Epigenetic gene promoter methylation at birth is associated with child's later adiposity.

This exceptional paper shows, for the first time, a link among nutrient intake in human pregnancy, epigenetic marks in non-imprinted genes apparent at birth, and offspring phenotype (adiposity) when the child reaches 9 years of age.

There is increasing evidence that adult adiposity is influenced by factors operating in utero which are independent of (but may synergise with) the effects of lifestyle of the adult. For example, recent data from Scotland suggest that percentage body fat in adults aged 30 is positively (and independently) influenced both by maternal body mass index (BMI) during pregnancy and by maternal weight gain {1}. The mechanism by which the in utero environment impacts long-term offspring outcomes is unknown, although epigenetic mechanisms are likely to be involved.

This exceptional paper by Godfrey and colleagues, also discussed in an accompanying editorial {2}, provides evidence of a link among maternal nutrient intake, epigenetic marks apparent at birth, and offspring phenotype (adiposity) when the child reaches 9 years of age. Low maternal carbohydrate intake and increased offspring percentage fat mass (calculated by DEXA scanning when the child was 9) were both positively correlated with hypermethylation of a promoter region of the retinoid X receptor-alpha (RXRA) gene. There was no correlation between RXRA methylation and offspring weight at birth nor with ponderal index. The absolute difference in body fat percentage between the groups with the lowest and the group with the highest quartile of methylation was 4%, with gene methylation and offspring sex explaining 25% of the variation in body fat. The validity of these results is strengthened by Godfrey and colleagues being able to show similar associations in a second population where offspring body fat was measured over a broader age range.

These results are a major step forward in understanding the links between the uterine environment and the human adult phenotype but, as with most good research, raise further questions. Are these links merely associations, or does RXRA hypermethylation contribute causally to offspring phenotype? If the latter, what is the mechanism of this? Is RXRA hypermethylation apparent in samples other than the umbilical cord (e.g. blood), which might be more readily available? Perhaps most importantly, could RXRA hypermethylation be used as a biomarker at birth or in infancy (before differences in body weight are apparent) to facilitate stratified medicine in the prevention of childhood and adult obesity?