0.5% to 3%: Do we now have a better sense of what makes people smart?

The Economist “A potent source of genetic variation in cognitive ability has just been discovered”

A new gene variant, KL-VS, appears to account for up to 3% of variation in IQ score; this would be a radical discovery given that past gene screens have only found variants accounting for as much as 0.5% variation in IQ. This is an interesting discovery that needs to be put into context; this article does a relatively good job of cautioning the reader from jumping to wild conclusions, but not a perfect job. If you are interested in getting a bit more background on this issue, Carl Zimmer provided a nice overview of “The Search for Intelligence” in Scientific American a few years back (Zimmer 2008a).

What’s always really tough for the public to understand with these kinds of findings is how likely they are to occur due to random chance. As this article correctly states about single-gene discoveries, “flukey correlations between one of them and some human trait are common”. A sample size of 718 sounds good, but the problem is that these studies are screening thousands (if not millions) of candidate genes for correlations with traits, so the chance of one of them appearing to cause variation in traits has to be assessed relative to the sample size used.

As the article points out, to really make a case for this KL-VS allele as an intelligence booster, one needs to show the mechanism by which it boosts intelligence. The studies in mice are a nice first step in this direction.

And for those who might be reading this randomly off a web search, I will provide my usual two reminders on the issue of single-gene discoveries about intelligence:

1. Even if a gene shows significant impact on a particular trait across the population, it is very hard to translate what that means for a particular person who has or does not have a copy of that gene. Genes interact significantly with other genes; for this reason it would be pretty difficult to compare two people’s “innate IQ score potential” based on the fact that one person possesses KL-VS and another does not.
2. Searching for the “genetic basis of intelligence” usually means searching for the “genetic basis of scores on the IQ test”, a shorthand that can really mislead. As any evolutionary biologist can tell you, a trait is only as good as the environment it finds itself in, and the IQ test is only one environment by which to judge intelligence. There are plenty of adaptively-significant forms of intelligence that are not captured by the IQ test, and there is plenty of reason to think that human social groups require multiple forms of intelligence in order to function.