Shared genetic architecture between gastric cancer and medication-use proxy phenotypes in East Asians highlights lipid-metabolic pathways and candidate cellular contexts

Gastric cancer (GC) remains a major cause of cancer-related mortality in East Asia, and metabolic and cardiovascular comorbidities are common in the same aging populations. However, the extent to which GC shares inherited genetic architecture with medication-use proxy phenotypes that reflect these clinical backgrounds remains incompletely defined. We sought to characterize this genetic overlap using large-scale East Asian genome-wide association study summary statistics. We analyzed germline GWAS summary statistics for GC and 20 medication-use proxy phenotypes from BioBank Japan. These proxy phenotypes were treated as composite traits reflecting underlying disease liability, treatment indication, prescribing patterns, and healthcare utilization rather than direct pharmacological exposure. Genome-wide and local genetic correlations were evaluated, followed by cross-trait association testing, gene-level prioritization, quantitative trait locus integration, and single-cell transcriptomic contextualization. GC showed significant genetic correlations with nine medication-use proxy phenotypes related mainly to metabolic and cardiovascular disease backgrounds. Cross-trait analyses identified 19 independent genomic regions and prioritized 13 candidate genes, including HMGCR , PSCA , and ARC . Functional enrichment analyses implicated cholesterol metabolism, lipoprotein transport, and related lipid-metabolic pathways. Integration with gastric cancer single-cell RNA-sequencing data suggested that cross-trait signals from the HMG-CoA reductase inhibitor-use proxy phenotype showed statistical enrichment in proliferative tumor cells, whereas signals from the antidiabetic medication-use proxy phenotype showed statistical enrichment in fibroblasts. Targeted in vitro follow-up provided independent cellular observations for HMGCR and ARC : HMGCR overexpression was associated with increased MKN45 proliferation, whereas ARC knockdown in CAFs reduced GC cell migration in a co-culture model. These experiments were not used as validation of the cross-trait genetic signals. GC shares inherited genetic architecture with medication-use proxy phenotypes that partly reflect metabolic and cardiovascular disease backgrounds in East Asians. The results prioritize lipid-metabolic pathways and candidate cellular contexts for future investigation, while not establishing medication effects or causal mechanisms.