Table 2.
Tests of association between DNA methylation at the 3 CpG sites of interest and maternal smoking exposure during gestation while controlling (or not) for the genotype of the methylQTL (linear regression tests) in 736 participants of the California Childhood Leukemia Study.
| Regression coefficient | Standard Error | t-value | P | ||
|---|---|---|---|---|---|
| GFI1 (cg18146737) | |||||
| model without genotype | −0.02 | 0.02 | −1.48 | 0.139 | |
| model with genotype at rs141819830 | −0.03 | 0.01 | −2.43 | 0.015 | * |
| AHRR (cg05575921) | |||||
| model without genotype | −0.04 | 0.01 | −6.40 | 2.9E-10 | *** |
| model with genotype at rs148405299 | −0.04 | 0.01 | −6.38 | 3.2E-10 | *** |
| MYO1G (cg12803068) | |||||
| model without genotype | 0.04 | 0.01 | 3.28 | 0.001 | ** |
| model with genotype at rs61087358 | 0.03 | 0.01 | 2.86 | 0.004 | ** |
†
Models were adjusted for: leukemia case/control status, ancestry, cell-mixture, gender, gestational age and DNA methylation array batch number.
*P < 0.05
**P < 0.01
***P < 0.001