Reliable, minimally invasive biomarkers for Alzheimer’s disease (AD) remain a critical unmet need. Considering mounting evidence for gut–brain interactions in AD, we explored whether fecal samples contain host-derived molecular signals that reflect disease-related changes and could serve as accessible biomarkers. Fecal small RNA sequencing and quantitative proteomics were performed in symptomatic 15-month-old 3×Tg-AD mice and controls. Pathway and network analyses identified key dysregulated molecules. Selected miRNAs and proteins were validated by qPCR or Western blotting in independent pre-symptomatic, early symptomatic, and severe cohorts. To assess whether fecal signals mirrored tissue alterations, colon samples were analyzed for tight-junction proteins, inflammatory markers, epigenetic regulation, and expression of selected miRNAs and proteins. Multi-omics identified 31 miRNAs and 81 proteins modulated in AD fecal sample, revealing a distinct AD-associated fecal signature involving pathways linked to neuroinflammation, synaptic dysfunction, vascular imbalance, and metabolism. Several miRNAs— including miR-322-5p, miR-194-1 and miR-223-3p were significantly altered and validated across disease stages (3-,7-,15-old months mice, respectively). Notably, miR-146b-5p expression is inversely correlated with AD severity, exhibiting a stepwise decline from asymptomatic individuals to those with mild and severe stages. IgKappa and Ela3B emerged as the strongest protein candidates, showing opposite and progressive modulation. Notably, IgKappa and miR-146b showed the most consistent trajectories. Colon analyses uncovered reduced Claudin-7, strong IgKappa upregulation driven by enhancer hypomethylation, and partial overlap between fecal and tissue miRNA profiles, indicating that fecal signals reflect subtle epithelial and immune perturbations associated with AD. Multi-omics at severe stage and a stage specific validation support fecal miRNAs and proteins as promising, host-derived biomarkers for AD. Fecal profiling offers a practical, repeatable, and low-cost approach with translational potential for early detection and longitudinal monitoring in neurodegeneration. © The Author(s) 2026.

Cited by: 0; All Open Access; Green Open Access; Hybrid Gold Open Access