Table 5.
Summary of polygenic risk score (PRS) analyses. PRS analysis in FMD using known AAA SNPs based on FMD GWAS samples, and PRS analysis in AAA using known FMD, CeAD, or SCAD SNPs based on Million Veteran Program (MVP) AAA GWAS samples. PRS tests were constructed based upon logistic regression of individual genotypes, adjusting for age, sex (if applicable) and the first five PCs, except where indicated by “GTX,” the PRS was derivative of summary statistics using the GTX program. Both logistic regression and GTX methods yielded similar PRS association results. The reported PRS p-values were un-adjusted, and when meeting the Bonferroni-corrected significance threshold (0.05/10=0.005) the result is shown in bold font. For replication purposes, the association of PRSFMD and AAA trait was tested in the UK Biobank (UKB) cohort. The replication p-value thresohold after multiple test correction was 0.05/1=0.05. Despite of the smaller sample size and limited power, the replication in UKB was positive (p-value=0.04) and displayed a similar odds ratio as that in the MVP analysis.
| Phenotype | Cohort | PRS* | Total N | N cases | N controls | PRS β | PRS SE |
PRS OR[95%CI] | PRS P- value |
|---|---|---|---|---|---|---|---|---|---|
| PRSAAA association with FMD | |||||||||
| FMD | UM | PRS AAA † | 4,011 | 573 | 3,438 | 0.40 | 0.11 | 1.50[1.20-1.87] | 3.07E-04 |
| FMD | UM | PRS AAA- GTX | 7,777 | 584 | 7,193 | 0.43 | 0.11 | 1.53[1.24-1.90] | 9.04E-05 |
| PRSFMD association with AAA in the MVP | |||||||||
| AAA-all | MVP | PRS FMD | 303,742 | 9,693 | 294,049 | 0.03 | 0.01 | 1.03[1.01-1.05] | 2.62E-03 |
| AAA-male | MVP | PRS FMD | 282,391 | 9,633 | 272,758 | 0.03 | 0.01 | 1.03[1.01-1.05] | 2.11E-03 |
| AAA-female | MVP | PRSFMD | 21,351 | 60 | 21,291 | −0.10 | 0.13 | 0.91[0.70-1.18] | 0.48 |
| PRSSCAD association with AAA in the MVP | |||||||||
| AAA-all | MVP | PRS SCAD | 303,742 | 9,693 | 294,049 | 0.03 | 0.01 | 1.03[1.01-1.05] | 2.46E-03 |
| AAA-male | MVP | PRS SCAD | 282,391 | 9,633 | 272,758 | 0.03 | 0.01 | 1.03[1.01-1.05] | 2.01E-03 |
| AAA-female | MVP | PRSSCAD | 21,351 | 60 | 21,291 | −0.08 | 0.13 | 0.92[0.71-1.19] | 0.53 |
| PRSCeAD association with AAA in the MVP | |||||||||
| AAA-all | MVP | PRS CeAD | 303,742 | 9,693 | 294,049 | 0.03 | 0.01 | 1.04[1.01-1.06] | 8.39E-04 |
| AAA-male | MVP | PRS CeAD | 282,391 | 9,633 | 272,758 | 0.03 | 0.01 | 1.03[1.01-1.06] | 1.01E-03 |
| AAA-female | MVP | PRSCeAD | 21,351 | 60 | 21,291 | 0.09 | 0.13 | 1.10[0.86-1.41] | 0.46 |
| Replication analysis: PRSFMD association with AAA in UKB | |||||||||
| AAA-all | UKB | PRS FMD | 486,906 | 998 | 485,908 | 0.06 | 0.03 | 1.07[1.00-1.13] | 0.04 |
Weighted PRS scores
Logistic regression PRS analysis of 573 FMD cases versus matched 3,438 MGI controls (1:6 case control ratio, after removing close relatives), adjusted for age, sex, PC1-PC5; (see eFigure 3 for PRS distribution plot).