Parkinson’s disease involves neurodegeneration and astroglial activation. GFAP, an intermediate filament protein expressed by astrocytes, reflects neuroinflammatory activity. Elevated GFAP levels have been linked to cognitive impairment and faster disease progression in PD, yet results across studies have been inconsistent. To clarify these findings, we conducted a systematic review and meta-analysis synthesizing available longitudinal evidence.

We systematically searched PubMed, Embase, Scopus, and Web of Science through October 2025, following PRISMA 2020 guidelines. Observational cohort studies reporting associations between baseline plasma or CSF GFAP levels and subsequent clinical outcomes in PD were included. Two reviewers independently extracted data and assessed quality using the Newcastle-Ottawa Scale. A random-effects meta-analysis estimated pooled standardized mean differences (SMDs) or hazard ratios (HRs) for progression outcomes. Heterogeneity was evaluated using I², and the certainty of evidence was graded using GRADE criteria.

 From 463 identified studies, eight met eligibility criteria, encompassing approximately 2,600 PD patients. Higher baseline GFAP levels were associated with cognitive decline (pooled SMD = 0.41; 95% CI 0.20–0.62; I² = 66%) and faster motor progression (SMD = 0.33; 95% CI 0.08–0.58). Elevated GFAP predicted conversion from mild cognitive impairment to dementia (HR = 2.7; 95% CI 1.9–3.5). Stronger associations were observed for CSF versus plasma GFAP and for follow-up periods ≥ 4 years.

Higher GFAP levels were associated with worse motor and cognitive outcomes in PD, supporting its potential role as a prognostic biomarker of disease progression. Larger longitudinal studies using standardized assays are needed to confirm its clinical applicability.