In the largest collaborative study of the brain to date, scientists using imaging technology at more than 100 centres worldwide have for the first time zeroed in on genes that play a role in intelligence and memory. The findings have been published in the journal Nature Genetics.

More than 200 researchers involved in Project ENIGMA (Enhancing Neuro Imaging Genetics through Meta-Analysis ) analysed the MRI scans and DNA samples of 21, 151 healthy people and discovered that a single gene - HMGA2 - had a measurable impact on a person's brain size as well as intelligence. People who carried a 'C' variant of the gene had brains that were about one per cent larger than those of people who carried the 'T' variant and also scored higher on standardised IQ tests - although the scores differed by only 1. 29 points. The two variants are equally distributed - about half of people have one and half have the other.

The study also marks the first time when researchers from across the globe came together to study the brain. "What's really new here is this movement toward crowdsourcing brain research... This is an example of social networking in science, and it gives us a power we have not had, " says Dr Paul Thompson, a professor of neurology at the University of California, Los Angeles, and senior author of one of the papers.

Brain imaging studies are expensive and, as a result, far too small to reliably tease out the effects of common gene variations. These effects tend to be tiny and difficult to distinguish from the background "noise" of other influences. And brain imaging is notoriously noisy: not only does overall brain size vary from person to person, but so do the sizes of brain regions like the hippocampus, which is critical for memory formation.

To solve the numbers problem, Dr Thompson and three geneticists - Nick Martin and Margaret Wright, both of the Queensland Institute of Medical Research in Australia, and Barbara Franke of the Radboud University Nijmegen Medical Center in the Netherlands - persuaded researchers around the world to pool resources and create a large database. As the study was being completed, the group learned that another consortium, led by Boston University researchers, was doing a similar analysis.

The two teams' findings did not completely line up. One found size-related genes that the other did not. But they agreed on two findings: the HMGA2 gene that correlated strongly with overall brain size, and another that correlated with the rate at which the hippocampus atrophies, or shrinks, with age.

The collaborators found that about 10 per cent of people carried a gene variant that correlated with a slightly accelerated rate of atrophy in the hippocampus. The hippocampi - there are two, each deep in the brain, one in the right side and one in the left, about level with the ears - are needed to form new memories. People with dementia often show pronounced atrophy in this region. The study was not set up to find a link between the gene variant and dementia, but experts suspect a connection.

In a separate analysis in Australia, Dr. Martin and Dr Wright found that size correlated with IQ. People with the larger brains scored slightly higher on a standardised test. The results are all averages, meaning that they hold for the group but say nothing about any individual. (Some very smart people have relatively small brains. )

The collaboration is not likely to lead to new treatments any time soon, the authors said, and, as always, the findings will need replication before they are conclusive. It is more a beginning than an end, and it illustrates how far the field has to go to get any real traction - and what it will take.

"It means sharing your data, pooling everything, " Dr Thompson said, "and this is not usually how scientists work. "