In this study, we demonstrated a high degree of genetic correlation between differing diagnostic and frequency-based case definitions of OA and across different types of controls (opioid exposed, unexposed, and public controls), which allowed us to conduct the GENOA meta-analysis and apply gSEM successfully to GENOA and existing summary statistics to conduct the largest GWAS among European ancestry cohort participants to date (23,367 cases and total effective sample size of 88,114 individuals) with the OA case definition. The GENOA GWAS of European ancestry identified a single GWS association (rs28386916), but the variant was not available in the MVP or PH cohorts and, consequently, was not tested for replication and was not present in the gSEM GWAS. Given the position of this variant between simple repeats and that variants in high LD with it were not significant in the gSEM GWAS, it seems likely that it represents a false positive. In the gSEM GWAS we found the strongest statistical evidence to date linking variants in intron 1 of the OPRM1 gene to OA, extending previous candidate gene studies focused on this
