Impact of intrauterine exposure to maternal diabetes on preterm birth: fetal DNA methylation alteration is an important mediator

Table 4 Mediation effect of CpGs on the association between diabetes in pregnancy and preterm birth

Mediator	TE/DE (95%CI)	IE (95%CI)	RE (%)
Total effect	1.910(1.094, 2.845)
Direct effect	0.516(− 0.125, 1.105)		0.270(− 0.072. 0.536)
cg08810410		− 0.131(− 0.353, 0.071)	− 0.068(− 0.198, 0.043)
cg25953130		0.272(0.039, 0.476)	0.143(0.026, 0.240)
cg09191149		0.224(0.076, 0.405)	0.117(0.029, 0.230)
cg09915396		0.349(0.142, 0.609)	0.183(0.068, 0.329)
cg07408552		0.440(0.191, 0.773)	0.230(0.100, 0.408)
Joint effect of CpGs		1.169(0.714, 1.722)	0.612(0.380, 0.900)

DE Direct effect, IE Indirect effect, RE Relative effect, TE Total effect
The model was adjusted for maternal age at delivery, race and ethnicity, educational attainment, smoking status during pregnancy, prepregnancy overweight or obesity, neonatal sex, and neonatal birthweight. The response variable is on the scale of log-odds of being preterm. Significance level was set at 0.1

ISSN: 1868-7083