![](data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="16" height="16"><rect fill-opacity="0"/></svg>)
RSS
We identify several common genetic variants associated with cognitive performance using a two-stage approach: we conduct a genome-wide association study of educational attainment to generate a set of candidates, and then we estimate the association of these variants with cognitive performance. In older Americans, we find that these variants are jointly associated with cognitive health. Bioinformatics analyses implicate a set of genes that is associated with a particular neurotransmitter pathway involved in synaptic plasticity, the main cellular mechanism for learning and memory. In addition to the substantive contribution, this work also serves to show a proxy-phenotype approach to discovering common genetic variants that is likely to be useful for many phenotypes of interest to social scientists (such as personality traits).
Ewen Callaway has a write up in Nature. The issue here is that it’s been evident for the past 10 years or so intelligence variation is not due to alleles segregating at high frequencies with at least modest effects, so they’re hard to pick up in association studies (contrast with pigmentation, which is mostly controlled by a number of loci on the order of 10). Some, such as Kevin Mitchell, don’t think that common variants are the way to go, period. Common as in variants which are found across the population, even if their effect on the trait is very small. This group disagrees. One of the authors, Peter Visscher, has written up his own view of this line of research, Intelligence inheritance – three genes that add to your IQ score:
This study of normal variation in cognitive performance confirms that there is no gene with a large effect on this trait. There is no “gene for intelligence” – instead, cognitive performance is likely to be influenced by thousands of genes, each having a small effect.
While the individual effect of the genetic variants are extremely small, their identification may lead to knowledge of the biological pathways involved in cognitive performance and cognitive ageing. This insight may eventually lead us into a better understanding of the mechanism involves in memory loss and dementia.
Finally, because individual gene effects are small, an implication of the study is that even larger studies, for example on millions of people, will lead to the discovery of many more gene variants.
In sum, because intelligence is at least moderately heritable, but the causal variants are so diffuse and numerous, the best bet for having a smart child is picking a spouse with a deviated phenotype. Look for smart people to marry….
![](data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="217" height="51"><rect fill-opacity="0"/></svg>)
- the Greatest Dancer - BBC
An additional qualifier for that, if you were really serious about it, is to find a spouse who’s not only smart, but who’s from a smart family – at the very least, has smart parents.
And presumably a smart family with a different ethnic background from oneself to pick up a wider variety of good alleles?
Depends on the ethnic background, no? The polygenic allele score for the three well replicated alleles discussed in the study shows large global variance. So if IQ is additive …. (Granted a larger batch of alleles could show different polygenic score differences, or maybe even none.)
“And presumably a smart family with a different ethnic background from oneself to pick up a wider variety of good alleles?”
That depends very much on whether you accept the Conventional Wisdom that behavioral gene effects are largely additive and linear, or whether you’re one of these weirdos who keep mumbling about the possibility of large-scale epistasis / gene-gene interactions (such as yours truly).
The conventional wisdom says “yes”, outmarry as much as possible. But if epistasis is significant For Realz, you want to keep the gene combinations as similar as possible (though not to a degree where you would risk inbreeding).
/Ready to get schooled in case I’m hopelessly confused
#4, can i get some lit references to inform me as to your prior here?
I dated two women from Mensa many years ago. They were clearly very bright. One was a crazy alcoholic. The other was just crazy.
This study of normal variation in cognitive performance confirms that there is no gene with a large effect on this trait. There is no “gene for intelligence” – instead, cognitive performance is likely to be influenced by thousands of genes, each having a small effect.
I think it’s neither a large number of genes with a small effect nor a few genes with a large effect. It is most likely a medium number of genes with an “encrypted” effect. That is the individually relevant genes do not act orthogonally or independently in their effects. For example two separate genes A and B with alleles a, A and b, B, aB and Ab may be intelligent and aa and BB may be dumb. Now instead of two, think perhaps 200 and it’s a job for the NSA.
Not only is IQ polygenic but “there is no gene with a large effect on the trait.” Yet somehow, someway, there is an enormous difference between yours truly and the average Joe. I am not expecting answers to this in the near future, but that doesn’t stop me from wondering.
I previously asked this question in earnest over at West Hunter and was laughed away (actually with some funny responses). At risk of that happening again I’ll ask here anyways.
What’s better for the expected IQ of your child: marrying someone with a 130 IQ from a population with a 100 IQ average, or someone with a 115 IQ from a population with a 115 IQ average? Assume all else equal, parents/family are same as population norms, etc.
#4 – I have seen either you or someone else raise this before.
Given that you are talking to real people here who have done precisely what you describe, some references would be appreciated.
Razib:
Sorry, don’t bother. My source for the Conventional Wisdom is R. Plomin, Personal Communication (it doesn’t get more Conventional than that). Unfortunately I didn’t ask for the evidence of this, and I remember that I couldn’t find any justification in his textbook or his articles.
As you know, there is no “smoking gun” of epistasis in IQ. It’s just a plausible alternative that has the advantage of explaining not just the “missing heritability”, but also the surprising result of near-zero shared-environment influences (because if epistasis is high, the twin-based methods will overestimate additive genetic variance and thus under-estimate shared environment).
#7, if you know of some people who have actually looked at epistasis in IQ and published some kind of result, whether negative or positive, I’ll be glad to shut up about it.
It is most likely a medium number of genes with an “encrypted” effect.
you’re just talking about epistasis or gene-gene interactions/networks. empirically it doesn’t seem likely for the normal distribution.
What’s better for the expected IQ of your child: marrying someone with a 130 IQ from a population with a 100 IQ average, or someone with a 115 IQ from a population with a 115 IQ average? A
fwiw, i’d go with the 130 person. in developed nation scenarios i don’t think the regression is going to be 1/2 because narrow sense heritability is probably greater han 0.50. also, you might have a good sense of the population background, though you are confident of that individua’s IQ. but mostly the former.
Razib, there appears to be an association between good health and IQ. According to this paper, there may be a “fitness factor” that partially explains the relationship.
http://my.psychologytoday.com/files/attachments/95822/intelligence-and-physical-health.pdf
What are your thoughts on the “fitness factor”?
Do you believe this could explain the association?