Table 4.
Effect of polygenic scores on clinical conversion obtained from Cox proportional hazards models
| Sample size (N) | Score | Beta | SE | P-value |
|---|---|---|---|---|
| 1092 | PHS (status) | 0.290 | 0.09 | 0.002 |
| PHS (burden) | 0.372 | 0.15 | 0.013 | |
| APOE-ε44 | 0.295 | 0.17 | 0.088 | |
| 813 | PHS (burden) | −0.005 | 0.20 | 0.99 |
| PRS1 | −0.029 | 0.15 | 0.84 | |
| PRS2 | −0.016 | 0.08 | 0.84 | |
| PRS2* | 0.290 | 0.11 | 0.012 | |
| PRS-cs | −0.181 | 0.12 | 0.11 |
Clinical conversion was defined as a dementia diagnosis in previously not demented participants. Rows correspond to different polygenic scores and estimates for the full dataset (top) and after removal of ADGC subjects (bottom). Beta = log hazards ratio; SE = standard error. *was additionally adjusted for genetic population structure.