This 2015 UK human study by many of the coauthors of What’s the origin of the problem of being fat? applied the Horvath epigenetic clock method to the same UK mother-child pairs and a Danish cohort:
“There has been no investigation on prenatal and antenatal factors that affect AA [age acceleration] in children. It is possible that the detrimental consequences of a higher AA may accrue over time, initiating in childhood. Conversely, it could be postulated that having a positive AA during early life and childhood is developmentally advantageous. To reflect this, we could refer to AA as an epigenetic measure of development in children.
We found associations between AA and sex, birth weight, caesarean section delivery and several maternal characteristics, namely smoking in pregnancy, weight, BMI, selenium and cholesterol level.
Offspring of non-drinkers had higher AA on average at birth, but this appeared to resolve during childhood. Offspring of smokers had higher AA on average and this difference became larger during childhood and adolescence.
The lack of correlation between AA and several clinical variables may also indicate that AA reflects an ‘intrinsic’ aging rate that is independent of various aging factors.
The observation that the estimated genetic component of AA increased in older study participants may indicate that the AA measure is more biologically meaningful in adults rather than children, though alternatively it could be a reflection of a decreasing environmental influence on DNA methylation patterns over time.
This accords with our finding of strengthening within subject correlation over time, which suggests the period of rapid early life changes in methylation affects epigenetic age during development to a greater extent than adulthood changes in methylation.”
The heritability of age acceleration was analyzed:
“The heritability estimate from our study (h = 0.37) is lower than that reported Horvath (h = 1.0), which was based on a small number of cord blood samples from twin pairs. Both of these heritability estimates were based on relatively few samples. Future large scale studies will be needed to arrive at precise estimates of the heritability of AA in newborns and minors.
While our heritability estimate may seem low, empirical evidence has suggested that fitness related traits tend to have lower heritability than morphological traits because selection acts to purify deleterious genetic variation, and one might consider age accelerated residuals in the former category.”
Like the coauthors’ follow-on study, causality couldn’t be definitively determined:
“Assessing the causal relationship between exposures and AA (through Mendelian randomization) is underpowered in our current data.”
Epigenetic age acceleration at birth seemed to be overall “developmentally advantageous” for offspring of non-drinking mothers. That age acceleration continued for the offspring of smokers at the second and third measurement times (ages 7 and 15-17) seemed to have “detrimental consequences.” I’d guess that the methylation state of specific CpG sites would be more informative than the overall rate in these cases.
The point about “AA..is independent of various aging factors” was similar to one made in Using an epigenetic clock to distinguish cellular aging from senescence:
“Cellular ageing is distinct from cellular senescence and independent of DNA damage response and telomere length.”
The study was a step toward establishing the Horvath epigenetic clock for widespread usage. The Hannum method was also compared and contrasted.
http://hmg.oxfordjournals.org/content/25/1/191.full “Prenatal and early life influences on epigenetic age in children: a study of mother-offspring pairs from two cohort studies”