Multiple Cox regression analysis including the factors that showed significant associations to autoantibody positivity when analyzed univariately
| Chromosome | SNP | Gene | Autoantibody-positive subjects (n = 350) vs. autoantibody-negative subjects (n = 4,814)* | P |
|---|---|---|---|---|
| 1p13.2 | rs2476601 | PTPN22 | 1.54 (1.26–1.87) | <0.0001 |
| 1q31.2 | rs2816316 | RGS1 | 1.25 (1.04–1.51) | 0.0185 |
| 4p15.2 | rs10517086 | 0 | 1.22 (1.04–1.43) | 0.0142 |
| 7p12.1 | rs4948088 | COBL | 0.62 (0.40–0.97) | 0.0346 |
| 11p15.5 | rs1004446 | INS | 0.77 (0.65–0.90) | 0.0010 |
| 12q13.2 | rs2292239 | ERBB3 | 1.31 (1.12–1.52) | 0.0005 |
| 12q24.12 | rs3184504 | SH2B3 | 1.38 (1.19–1.61) | <0.0001 |
| 16p13.13 | rs12708716 | CLEC16A | 0.84 (0.71–0.98) | 0.0324 |
The analysis was corrected for HLA category, sex, and country for European sites as covariates. For U.S. sites, the principal components were estimated from our data from the U.S. population to be included in the model as a covariate for the two major principal components. The principal components were used to adjust for heterogeneity within the Caucasian population in the U.S. sites. The factors indicating nominal significant risk or protection are indicated in bold.
*HR (95% CI).