HAI Book 2025 - Flipbook - Page 338
Bilgel, Murat
71
Longitudinal change in five dominant dimensions of brain aging in
relation to brain amyloid status and Alzheimer9s-related plasma biomarkers
Murat Bilgel1, Ishaan Shah1, Jasmine Cooper1,2, Yang An1, Keenan Walker1, Abhay Moghekar3,
Zhijian Yang4, Guray Erus4, Christos Davatzikos4, Susan Resnick1
1
Laboratory of Behavioral Neuroscience, National Institute on Aging, Baltimore, MD, US
Department of Psychology, University of Michigan, Ann Arbor, MI, US
3
Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD, US
4
Artificial Intelligence in Biomedical Imaging Laboratory (AIBIL), Center for and Data Science for Integrated
Diagnostics (AI2D), Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, US
2
Introduction: Plasma biomarkers are associated with regional brain volumes, but it remains unclear if these
associations are driven by specific patterns of neurodegeneration among cognitively unimpaired individuals.
Methods: We used data for 179 participants from the Baltimore Longitudinal Study of Aging who were cognitively
normal at the time of their plasma A´42/A´40, GFAP, NfL, and p-tau181 measurements with Quanterix Simoa
assays. All participants had undergone at least one assessment with 11C-Pittsburgh Compound B (PiB), allowing for
the ascertainment of their amyloid status at the time of plasma biomarker measurements. Structural MRIs
concurrent with and following plasma biomarker measurements (707 MRI visits) were used to calculate R-indices
describing five dominant dimensions of brain aging: subcortical atrophy (R1), mediotemporal lobe atrophy (R2),
parieto-temporal atrophy (R3), diffuse cortical atrophy (R4), and perisylvian atrophy (R5) (Figure 1). Using linear
mixed effects models adjusted for baseline age, age2, sex, age × sex, race, and estimated glomerular filtration rate
(eGFR), we examined if baseline plasma biomarkers are associated with longitudinal change in R-indices and
whether these associations differ by PiB group.
Results: There were statistically significant differences between PiB3 and PiB+ individuals in associations of
baseline plasma biomarkers with longitudinal change in R-indices (i.e., PiB group × baseline plasma biomarker ×
time interaction): among PiB+ vs. PiB3 individuals, higher GFAP and NfL at baseline were associated with steeper
longitudinal increases in R2 and R3, whereas lower A´42/A´40 was associated with steeper increases in R2 and
R5 (Table, Figure 2).
Discussion: Among individuals with elevated brain amyloid, plasma A´42/A´40, GFAP, and NfL may be useful for
predicting future mediotemporal lobe, parieto-temporal, and perisylvian atrophy, suggesting that these plasma
biomarkers may be able to capture impending Alzheimer9s-related patterns of neurodegeneration.
Acknowledgments: This study was supported by the National Institute on Aging Intramural Research Program,
National Institutes of Health.
HAI2025 - 338
