ABSTRACT Background and Aims Polycystic ovary syndrome (PCOS) is a prevalent endocrine and metabolic disorder with complex pathogenesis and limited targeted therapeutic options. This study presents a proteomics‐informed approach to identify potential plasma protein targets for PCOS intervention using causal inference methods. Methods We conducted a comprehensive proteome‐wide Mendelian randomization analysis by integrating 1 Mb plasma cis‐acting protein quantitative trait loci datasets with PCOS genome‐wide association study summary statistics. To enhance robustness, we applied complementary approaches including summary‐data‐based MR, the Heterogeneity in Dependent Instruments test, and colocalization analysis. Identified targets were further assessed for druggability using curated drug databases. Results The MR analysis revealed 33 plasma proteins significantly associated with PCOS risk. Among these, FGF23 and SH2B3 showed the strongest evidence of a causal role, supported by SMR, HEIDI, and colocalization analyses. FGF23 was positively associated with PCOS risk and implicated in inflammatory and metabolic pathways, while SH2B3 was inversely associated and linked to anti‐inflammatory signalling. Conclusions This study establishes a causal link between specific plasma proteins and PCOS and identifies FGF23 and SH2B3 as promising candidates for targeted drug development. These findings demonstrate the value of proteomics‐integrated genetic analyses in uncovering novel therapeutic avenues for complex diseases like PCOS.
Yu et al. (2025) studied this question.