Skip to main content

Table 4 Final linear regression model adjusting for all key variables with average HIF3A.2 methylation as outcome

From: Early-life determinants of hypoxia-inducible factor 3A geneĀ (HIF3A) methylation: a birth cohort study

Ā 

Average across HIF3A.2 (nā€‰=ā€‰513)

Ī² (SE)

p

95% CI

GDM

4.62 (1.77)

0.009

1.14 to 8.09

Pre-eclampsia

āˆ’ā€‰5.20 (2.19)

0.02

āˆ’ā€‰9.51 to āˆ’ā€‰0.89

Sex (male)

āˆ’ā€‰3.77 (0.76)

<ā€‰0.001

āˆ’ā€‰5.26 to āˆ’ā€‰2.29

Gestational age (weeks)

1.21 (0.27)

<ā€‰0.001

0.68 to 1.74

rs3810298

Ā Ā Ā 

ā€ƒC/T

āˆ’ā€‰8.31 (0.97)

<ā€‰0.001

āˆ’ā€‰10.22 to āˆ’ā€‰6.41

ā€ƒT/T

āˆ’ā€‰16.91 (2.73)

<ā€‰0.001

āˆ’ā€‰22.27 to āˆ’ā€‰11.54

  1. Effect sizes (Ī²) given as percentage methylation. Ī² for rs3810298 categories is the difference from homozygote major allele (C/C). GDM gestational diabetes, SE standard error, CI confidence interval. No other SNPs were strongly associated with HIF3A.2 methylation after adjustment for rs3810298, and a linear regression model with just rs3810298 accounted for a similar amount of variation in HIF3A.2 methylation measures and the model with six SNPs (R2 for minimal modelā€‰=ā€‰0.162, R2 for full modelā€‰=ā€‰0.177, pā€‰=ā€‰0.29)