Associations of %HbF and PGSs with VOP event rate and occurrence in the SCCRIP and the SAC participants
| . | SCCRIP . | SAC (censored at 11 y) . | ||||
|---|---|---|---|---|---|---|
| EST . | SE . | P . | EST . | SE . | P . | |
| VOP event rate | ||||||
| %HbF* | −0.18 | 0.05 | 4.7 × 10−4 | ND | ND | ND |
| PGSHbF | 0.27 | 0.07 | 1.2 × 10−4 | 0.11 | 0.07 | .045 |
| PGSCOMT | 0.29 | 0.06 | 2.7 × 10−5 | 0.07 | 0.04 | .04 |
| PGS5snps | 0.60 | 0.09 | 5.8 × 10−10 | ND | ND | ND |
| PGSHbF+COMT | 0.27 | 0.05 | 3.9 × 10−8 | 0.1 | 0.04 | .06 |
| PGSHbF+COMT+5snps | 0.35 | 0.04 | 5.9 × 10−14 | ND | ND | ND |
| VOP event occurrence | ||||||
| %HbF* | −0.22 | 0.07 | 7.5 × 10−4 | ND | ND | ND |
| PGSHbF | 0.33 | 0.08 | 8.8 × 10−5 | 0.19 | 0.09 | .014 |
| PGSCOMT | 0.29 | 0.08 | 2.0 × 10−4 | 0.07 | 0.04 | .055 |
| PGS5snps | 0.73 | 0.12 | 3.2 × 10−9 | ND | ND | ND |
| PGSHbF+COMT | 0.32 | 0.06 | 1.0 × 10−7 | 0.08 | 0.05 | .065 |
| PGSHbF+COMT+5snps | 0.42 | 0.06 | 4.1 × 10−13 | ND | ND | ND |
| . | SCCRIP . | SAC (censored at 11 y) . | ||||
|---|---|---|---|---|---|---|
| EST . | SE . | P . | EST . | SE . | P . | |
| VOP event rate | ||||||
| %HbF* | −0.18 | 0.05 | 4.7 × 10−4 | ND | ND | ND |
| PGSHbF | 0.27 | 0.07 | 1.2 × 10−4 | 0.11 | 0.07 | .045 |
| PGSCOMT | 0.29 | 0.06 | 2.7 × 10−5 | 0.07 | 0.04 | .04 |
| PGS5snps | 0.60 | 0.09 | 5.8 × 10−10 | ND | ND | ND |
| PGSHbF+COMT | 0.27 | 0.05 | 3.9 × 10−8 | 0.1 | 0.04 | .06 |
| PGSHbF+COMT+5snps | 0.35 | 0.04 | 5.9 × 10−14 | ND | ND | ND |
| VOP event occurrence | ||||||
| %HbF* | −0.22 | 0.07 | 7.5 × 10−4 | ND | ND | ND |
| PGSHbF | 0.33 | 0.08 | 8.8 × 10−5 | 0.19 | 0.09 | .014 |
| PGSCOMT | 0.29 | 0.08 | 2.0 × 10−4 | 0.07 | 0.04 | .055 |
| PGS5snps | 0.73 | 0.12 | 3.2 × 10−9 | ND | ND | ND |
| PGSHbF+COMT | 0.32 | 0.06 | 1.0 × 10−7 | 0.08 | 0.05 | .065 |
| PGSHbF+COMT+5snps | 0.42 | 0.06 | 4.1 × 10−13 | ND | ND | ND |
PGSHbF, sum of low HbF–associated alleles across 11 SNPs in BCL11A, HBS1L-MYB, and the extended β-globin locus (supplemental Methods); PGSCOMT, sum of high pain risk alleles across 5 COMT SNPs: rs6269, rs4633, rs4818, rs4680, and rs165599; PGS5snps, sum of high pain risk alleles across 5 additional pain-related SNPs: rs6858735 (TBC1D1), rs2835914 (KCNJ6), rs2295632 (FAAH), rs2963155 (NR3C1), and rs1800587 (IL1A); PGSHbF+COMT+5snps, PGSHbF + PGSCOMT + PGS5snps. The PGSs were analyzed as continuous variables. Covariates for the SCCRIP analysis included sex, hydroxyurea therapy, 5 PCs, −α3.7, and PGS–age interaction (except for %HbF). The −α3.7 status was not available for the SAC. PGSCOMT–age interaction was not included in the SAC analysis because the Bayesian information criterion for the model with PGSCOMT–age interaction was higher. P values for the SCCRIP and SAC studies were determined by the Wald test calculated from a generalized linear mixed model. P values for the SAC validation cohort were 1-sided. Bold P values indicated significant associations (P < .05). EST, β estimates; ND, not determined; SE, standard error.
A total of 1095 longitudinal RBC %HbF values were available for 284 of 327 SCCRIP participants analyzed for VOP event rate and event occurrence.