Plasma glial fibrillary acidic protein as a potential biomarker for differentiating amyloid-negative subjective cognitive decline and mild cognitive impairment

Abstract

Background: Differentiating subjective cognitive decline (SCD) from mild cognitive impairment (MCI) in amyloid-negative (A-) individuals is clinically relevant for early evaluation along the Alzheimer's disease (AD) pathway. Objective: To assess whether plasma biomarkers discriminate between A-SCD and A-MCI and compare their performance. Methods: We studied 114 amyloid PET-negative participants (93 A-SCD, 21 A-MCI) aged >= 60 years from multicenter cohorts. Plasma A beta(42)/A beta(40) ratio, glial fibrillary acidic protein (GFAP), neurofilament light chain (NFL), and phosphorylated tau 181 (pTau181) were measured using Simoa. Associations with MCI were estimated with logistic regression; discrimination was assessed by ROC analysis with AUCs and pairwise AUC comparisons using 1000 bootstrap resamples. Results: Standardized log (GFAP) was associated with MCI in univariate analysis (OR 2.32; 95% CI 1.35-4.00; p = 0.002) and after adjustment for age and sex (OR 2.45; 1.27-4.74; p = 0.008). GFAP showed moderate discrimination (AUC 0.71; 0.59-0.82), whereas A beta(42)/A beta(40), NFL, and pTau181 did not. Pairwise AUC differences favored GFAP over A beta(42)/A beta(40) (Delta AUC 0.34; p = 0.020), NFL (0.21; p = 0.032), and pTau181 (0.29; p = 0.005). Conclusions: In this A- cohort, plasma GFAP was the most informative biomarker for distinguishing MCI from SCD, likely reflecting astrocytic activation. Given modest accuracy and limited disease specificity, GFAP may be most useful as part of a multimodal panel rather than a stand-alone test.