AbstractBackgroundGlial fibrillary acidic protein (GFAP) is an astrocytic marker that has been recently associated with fluid and imaging biomarkers of amyloid and tau pathologies in the Alzheimer’s disease (AD) spectrum. Plasma GFAP has also shown to be a strong indicator of early brain amyloidosis in preclinical AD. However, it remains to be confirmed how plasma GFAP associates with neuropathological hallmarks of AD and which pathological entities are being reflected by this plasmatic marker.MethodPlasma GFAP concentrations were measured using a commercial Simoa assay on participants of the UCSD Shiley‐Marcos Alzheimer’s Disease Research Center (ADRC), consisting of 271 participants with brain autopsy data. Amyloid and tau pathologies were assessed following the CERAD, Braak and NIA‐Reagan criteria, as well as with tangle and plaque counts (147 participants). Cerebral amyloid angiopathy (CAA) was graded as recommended by NACC Neuropathology Working Group. Non‐AD pathologies were also assessed: Lewy body disease, hippocampal sclerosis, TDP‐43 and vascular pathology. The distribution of GFAP concentration was compared across groups using ANCOVA models adjusting for age and sex, followed by post‐hoc Tukey’s HSD test. The association between GFAP and the pathological indexes was tested using linear models and adjusted by covariates.ResultConcentrations of plasma GFAP were higher in females as compared to males (P=0.008) and no age effect was observed. GFAP also increased with severity of AD pathological stages (P<0.001), with increased neuritic (P<0.001) and diffuse plaque density or counts (P=0.001), as well as with Braak stages and tangle counts (P<0.05). We found no association between tangle counts and GFAP when accounting for the effect of plaque counts. The presence of co‐pathologies did not affect average GFAP levels. We found no association between GFAP and CAA.ConclusionPlasma GFAP showed again here to highly reflect amyloid deposition. Results indicate, however, that this association is due to the accumulation of extra‐cellular amyloid in the form of plaques rather than its deposition in blood vessels. Abnormal accumulation of non‐AD proteins did not show any marked effect on GFAP levels. More detailed analysis, including longitudinal retrospective data, is ongoing.