Abstract
BACKGROUND: Osteosarcopenia (OS) significantly elevates fracture risk in the elderly due to concurrent skeletal and muscular metabolic decline. Mitochondrial autophagy dysregulation is a significant mechanism driving myoskeletal metabolic imbalance, and myofibroblast-derived factor 5 (FNDC5)/irisin is a crucial mediator in the myoskeletal dialogue. It is unclear if the medicinal ingredient icariin (ICA) from the traditional Chinese herb Epimedium can work in concert to prevent and treat OS through this pathway, despite research suggesting that exercise may enhance muscular performance by modulating the FNDC5–PINK1/Parkin pathway. METHODS: With the use of dexamethasone (DEX) and ovariectomy (OVX), a female SD rat OS model was created. Low, medium, and high dosages (40, 80, and 120 mg/kg/d) of ICA were then administered. Micro-computed tomography (Micro-CT), forelimb grip strength tests, and dual-energy X-ray absorptiometry (DXA) were used to evaluate bone microarchitecture, muscle performance, and body composition, respectively. Using hematoxylin and eosin (HE) staining, the morphology of muscle tissue was investigated. In skeletal muscle, Western blotting revealed the expression of FNDC5 and mitochondrial autophagy-related proteins (PINK1, Parkin, LC3, p62). ELISA was used to evaluate the serum levels of FNDC5, muscle growth inhibitor (GDF-8), and inflammatory markers interleukin-6 (IL-6) and C-reactive protein (CRP). To directly verify the causal role of the PINK1/Parkin pathway in mediating ICA’s effects, an in vitro experiment using the mitochondrial autophagy inhibitor Mdivi-1 was performed alongside the CCCP-induced mitophagy model in C2C12 myoblasts. We used Bayesian co-localization, HEIDI heterogeneity testing, and summary-data-based Mendelian randomization (SMR) to further integrate expression quantitative trait locus (eQTL) data with genome-wide association study (GWAS) data from eQTLGen and IEU Consortium in order to genetically validate the causal relationship between FNDC5 and OS. RESULTS: DEX + OVX effectively produced a classic OS phenotype in rats, which was marked by a marked increase in fat percentage (FAT) (p < 0.01) and a significant decrease in bone mineral density (whole body, femur, spine), bone mineral content (BMC), muscle mass, muscle mass index (SMI), and grip strength (p < 0.01). These alterations were significantly reversed by ICA intervention, especially in the medium and high-dose groups. It increased bone mineral density and BMC (p < 0.001), improved bone microstructural parameters (increased BV/TV, Tb.Th, and Tb.N; decreased Tb.Sp), enhanced muscle mass and SMI, enlarged muscle fiber cross-sectional area, and reduced fibrosis and fat infiltration. ICA also significantly reduced serum GDF-8, CRP, and IL-6 levels (p < 0.0001) while increasing FNDC5 concentration (p < 0.0001). Mechanistic studies revealed that ICA suppressed excessive mitochondrial autophagy in muscle tissue (downregulating PINK1 and Parkin proteins) while upregulating FNDC5 expression (p < 0.0001) and promoting the expression of osteogenic-related proteins Runx2 and OCN. Additionally, in vitro tests demonstrated that ICA (300 µM) reversed the inhibition of FNDC5 expression (p < 0.01), prevented excessive activation of the PINK1/Parkin pathway (p < 0.01), and mitigated the drop in mitochondrial membrane potential caused by CCCP. Crucially, Mdivi-1 suppression of the PINK1/Parkin pathway replicated the effects of ICA, resulting in elevated expression of FNDC5. Additionally, co-treatment with ICA and Mdivi-1 did not result in an additive impact on FNDC5 overexpression, suggesting that ICA modulates FNDC5 via the PINK1/Parkin pathway. Its preventive causative function was supported by SMR analysis, which showed that genetically predicted greater FNDC5 levels were strongly linked with decreased OS risk. CONCLUSIONS: This study suggests that ICA improves OS in association with inhibition of pathological mitochondrial autophagy and modulation of the PINK1/Parkin–FNDC5 pathway, providing a basis for further mechanistic studies. Genetic evidence further supports FNDC5 as a potential therapeutic target for OS. These findings provide preclinical evidence for ICA’s potential in postmenopausal osteosarcopenia, though applicability to age-related, male, or disuse-related forms requires validation. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12967-026-08017-0.