In utero exposures have been shown to impart increased risk for obesity in the offspring and are associated with epigenetic changes. DNA methylation was measured in 351 umbilical cord blood (UCB) samples within the Healthy Start Study, a well-characterized prebirth cohort that is following mother-infant pairs from pregnancy-early childhood. The primary objective was to test if DNA methylation at birth was associated with fat mass (FM) gains and infant growth (weight-for-age Z-score, WAZ) during the first 6 months (mos) of life to identify potential biomarkers of future obesity risk. DNA methylation was assessed in UCB via the Illumina 450K array. Adiposity was measured at birth and 4-6 mos of age via PeaPod. Infant weight was obtained from medical records. After adjusting for cellular heterogeneity, sex, and race we found 742 methylation sites (CpGs) through a univariable analysis that were associated with the rate of FM (g/day) in the first 6 mos of life (p < 0.001). Then, partial least squares (PLS) fitting using a multivariable analysis selected a subset of 235 predictor CpGs in this list for KEGG pathway analysis. Similarly, 1314 CpGs were associated with WAZ (p < 0.001); PLS selected a subset of 198 CpGs for KEGG analysis. Using a FDR < 0.001 for KEGG analysis, 3 pathways were associated with FM gain (Metabolic pathways, Glyoxylate and dicarboxylate metabolism, Neurotrophin signaling) and 8 pathways for WAZ gain (Gastric acid secretion, Oocyte meiosis, Metabolic pathways, Endocytosis, Tight Junction, Oxytocin signaling, GnRH signaling, and Inflammatory mediator regulation of TRP channels). The KEGG: Metabolic pathways was identified for both FM and WAZ gain, and 3 gene targets enriched in this pathway, ATP5C1, CYP26A1, and IDH3A are known to be involved in fat metabolism. Our studies have identified potential candidate methylation biomarkers, present at birth, which can be used for early risk detection in order to mitigate development of childhood obesity.


S.J. Borengasser: None. A.P. Starling: None. W. Zhang: None. J. Friedman: Advisory Panel; Self; Janssen Pharmaceuticals, Inc.. R.F. Hamman: None. I. Yang: None. K. Kechris: None. D. Dabelea: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at