|Figure 1: Rise in childhood obesity from 1963-2008 (from CDC.gov)|
Godfrey et al. analyzed the methylation status (if a gene is methylated it is turned off) of several genes in the cord blood of newborns from two independent cohorts (cohort 1: 68 samples; cohort 2: 31 samples) and found that
[i]n cohort 1, retinoid X receptor-α (RXRA) chr9:136355885+ and endothelial nitric oxide synthase (eNOS) chr7:150315553+ methylation had independent associations with sex-adjusted childhood fat mass (exponentiated regression coefficient [β] 17% per SD change in methylation [95%CI 4-31], P = 0.009, n = 64, and β = 20% [9-32], P < 0.001, n = 66, respectively) and %fat mass (β = 10% [1-19], P = 0.023, n = 64 and β =12% [4-20], P = 0.002, n = 66, respectively).Furthermore, additional statistical analyses revealed that [in cohort 1] these “epigenetic marks explained >25% of the variance in childhood adiposity”.
While the correlation with eNOS, which was found in cohort 1, could not be established in the second cohort, the correlation between retinoid X receptor-alpha methylation and childhood obesity [measured at the age of 9 years] was observed (to a lower degree, β(BodyFat%)=4%) in cohort 2, as well. The exact mechanism, by which these nuclear receptors, which mediate the biological effects of retinoids on our bodies, influence the deposition of body-fat, is still not completely elucidated, the activation of PPAR-gamma, which by itself has attracted some attention as a potential target for treatment strategies for the metabolic syndrome, could yet play a major role.
|Figure 2: Exponentiated regression coefficients β of retinoid X receptor-α (RXRA) and endothelial nitric oxide synthase (eNOS) methylation with body fat % of children at the age of 9 years.|
(data adapted from Godfrey. 2011)
On a side note: If you are interested in the question how reliable scientific "data" (which is by no means, as the name /lat. dare = to give/ would imply, "given" to us) actually is and have not yet listened to my latest appearance on Carl Lenore's Super Human Radio, where I discuss some factors to consider, when you look at the data from scientific studies, I suggest you go and download the podcast ;-)