Brolly · Clinical epigenetics 2026 · Prospective multicenter cohort study with epigenome-wide association and mediation analysis · n=3,116 mother-child pairs

Maternal BMI associations with cord blood DNA methylation and mid-childhood adiposity.

Cited 1 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective observational birth cohort follow-up study

PubMed 42070037 · doi:10.1186/s13148-026-02119-8 · record verified 2026-08-27

What was done

Investigators conducted an epigenome-wide association study (EWAS) and mediation analysis using cord blood DNA methylation (Infinium MethylationEPIC 850K BeadChip) from the prospective, multicenter Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study and its follow-up study (HAPO FUS). Linear regression models adjusted for population substructure, cell counts, and maternal and child covariates with Bonferroni correction. Mediation analysis assessed whether maternal BMI-associated CpG sites explained the relationship between maternal BMI and offspring BMI z-score at a mean follow-up age of 11.5 years (offspring aged 10–14 years).

What was found

In 3,116 mother-child pairs (48% White, 21% Asian, 19% Black, 12% Hispanic; 36% of mothers and 28.3% of children overweight or obese), maternal BMI was significantly associated with DNA methylation at 7 CpG sites: cg00579423, cg07138793, cg12188424, cg19345626, cg20020844, cg02988288, and cg26974062 (including two sites on the TXNIP gene). Cord blood DNA methylation at cg20020844 (SP6) mediated 1.2% of the association between maternal BMI and child BMI z-score.

Why it matters

This study provides multi-ethnic evidence that maternal BMI is linked to altered cord blood DNA methylation at birth. However, the finding that the top site mediated only 1.2% of the child BMI association demonstrates that single-site cord blood methylation explains very little of the transgenerational adiposity relationship.

Limits

The observational design cannot establish causal fetal programming. The proportion of the association mediated was extremely small (1.2%), leaving most of the maternal-child BMI relationship unexplained by these epigenetic marks. The abstract omits effect sizes, beta coefficients, and confidence intervals for the 7 CpG sites, and measurements in cord blood leukocytes may not mirror methylation patterns in target metabolic tissues.

Cited by