Evidence for paternal programming from obesity?

It is well understood that the diet and health status of mothers can affect offspring by epigenetic mechanisms (see this recent review for example).  This is often called “fetal programming.”

A recent study (published yesterday) suggests there may be a paternal influence as well- epigenetic programming on the offspring based on the health status of the father.  Maybe.  This study has major limitations and it only found an effect on female offspring, and the results certainly aren’t anywhere close to the influences observed in fetal programming research.  It shouldn’t be getting the dramatic news headlines that i’ve seen.  But it is interesting work and an interesting area, so without too much detail, here is a summary of the study:

The investigators made male Sprague-Dawley rats fat with a high fat diet and had another male group as a control that were fed a diet that kept them slim.  Both groups were mated with slim females on a control diet.  The resulting offspring were examined.

The authors note that human paternal obesity is associated with offspring that have a low birth weight.  They found that female offspring on the first day of birth from the obese fathers had a lower birth weight.  This effect wasn’t statistically significant in males.  The gender differences may be true or have something to do with the low number of offspring examined (34 in total between the groups).

The authors note additionally that associations exist in humans for adiposity and insulin resistance between girls and obese fathers.  They state that an unpublished pilot study on mice found a gender specific female glucose intolerance after weaning in (an apparently) similar design.  However in this study they found that female offspring from obese fathers did not have altered body weight, growth rate, energy intake, or energy efficiency.  Adiposity, muscle mass, fasting leptin, triglycerides, or non-esterified fatty acids were also not altered.  The lack of effect on obesity through paternal lineages is unlike observed in maternal.

Glucose tolerance, insulin sensitivity, and insulin secretion were also examined.  The obese fathers did not effect their daughters’ fasting blood glucose or plasma insulin but did negatively effect glucose tolerance (insulin secretion impaired).  Islet and beta cell area were reduced (the latter a non significant trend).  The authors suggest that beta cells were sufficient to maintain fasting glucose and insulin but not enough to maintain insulin secretion with a glucose challenge.

Analysis of a number of genes found differences, and they focused on the most dramatic difference at Il13ra2 (regulate islet function). Methylation of a site on this gene was found to be reduced in the offspring of obese fathers, which would be one epigenetic alteration that might mediate these effects.

Obviously, epigenetic changes from the father would need to be influenced through the sperm, in which enough evidence exists to show sperm composition and quality can be altered in a number of ways (and by diet).

This is certainly interesting but I am wary of some of the results because of small sample sizes and small effects.  It will be interesting to see what research follows up.

*Edit Oct. 22: See this post from Larry Parnell who lists 71 additional genes with known or computationally predicted paternal imprinting.  Lots yet to learn!

*Edit Oct. 26: And this one also from Larry on the gene S1PR5 from this study.


Ng, S., Lin, R., Laybutt, D., Barres, R., Owens, J., & Morris, M. (2010). Chronic high-fat diet in fathers programs β-cell dysfunction in female rat offspring Nature, 467 (7318), 963-966 DOI: 10.1038/nature09491

  • anonymous

    The results from this paper are just experimental artifact. It should be retracted by Nature.

    Foremost, the impairment of glucose metabolism that the authors report is ridiculously small (look at Fig. 2). This data is based almost entirely on glucose tolerance tests, which are difficult/impossible to use to reliably measure such small differences.

    To understand the mechanism of this “dysfunction”, the authors performed microarray analysis from the pancreas of affected and unaffected progeny, and reported a table of genes that were differentially expressed in the two cohorts. Incredibly, the single largest gene expression difference in this table was 1.7-fold, and many of the changes were 1.1-fold. Changes of this magnitude are not real — they are just microarray noise. The authors report p-values for each of their 1.1-fold changes in gene expression, but it is unclear whether they corrected for multiple hypothesis testing ,since they are looking across 25,000 genes. Of course, these differences couldn’t be validated even if the authors tried (and the authors didn’t try, as none of the reported gene expression changes were shown to be functionally relevant.)

    This paper is embarrassing for Nature.

  • http://www.recomp.com Colby

    Bookmarking this here: http://www.scientificamerican.com/article.cfm?id=fat-fathers-affect-daughters-h

    But not everyone is convinced. Stephen O’Rahilly, a clinical biochemist at the University of Cambridge, UK, notes that “the difference in glucose tolerance between these animals is pretty slim, and the number of animals in the study was too few to give a robust signal”.

  • Pingback: A low-protein paternal diet alters hepatic gene expressions in progeny | Nutritional Blogma