A pan-matrix meta-analysis of microRNAs signature in Alzheimer's disease and mild cognitive impairment

Effective treatments for Alzheimer’s disease (AD) are still lacking and early and precise diagnosis is crucial for the development of successful therapies. MicroRNAs (miRNAs) have emerged as potential biomarkers and therapeutic targets for AD. This meta-analysis is the first to combine altered miRNA expression in AD and mild cognitive impairment (MCI) across a wide range of biological matrices. A comprehensive search of PubMed and Embase was conducted to identify relevant studies investigating miRNA expression in AD or MCI patients compared to healthy controls (AD/HC and MCI/HC) in brain tissue, plasma, cerebrospinal fluid, blood, serum, and extracellular vesicles. Data on miRNA expression levels were extracted, and a meta-analysis was performed using P-value meta-analysis. Results obtained from 224 data sets across 151 independent studies were quantitatively assessed in 655 meta-analyses. In the AD/HC comparison, 356 miRNAs showed differential expression across these matrices. The MCI/HC comparison revealed 16 differentially expressed miRNAs. The most reproducible miRNA alterations converged on biological processes relevant to AD pathophysiology, including β-amyloid protein (Aβ) metabolism, tau pathology, neuroinflammation, apoptosis/cell death, immune response, lipid metabolism, epithelial-to-mesenchymal transition, and adipogenesis. By preserving brain tissue, CSF, plasma, serum, whole blood, and extracellular vesicles as distinct matrices, this pan-matrix approach highlights both concordant and matrix-specific miRNA signatures that may inform future biomarker discovery and disease-stage characterization.