Optical Coherence Tomography as a Proxy for Accelerated Brain Aging and Multimodal Neurodegeneration in Multiple Sclerosis

Neurodegeneration and accelerated brain aging are hallmarks of multiple sclerosis (MS). Optical coherence tomography (OCT) captures retinal neuroaxonal loss, that mirrors central nervous system damage. Brain-predicted age (BA) and the brain age gap (BAG) are machine learning derived MRI markers that quantify accelerated aging. Whether OCT measures relate to brain aging estimates in MS remains unexplored. In 89 patients with MS, BA was estimated from 3T MRI using brainageR package (Gaussian Process Regression). OCT-derived retinal layer volumes were obtained with Heidelberg SPECTRALIS. Spearman correlations assessed relationships between OCT measures and BA, BAG, clinical disability and cognitive outcomes (EDSS, SDMT, PASAT-3, and others), and MRI volumetric markers. Higher Brain-predicted age correlated with reduced temporal peripapillary retinal nerve fiber layer thickness (ppRNFL) (r = −0.419, p < .001), papillomacular bundle (PMB; r = −0.341, p = .001), macular RNFL volume (r = −0.263, p = .013), and ganglion cell-inner plexiform layer (GCIPL) volume (r = −0.275, p = .009). BAG showed equally robust associations with these OCT measures (all p < .01). Chronological age did not correlate with OCT measures. Temporal thickness and PMB thickness correlated with SDMT and PASAT-3 performance. Associations between OCT metrics and brain aging were further supported by correlations with T2 lesion volume, deep gray matter volume, and choroid plexus volume. OCT-derived retinal measures, specifically temporal ppRNFL thickness, PMB, and macular GCIPL volume are significantly associated with brain-predicted age and BAG in MS, independent of chronological age. These metrics serve as accessible noninvasive biomarkers that reflect the accelerated brain aging and the neurodegenerative pathology across clinical, cognitive, and structural MRI domains.