Life ScienceA few weeks ago I reported some research that seemed to show a relationship between a gain in IQ due to breastfeeding and a particular genetic variant. Looks like I spoke too soon. p-ter has the goods:
The fact that they have a measure of maternal IQ but don't directly include it in the published multiple regression suggests that they tried it, but didn't like the results. They didn't include parental phenotype in any of their previous studies, but there, at least, there was some functional evidence linking the polymorphism and the phenotype. Here, there's nothing. Considering the fact mentioned in a previous post that other researchers find absolutely no evidence for link between IQ and breastfeeding (the entire basis for this study), this has to be classified as highly questionable. And regardless of the veracity of any gene-environment interactions here, the huge effects of breastfeeding on IQ shown by the authors are clearly artefacts of the heritability of IQ, and it's unfortunate that they are being propagated.
This is the classic problem that Judith Rich Harris points on in her work, many researchers just don't control for the correlation between parents and offspring genetically.
Comments
There was an interesting study published in 2006 by Der et al (BMJ. 2006 Nov 4;333(7575):945) that did a pretty good job showing that:
1. When maternal IQ is controlled for the breastfeeding-IQ increase dissapears.
2. Siblings who were breastfed did not have higher IQ than siblings who were not breasfed.
Posted by: ozzy | November 18, 2007 11:09 PM
Sorry that was using sibling pairs and directly comparing the breastfed sibling to the non-breastfed sibling.
Posted by: ozzy | November 18, 2007 11:10 PM
Did Derr et al control for regression to the mean when controlling for maternal IQ?
At what ages were children tested? It is possible that breastfed children mature more slowly.
Posted by: Rob | November 19, 2007 7:49 PM
Rob:
I'm not familiar with controlling for regression to the mean. If you have time, perhaps you can give the paper a read and tell me if they did and if not how would someone do it. I would greatly appreciate it. I think I have an idea of what regression to the mean is.
Also, the children were tested every two years from age 5 to 14.
Posted by: ozzy | November 20, 2007 10:30 PM