Maternal BMI at the start of pregnancy and offspring epigenome-wide DNA methylation: findings from the pregnancy and childhood epigenetics (PACE) consortium.
Level 3 - non-randomized controlled study
Meta-analysis of prospective observational cohort studies (CEBM Level 3 by design analogy).
PubMed 29016858 · doi:10.1093/hmg/ddx290
What was done
Researchers within the Pregnancy and Childhood Epigenetics (PACE) Consortium meta-analysed associations between pre-pregnancy maternal BMI and DNA methylation at over 450,000 CpG sites in newborn blood across 19 cohorts (9,340 mother-newborn pairs). Persistence of associations was assessed in 4 additional cohorts of adolescents (1,817 mother-child pairs). Causality was evaluated by comparing maternal versus paternal BMI effect sizes and integrating offspring genetic variation.
What was found
Maternal BMI was univariately associated with small (<0.2% methylation change per 1 kg/m2 BMI increase, P < 1.06 x 10^-7) differences across 9,044 CpG sites in newborns. Adjusting for estimated blood cell-type proportions attenuated significance to 104 CpG sites, including 86 sites shared with the unadjusted model. At 72 of these 86 sites, the direction of effect persisted into adolescence. However, causal inference analyses supported a direct intrauterine effect for only 8 of the 86 sites.
Why it matters
This large-scale study demonstrates that while maternal adiposity correlates with offspring epigenetic variation, effect sizes are minimal and largely explained by shared genetics or postnatal lifestyle rather than direct intrauterine programming.
Limits
Measurements were restricted to peripheral blood rather than target metabolic tissues. Effect sizes were extremely small (<0.2% per BMI unit), cell composition accounted for most initial associations, and causal inference relied on observational genetic/paternal comparisons rather than experimental intervention.
Cited by
- context Children born to obese mothers are epigenetically predisposed toward developing obesity and type 2 diabetes.