Maternal Micronutrient Deficiency, DNA Methylation, and Fetal Neurodevelopment in Low-Resource Settings: A Secondary Analysis of Growth-Associated EWAS Signatures

Maternal undernutrition is common in urban slums and other low-resource settings, where both low birthweight and childhood stunting have been associated with poorer brain development. One proposed explanation is epigenetic: a shortage of folate and related nutrients lowers the methyl-group supply during the period when the fetal brain is being programmed, altering DNA methylation. This is hard to test directly, because no epigenome-wide association study (EWAS) has yet measured neurodevelopmental outcomes in these populations. We therefore re-annotated published methylation signals for two nutrition-sensitive growth phenotypes: neonatal birthweight from the PACE consortium meta-analysis (914 CpG sites, 729 genes) and child height at the SOCS3 gene in low- and middle-income cohorts. The birthweight genes were sorted into functional categories, compared with three studies of methylation in undernourished children, and annotated against fetal-brain and placental chromatin states and fetal-brain expression. We found no enrichment for one-carbon metabolism; DNMT3A was the only related gene present, and the core folate-cycle enzymes were absent. Neurodevelopmental and synaptic genes made up a sizeable fraction of the signature (92 genes, 12.6%), but this reflects the composition of the gene set rather than a tested over-representation, and the fetal-brain enhancer enrichment we observed applied to the whole birthweight signal, not to these genes in particular, and was not specific to fetal brain. The most consistent finding was SOCS3, a nutrition-responsive locus that recurred across independent undernutrition datasets, altered its expression with methylation in vitro, and is expressed in the prenatal brain. Overall, the data support a nutrition-sensitive methylation signature at growth and immune loci that overlaps enhancers in fetal brain as in other tissues, with SOCS3 as a candidate link to neurodevelopment. The one-carbon-to-neurodevelopment pathway remains plausible but hypothesis-generating, and testing it will need formal enrichment analysis and primary methylation studies where undernutrition is most common.