The objectives of our study were to determine whether plasma GFAP was associated with amyloid and tau PET, temporal meta-ROI cortical thickness, white matter hyperintensities (WMH), cerebral microbleeds (CMBs), and cognition. In addition, we determined whether associations of GFAP and outcomes were more specific to those with elevated brain amyloid.

A subsample of 200 subjects from the population-based Mayo Clinic Study of Aging who underwent amyloid and tau positron emission tomography (PET), MRI and had plasma GFAP measured were included. Multiple linear regression and hurdle models were used to investigate associations controlling for age and sex. Additionally, we stratified by elevated brain amyloid status.

Higher plasma GFAP levels were associated with higher amyloid (regression coefficient b=0.1, p<.001) and tau PET burden (b=0.03, p=.008) after adjusting for age and sex. After adjusting for amyloid, the association with tau PET was no longer significant (b=0.01, p=.276). In the amyloid stratified analyses, GFAP was only associated with cortical thickness (b=-0.05, p=.047), WMH (b=0.2, p=.04) and lobar CMBs (OR = 2.354, p=.020) among those who were amyloid positive. There was no cross-sectional association between GFAP and global cognition after adjusting for age and sex (b=0.1, p=.93).

This cross-sectional analysis demonstrates the utility of plasma GFAP as a biomarker for AD-related pathologies, including amyloid PET burden.