Our GWAS analyses found that CNIH3 SNPs are associated with protection against progression to ODE in OUIP individuals. Since this effect was observed in analyses limited to opioid misusers, it likely represents liability unrelated to that for initiation of opioid misuse. The discovery and confirmation sample OUIP groups have substantially higher RAFs than an Australian (EA) general population sample59 (Supplementary Table 8) to which the ODE groups’ RAFs are more similar. Post-hoc SNP-based association analyses comparing the CATS OUIP group and this general population sample found substantial differences (p values ≤ 6.4E-5); similar comparisons with the CATS ODE group found more modest differences (p≥2.3E-2) in the opposite direction. The phenotypic variance in ODE explained by rs10799590 in our meta-analysis is estimated at 1.17% to 5.85% (Supplementary Table 9), indicative of a strong effect. Overall, our findings suggest that these CNIH3 SNPs enable greater, but not complete control in the use of these otherwise highly addictive drugs.
