Cross-Phenotype Genomics Reveals Heterogeneous Genetic Susceptibility Architecture Relevant to Menopause

Background: Menopause is accompanied by substantial interindividual variation in hormonal, metabolic, cardiovascular, skeletal, and thrombotic phenotypes. Although the genetics of reproductive ageing and individual menopause-associated traits are increasingly characterised, the cross-phenotype architecture of individual menopause-relevant germline variants remains less well defined. Methods: We performed a variant-centred genomic analysis of 31 prespecified menopause-relevant germline variants. Cross-trait evidence comprised 766 variant-trait records across 52 phenotypes and was integrated with a targeted Pan-UK Biobank phenome-wide association study (PheWAS; 1462 tests across 57 endpoints), clinically relevant outcome analyses, linkage disequilibrium assessment, FinnGen R12 venous thromboembolism (VTE) associations, regional fine-mapping, and molecular annotation. Association strength, between-study heterogeneity, phenotypic breadth and sharing, and locus resolution were evaluated as distinct evidence dimensions. Results: In total, 25 of 31 variants showed at least one cross-trait association at p < 1 × 10−5 and 21 reached genome-wide significance. Phenotypic breadth varied markedly, from restricted profiles to 16 associated phenotypes, and strong associations frequently coexisted with substantial between-study heterogeneity. In Pan-UK Biobank, 222 associations involving 25 variants met false-discovery-rate correction, spanning hormonal, reproductive, cardiometabolic, skeletal, nutritional, hepatic, and neurological phenotypes. Cross-phenotype reconstruction identified recurrent connectivity through sex hormone-binding globulin, testosterone, adiposity, lipid, and skeletal traits, while linkage disequilibrium analyses distinguished phenotypic sharing from genetic redundancy. Clinically, F5 rs6025 and F2 rs1799963 showed strong VTE associations in FinnGen (OR 2.387 and 1.921, respectively), but markedly different locus resolution: rs6025 had posterior inclusion probability (PIP) 0.979 and constituted a single-variant 95% credible set, whereas rs1799963 had PIP 0.596 within a four-variant credible set. Conclusions: Menopause-relevant germline variation exhibits heterogeneous cross-phenotype architectures in which association strength, phenotypic breadth, between-study consistency, genetic independence, and locus resolution are not interchangeable. This variant-centred framework provides a basis for evaluating inherited susceptibility relevant to menopause and hormone exposure, while prospective individual-level studies remain necessary before translation into risk prediction or treatment selection.