Wednesday, October 28, 2009

"Grandma plays favorites"

Molly Fox and a group of coauthors recently published a paper on grandmothers and child survival that shows improved survivorship among children with more of the grandmother's genetic material. The idea then is that grandmothers with "longevity genes" (not "grandma jeans") pass along their habit of being long-lived to their (paternal, it turns out) granddaughters. Hence a yet-clearer evolutionary argument for long life past menopause.

The authors examine girls' and boys' mortality based on the presence of maternal or paternal grandmothers, and they pose the idea that differences in the inheritability of the X chromosome across sexes of grandkids and paternal/maternal sides ought to be interesting. At the top of this post is their neat Figure 1 that traces the heritability of genes.

They find higher mortality among boys with a paternal grandmother present compared to those with a maternal grandmother present, and that the presence of a paternal grandmother was statistically significantly better for girls than boys.

I was shocked that they had data that could inform this. I also wondered whether there are omitted variables. Is it just random whether a paternal or maternal grandmother is present, or whether one is present at all? Or is it connected to low SES or poor health somehow?