Table 2.

Total, direct and indirect effect among the pathways between prenatal (maternal pre-pregnancy BMI) and postnatal factors (infant birth weight, infant rapid weight gain, breastfeeding duration) and BMI z-score in early childhood: pooled effect from seven Australian and New Zealand cohorts.

	β or Odds ratio	95%CI		P value
Maternal pre-pregnancy BMI → Child BMI z-core
Total effect	0.04	0.03	0.05	<0.001
Direct effect	0.03	0.03	0.04	<0.001
Indirect effect
via infant birth weight	0.01	0.01	0.01	<0.001
via infant rapid weight gain	0.002	−0.01	0.01	0.70
via breastfeeding duration	−0.003	−0.01	0.003	0.28
Maternal pre-pregnancy BMI → Rapid weight gain in infancy (yes vs no)
Total effect	1.05	1.02	1.07	<0.001
Direct effect	1.00	0.99	1.02	0.70
Indirect effect
via infant birth weight	0.98	0.97	0.98	<0.001
via breastfeeding duration	1.07	1.05	1.09	<0.001
Maternal pre-pregnancy BMI → Breastfeeding duration (≥6 vs < 6 months)
Total effect	0.92	0.91	0.93	<0.001
Direct effect	0.92	0.90	0.93	<0.001
Indirect effect
via infant birth weight	1.00	1.00	1.01	0.002
Infant birth weight→ Child BMI z-score
Total effect	−0.84	−1.03	−0.65	<0.001
Direct effect	0.57	0.51	0.64	<0.001
Indirect effect
via infant rapid weight gain	−1.42	−1.62	−1.22	<0.001
via breastfeeding duration	0.01	−0.01	0.03	0.33
Breastfeeding duration(≥6 vs <6 months)→ Child BMI z-score
Total effect	−0.50	−0.65	−0.34	<0.001
Direct effect	0.04	−0.03	0.11	0.28
Indirect effect
via infant rapid weight gain	−0.54	−0.65	−0.34	<0.001

Adjusted for cohort, ages at infancy and early childhood assessments, maternal age, child sex, and intervention allocation. β-coefficients are presented for continuous outcome (BMI z-score), and odds ratios are presented for categorical outcomes.

CI confidence interval.