Association of maternal obesity with preterm birth phenotype and mediation effects of gestational diabetes mellitus and preeclampsia: a prospective cohort study

Table 4 Mediation effect of GDM/PE between maternal pre-pregnancy BMI and preterm birth

Preterm birth	Mediator	Adjusted OR(95%CI)			Proportion mediated
Preterm birth	Mediator	Natural direct effect (NDE)	Natural indirect effect (NIE)	Total effect (TE)	Proportion mediated
All preterm births	GDM	1.002 (1.001 to 1.003)***	1.0003 (1.0002 to 1.0005)***	1.002 (1.001 to 1.003)***	13.41%
Spontaneous preterm labor	GDM	1.0001 (0.9996 to 1.001)	1.0001 (1.00004 to 1.0002)***	1.0002 (0.9997 to 1.001)	32.80%
Premature rupture of the membranes	GDM	1.001 (1.0002 to 1.001)**	1.0001 (1.0000 to 1.0002)**	1.001 (1.0003 to 1.001)**	12.80%
Medically indicated preterm birth	GDM	1.001 (1.001 to 1.002)***	1.0001 (1.00003 to 1.0002)**	1.001 (1.001 to 1.002)***	7.50%
All preterm births	PE	1.001 (1.0005 to 1.002)***	1.001 (1.0001 to 1.001)*	1.002 (1.001 to 1.003)***	36.66%
Spontaneous preterm labor	PE	1.0003 (0.9997 to 1.001)	0.99997 (0.9999 to 1.000003)	1.0003 (0.9997 to 1.001)	0
Premature rupture of the membranes	PE	1.001 (1.0003 to 1.001)**	1.00004 (1.0000001 to 1.0001)*	1.001 (1.0003 to 1.001)**	4.51%
Medically indicated preterm birth	PE	1.0004 (0.9999 to 1.001)	1.001 (1.0002 to 1.001)*	1.001 (1.0004 to 1.002)***	64.30%

Abbreviations: BMI Body mass index, GDM Gestational diabetes mellitus, PE Preeclampsia, OR Odds ratio, CI Confidence interval
Continuous variable BMI was analyzed as independent variable in these causal mediation analyses
ORs were adjusted for maternal age, educational level, and assisted reproduction technology
Basically, the OR estimations were rounded to 3rd decimal. If the rounded result was 1.000, then the result would be revised and rounded to the non-zero decimal digit
*, P < 0.05; **, P < 0.01; ***, P < 0.001

ISSN: 1471-2393