Abstract
Background: Osteoarthritis (OA) is a prevalent musculoskeletal disorder causing chronic pain and disability, particularly in older adults. It is a multifactorial disease characterized by joint degeneration, with varying pathophysiological mechanisms across different OA subtypes (knee, hip, spine, hand, etc.). This study aimed to explore the genetic mechanisms underlying various OA subtypes using a novel approach combining protein level ratios (rQTLs) with Mendelian Randomization (MR) analysis. Method: We utilized publicly available Genome-Wide Association Study (GWAS) datasets on rQTLs as exposure variables and OA at various anatomical sites as outcome variables. The study involved conventional multi-related-SNP MR analyses, top-related-SNP MR analyses, advanced Bayesian MR analyses, sensitivity analyses and experiments to validate findings. Results: Key findings include significant associations between specific rQTLs and hip OA, such as DNMBP/FKBP5 and MME-related ratios, indicating their potential role in disease pathogenesis. For knee OA, rQTLs like INPP1/MPI were associated with increased risk, while FABP5/PPCDC and LYN/TACC3 were associated with reduced risk. In contrast, most rQTLs showed minimal influence on spine OA, hand OA, finger OA, and thumb OA. Advanced Bayesian MR analyses, sensitivity analyses and experiments confirmed a significant causal effect of the DNMBP/FKBP5 ratio on hip OA risk. Conclusions: This study provides new insights into the genetic and molecular mechanisms of OA subtypes, highlighting potential therapeutic targets. The integration of protein ratio GWAS with network MR offers a comprehensive approach to understanding the complex pathogenesis of OA and emphasizes the need for subtype-specific therapeutic strategies.