Wednesday, August 19, 2009

Losing your mind? Yes, if you believe these results

The British newspaper, The Independent, last week published a great study in confounding and interpretation of results based on prior assumptions--exactly the kinds of thing that we think is a problem both within science and for those who write about it for public consumption.

The headline is "Losing your mind? The answer is in the mirror." The article reports results of a Scottish study of hundreds of 11 year-olds given an IQ test in 1932, and a recent follow-up of surviving 83 year-olds. Apparently, men with asymmetric faces--but not women--are more apt to lose cognitive ability as they age. Why? The article says
scientists believe that it could be explained by the idea that a good set of genes for facial symmetry may be linked with an equally good set of genes for brain preservation.
The idea that facial symmetry indicates 'good' or 'healthy' genes is not new. Indeed, the sociobiological explanation for beauty itself is that it indicates good genes, and a healthy and long life (which apparently means that beauty is not in the eye of the beholder), so that beautiful women are more likely to be chosen by men than women who are less beautiful. Further,
[t]here is a growing body of emphasis suggesting that the "fitness" of the genes involved in the development of the embryo in the womb, and of the body during childhood, can be measured by analysing the left-right symmetry of the body.
But really, this is just a wild guess that raises a lot of questions. For example, how did these 83 year old men with asymmetric faces get to be 83 if it's symmetry that's linked with genes for longevity? And, what about the fact that throughout human history, mainly spent in small bands with limited mate choice, nearly everyone mates, even people with asymmetric faces? Do men have a private Y-linked 'good set of genes for brain preservation'? Or facial symmetry?

And for decades there have been suggestions that those who have erratic, different, odd, or unusual behavior may in fact be more appealing as mates--may seem more charismatic, for example. And senile dementias may be reflected at least to some extent by behavioral differences earlier in life, if the nun study is to be believed. Whether these ideas are correct has been controversial, but shows the subtleties that may be involved in 'fitness'.

(A lovely little report published in the New York Times on August 4 alludes to a related issue, the ethics of cosmetic surgery in Darwinian terms. The paper from which the article draws is published in The Journal of Evolution and Technology, and suggests that the choice to have cosmetic surgery is unethical because beauty is an indicator of good genes, and attaining beauty unnaturally cheats potential mates into believing you've got good genes. Or, perhaps, does it reveal that you have good IQ genes, that lead to your deviousness? This paper could be the subject of a whole post--except that it's too ludicrous to give even that kind of credence to, starting with the opening sentence ("Evolution continually selects the best genes to proliferate the species.").)

Could there be alternative explanations for the Scottish findings? Could the small sample size (219 of the original much larger sample) mean the results aren't replicable? They aren't in females, although the study authors suggest this is because these women aren't yet old enough to be losing cognitive ability. Their argument is rather convoluted, though--people die on average 4 years after developing dementia, women are older than men when they die on average, women with asymmetry in this study don't yet have dementia, thus they could still lose mental capacity and not die for at least another 4 years. Could be. But it could just as well be that the correlation between asymmetry and dementia is a false one.

What about the genetic interpretation? Are there genes 'for' symmetric faces? Are there genes 'for' longevity? As far as we know, such genes haven't been found--if there are genes 'for' longevity, how did they evolve, given that natural selection doesn't see beyond reproductive age? And, face shape is yet another complex polygenic trait, and genes 'for' face shape (that is, genes that contribute to facial development) will be found all over the genome, and yet the explanation for the results reported in this article suggests that it's a simple question of asymmetry genes linked to dementia genes.

Development involves all sorts of stochastic effects. Left and right sides separate early in the embryo, but from that point til any later age they are largely independent. They have the same cellular genome, but are independently affected by all sorts of aspects of cell division, somatic mutation, and chance interactions with the environment, chance variation in the level of signaling molecules and their response, or in timing of gene expression, and many other things to boot. None of those is heritable and hence none can be affected by natural selection. And of course none of these causes of asymmetry would be linked with risk of dementia, another polygenic complex trait. (With the caveat that early onset dementia is a Mendelian single gene trait, but this isn't likely to be what afflicts the 83 year-olds in this study because the early onset form is also associated with early mortality.)

Hopefully there is more to this story than meets the reporter. Otherwise, it exemplifies shallow if not stupidly simplistic studies that shouldn't be funded or published, but stories about them get into print because the selection that is really important is that you select The Independent from the very asymmetrical display of newspapers trying to get you to look at them to pick up something to read on the train.

No comments: