age Acceleration in Glaucoma Patients Using Deep Learning and Propensity Score-Matched...

Area of Science:

• Ophthalmology
• Gerontology
• Biomedical Engineering

Background:

• Glaucoma is a primary cause of irreversible blindness, often linked to aging.
• Understanding the relationship between glaucoma and biological aging is crucial for disease management.
• Accelerated biological aging may offer new insights into glaucoma pathogenesis.

Purpose of the Study:

• To investigate the association between glaucoma and accelerated biological aging using fundus-derived age prediction.
• To explore the roles of systemic factors and retinal vascular changes in this association.
• To assess the potential of fundus-derived predicted age as a biomarker for systemic aging.

Main Methods:

• Utilized a deep learning model (EfficientNet) to predict age from fundus images of 6023 participants (547 glaucoma patients, 547 matched controls).
• Analyzed systemic factors (BMI, blood pressure, lipids, liver function, glucose) and retinal vascular changes (Scheie classifications).
• Performed statistical comparisons and multivariate regression to evaluate glaucoma's impact on age acceleration, adjusting for confounders.

Main Results:

• Glaucoma patients exhibited significantly higher predicted age acceleration compared to controls.
• Glaucoma independently predicted higher chronological age and age acceleration (p=0.021).
• Elevated γ-GTP levels and specific retinal vascular changes (hypertensive, sclerotic) were associated with increased age acceleration in glaucoma patients.

Conclusions:

• Glaucoma is associated with accelerated biological aging, detectable via fundus-derived age prediction.
• Systemic factors like γ-GTP and retinal vascular changes may contribute to this accelerated aging.
• Fundus-derived predicted age shows potential as a non-invasive biomarker for monitoring systemic aging processes.