Abstract
BACKGROUND: Various plasma phosphorylated tau species have been shown to be associated with amyloid-β (Aβ) PET and Tau PET in Alzheimer’s disease (AD), but whether APOE ε4 affects the interaction between glial fibrillary acidic protein (GFAP) and phosphorylated tau (pTau), and whether a three-way interaction exists among APOE ε4, GFAP, and pTau that influences AD progression remain unclear. METHODS: The study included 563 participants from the Chinese Preclinical Alzheimer’s Disease Study (CPAS) and 243 from Alzheimer’s Disease Neuroimaging Initiative (ADNI), all of whom underwent Aβ PET, magnetic resonance imaging (MRI), neuropsychological assessments, and plasma biomarker analyses (GFAP, pTau181, pTau231, pTau217), with subsets undergoing Tau PET. The longitudinal data of 101 participants from ADNI were additionally included. We employed linear regression models with interaction terms to examine how APOE ε4 status and plasma GFAP levels modulate the relationships between plasma pTau biomarkers and AD pathology cross-sectionally and longitudinally. RESULTS: Plasma GFAP and pTau biomarkers (pTau181, pTau231, pTau217) are significantly elevated in Aβ-positive individuals, with stronger Aβ–pTau associations observed in APOE ε4 carriers (CPAS: β = 0.26, p = 0.003 for pTau231; ADNI: β = 0.45, p < 0.001 for pTau181). Across two cohorts, plasma GFAP levels significantly strengthened the associations between pTau biomarkers and Tau PET. Furthermore, subsequent analyses revealed that this modulatory effect of GFAP on the links between pTau and PET-derived pathological changes was more pronounced in APOE ε4 non-carriers, whereas in APOE ε4 carriers, a significant interaction between GFAP and pTau was only observed in specific Braak stage-specific regions within the CPAS cohort. In longitudinal analyses, we also observed stronger pTau181-associated longitudinal tau accumulation in individuals with high GFAP levels (Braak III-IV). CONCLUSION: We demonstrate that APOE ε4 status critically modulates the relationship between pTau and Aβ pathology, whereas plasma GFAP primarily influences pTau–tau pathology associations, particularly in individuals without APOE ε4 allele. These findings underscore the role of reactive astrogliosis in tau propagation and support the utility of plasma biomarkers for AD diagnosis and prognosis. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13024-025-00906-0.