Blood GFAP in Cerebral Small Vessel Disease and Vascular Cognitive Impairment: A Systematic Review of Cognitive Associations, Diagnostic Performance, and Prognostic Evidence

Background and Objectives: Blood glial fibrillary acidic protein (GFAP) may reflect astrocytic injury in cerebral small vessel disease (CSVD) and vascular cognitive impairment (VCI). This review examined its associations with vascular cognitive phenotypes, cognitive performance, and prognosis. Materials and Methods: The protocol was registered on the Open Science Framework (OSF). PubMed and Europe PMC were searched from inception through 4 October 2026. Two reviewers independently screened records and extracted data. Eligible observational reports measured blood GFAP in adults with independently defined sporadic CSVD or VCI, or evaluated incident vascular dementia. Narrative synthesis was supplemented by an exploratory random-effects meta-analysis (restricted maximum likelihood, modified Hartung–Knapp intervals) of unadjusted concentration differences, with sensitivity analyses by matrix, assay platform, leave-one-out, and methodological appraisal. Results: Twenty-one reports (2022–2026) were included: 17 contributed concentration contrasts, 7 cross-sectional cognitive associations, 5 diagnostic accuracy estimates, 3 longitudinal cognitive analyses, and 2 incident vascular dementia analyses. Five independent cohorts with compatible mean-and-standard-deviation data contributed 644 participants to the concentration meta-analysis. GFAP was higher in vascular groups than in cognitively unimpaired controls (Hedges g, 0.66; 95% confidence interval, 0.43–0.90; I2 = 0%; τ2 = 0; sensitivity estimates, 0.59–0.75). Cross-sectional cognitive associations varied across populations. Longitudinal associations were reported in established VCI and high-white-matter-hyperintensity subgroups, but differed by cognitive domain and Alzheimer biomarker status. Two population cohorts reported associations with incident vascular dementia on non-equivalent exposure scales. Diagnostic accuracy varied by comparator and covariate adjustment. Six further candidate reports could not be retrieved. Conclusions: Blood GFAP is associated with several vascular cognitive phenotypes, but the signal is not specific to vascular injury, and assay differences, mixed pathology, and confounding limit clinical interpretation. Current evidence supports a possible complementary role for GFAP within multimodal assessment; this depends on external validation, analytical standardization, and demonstration of incremental value over clinical and neuroimaging variables.