Speaker
Description
Brain atrophy is a sign of neuropathological aging, and it is associated with multiple neurodegenerative disorders, including Alzheimer's disease, Schizophrenia, Parkinson’s disease, and Multiple Sclerosis. Additionally, Alzheimer’s Disease (AD) related plasma biomarkers can be used as a proxy to measure neurodegeneration and assess AD related changes. To better understand the link between trajectories of brain atrophy and accumulation of AD plasma biomarkers, we used a sample of 639 participants from the Baltimore Longitudinal Study of Aging who at baseline, were cognitively normal, and had both MRI scans and plasma biomarkers. We generated five scores for different trajectories of brain atrophy, based on the Surreal-GAN framework, a deep learning approach that explores variations in brain aging. AD plasma biomarkers of interest included GFAP, Ptau-181, NfL, and Amyloid-Beta ratio (A𝛽42/A𝛽40). After running linear mixed effects models, we found that higher baseline levels of GFAP were associated with a faster rate of atrophy in earlier changing medial temporal regions, which corresponds to the transition from being cognitively normal to showing signs of Mild Cognitive Impairment. Next, we found that higher GFAP, lower A𝛽42/A𝛽40, and higher ptau181 were associated with faster rate of parietotemporal atrophy, which corresponds to the transition from Mild Cognitive Impairment to Alzheimer’s Disease. Importantly, we found an age modifying effect of Ptau-181 on parietal temporal atrophy, where those who showed signs of Ptau-181 accumulation earlier in older adulthood showed more advanced disease progression than their older counterparts. These results have implications for using AD plasma biomarkers for predicting and monitoring specific patterns of brain atrophy.
