A genome-wide association study looked at 3,500 people and found that intelligence… (Los Angeles Times )
Intelligence is in the genes, researchers reported Tuesday in the journal Molecular Psychology.
The international team, led by Ian Deary of the University of Edinburgh in Scotland and Peter Visscher of the Queensland Institute of Medical Research in Brisbane, Australia, compared the DNA of more than 3,500 people, middle aged and older, who also had taken intelligence tests. They calculated that more than 40% of the differences in intelligence among test subjects was associated with genetic variation.
The genome-wide association study, as such broad-sweep genetic studies are known, suggested that humans inherit much of their smarts, and a large number of genes are involved.
Booster Shots asked Deary to answer a few questions about the research. The following is an edited version of our questions and his emailed responses.
What exactly were you looking for when you looked at test subjects' genetic information?
We studied over 3,500 people. We looked at over 500,000 individual locations on the chromosomal DNA where people are known to differ. We looked at the association between those DNA differences and two types of intelligence. One type of intelligence was on-the-spot thinking (fluid intelligence) and the other was vocabulary (crystallized intelligence).
You wrote in your paper that 40% of the variation in crystallized intelligence and 51% of the variation in fluid intelligence is associated with genetic differences. How did you calculate those figures? And where does the rest of intelligence come from? Other genes, or environmental factors?
To estimate the proportion of variance associated with common genetic differences (in what are called single nucleotide polymorphisms, or SNPs) we used a new genetic statistics procedure invented by Professor Visscher and his colleagues in Brisbane, called GCTA. The rest of people's differences in those types of intelligence could come from genetic differences we were not able to capture, or from the environment.
Certainly, twin and adoption studies tell us that the environment makes an important contribution to intelligence differences throughout life, and especially in early childhood.
Is this the first time such a study has been attempted? How have scientists studied the relationship between genes and intelligence in the past?
There have been some studies looking at individual genes and sets of genes. And some smaller studies have been conducted with coarser genetic sweeps. This is the first study to use thousands of people, half a million genetic variants and to apply this new GCTA procedure to
estimate the genetic contribution directly from the genes.
Why would it be surprising that intelligence is an inherited trait? Many people might say this seems obvious.
It is not surprising to find that intelligence differences have some genetic foundation. Twin and adoption studies have been suggesting that for decades. But those studies make assumptions -- for example that the environment is just as similar for non-identical twins as for identical twins -- and people have questioned those assumptions.
Here, we bypass all that and test the DNA. What is not at all obvious is what the genetic contribution is. From our results, we can suggest that a substantial amount of the genetic contribution to intelligence differences comes from many, many small effects from genetic variants that are in linked with common variants (SNPs).
What parts of your study and analysis do you suspect might receive criticism, and on what grounds?
We don't point to individual genes among the 40%-50% of the variance we detected. We need far larger numbers to do that. We know now that it would be better to have ten times or more subjects than we tested.
We did not have exactly the same intelligence tests in each sample, so that might have led us to underestimate some effects. The GCTA procedure is not easy to understand, so it is hard for people to get their head round how the estimate for the genetic contribution is derived.