Spatiotemporal Analysis of Ferroptosis-Related Biomarker Alterations in Alzheimer's Disease and Mild Cognitive Impairment: A Meta-Analysis

Ferroptosis, an iron-dependent regulated cell death driven by lipid peroxidation and antioxidant system failure, has been implicated in Alzheimer's disease (AD), yet its spatial and temporal patterns across brain regions and disease stages remain unclear. Objective: To quantitatively synthesize human post-mortem and clinical brain-tissue studies comparing ferroptosis-related biomarkers among AD, mild cognitive impairment (MCI), and control groups in the frontal lobe, temporal cortex, parietal lobe, and hippocampus, and to explore the spatiotemporal distribution of these alterations. Methods: Following PRISMA guidelines, we systematically searched PubMed and Web of Science, included 35 studies, extracted effect sizes (standardised mean differences, SMD) for lipid peroxidation, antioxidant, and iron-homeostasis markers, and performed random-effects meta-analyses with subgrouping by brain region and disease stage, with heterogeneity assessed via I2 and Q statistics, alongside sensitivity and publication-bias analyses. Results: Frontal lobe AD vs. controls showed a pooled SMD of 0.547 (95% CI: −0.106 to 1.200, p=0.092, I2=86.9%); MCI vs. controls SMD=1.130 (p=0.108, I2=53.8%); AD vs. MCI SMD=0.638 (p=0.120, I2≈0). Temporal cortex AD vs. controls SMD=−0.024 (p=0.952, I2=96.1%). Hippocampus AD vs. controls SMD=1.121 (p=0.066, I2=88.9%). Parietal lobe SMD=0.684 (p=0.343, I2=60.1%). No overall comparison reached statistical significance, but directionally, the hippocampus and frontal lobe showed relatively larger effect sizes; sensitivity analyses indicated robustness in direction but high heterogeneity partly driven by individual studies, and publication bias was not evident. Conclusions: Ferroptosis-related biomarker alterations in AD and MCI exhibit regional selectivity, with the hippocampus and frontal lobe showing more prominent trends, while temporal cortex findings are highly inconsistent; changes may already emerge at the MCI stage, suggesting early involvement. However, current evidence is insufficient to determine the precise initial site or temporal sequence due to high heterogeneity, limited ferroptosis-specific markers, and scarcity of staging data, and future studies with standardised methods and Braak-stage matching are needed.