Skip to main content

Plasma GFAP outperforms CSF GFAP in detecting amyloid pathology and is associated with increased risk of clinical progression in early Alzheimer's disease.

The journal of prevention of Alzheimer's disease

Authors: Arda C Cetindag, Carola G Schipke, Hermann Esselmann, Niels Kruse, Jens Wiltfang, Anja Schneider, Klaus Fliessbach, Carolin Miklitz, Franziska Maier, Katharina Buerger, Daniel Janowitz, Michael Ewers, Sophia Stöcklein, Robert Perneczky, Boris-Stephan Rauchmann, Carolin Kurz, Stefan Teipel, Ingo Kilimann, Doreen Goerss, Christoph Laske, Sebastian Sodenkamp, Elham Najafpour, Michael Wagner, Sandra Roeske, Ingo Frommann, Melina Stark, Frederic Brosseron, Alfredo Ramirez, Luca Kleineidam, Josef Priller, Eike Jakob Spruth, Maria Gemenetzi, Slawek Altenstein, Emrah Düzel, Wenzel Glanz, Enise I Incesoy, Michaela Butryn, Chris Bauer, Frank Jessen, Oliver Peters

BACKGROUND: Early and accurate detection of Alzheimer's disease (AD) is essential for timely intervention and development of disease-modifying treatments. The DZNE-Longitudinal Cognitive Impairment and Dementia Study (DELCODE) provides a deeply phenotyped cohort covering preclinical and early clinical stages, including subjective cognitive decline (SCD) and mild cognitive impairment (MCI). Astrocyte reactivity and its biomarkers, particularly glial fibrillary acidic protein (GFAP), have gained increasing attention in AD research; however, the relationship between GFAP and amyloid in early disease, as well as its potential prognostic value beyond its association with amyloid status, remains insufficiently understood.

OBJECTIVES: To evaluate the performance of CSF and plasma GFAP across early disease stages, compare these measures according to amyloid status, and assess the prognostic value of GFAP for clinical progression across diagnostic stages during longitudinal follow-up.

SETTING: This study used data from the multicenter DELCODE cohort in Germany, including participants with available plasma and/or CSF samples and standardized clinical, cognitive, imaging, and biomarker assessments.

MEASUREMENTS: GFAP concentrations in plasma and CSF were quantified using validated immunoassay platforms. Standard CSF AD biomarkers and ApoE genotype were measured using established assays. Amyloid status was defined by the CSF Aβ42/40 ratio. Longitudinal follow-up occurred annually for up to ∼10 years, with clinical conversion determined according to NIA-AA criteria.

RESULTS: Plasma and CSF GFAP increased across the AD continuum, with higher levels in MCI and AD (p < 0.001). Plasma GFAP showed a stronger association with amyloid status than CSF GFAP across all groups. In MCI, plasma GFAP combined with age and ApoE4 yielded an AUC of 0.87. Elevated plasma GFAP predicted increased risk of conversion to MCI (HR = 2.19, p < 0.001; adjusted HR = 1.70, p = 0.0056) and AD dementia (HR = 3.5; adjusted HR = 2.49 both p < 0.001).

CONCLUSION: Plasma GFAP is a sensitive, minimally invasive biomarker with diagnostic relevance for amyloid detection and prognostic relevance for clinical progression in early AD.

Copyright © 2026 The Authors. Published by Elsevier Masson SAS.. All rights reserved.

PMID: 41905188

Participating cluster members