Autism and schizophrenia may be two sides of the same genomic coin. Copy number variations in the exact same genes determine whether patients suffer from one condition or the other, according to data presented on Friday (Apr. 3) at the Sackler Colloquium on Evolution in Health and Medicine in Washington, DC.

Both autism and schizophrenia involve disturbances in brain areas linked to social functions, but the two psychiatric disorders often display diametrically opposite traits. For instance, social cognition is underdeveloped in autism but hyper-developed in schizophrenia. Several recent studies have also implicated some of the same genes in the two types of conditions, which has led researchers to suggest a common underlying genetic basis for both brain disorders.

Last year, Bernard Crespi, an evolutionary biologist at Simon Fraser University in Burnaby, Canada, and Christopher Badcock, a sociologist at the London School of Economics, proposed that autism and schizophrenia result from an evolutionary tug of war between the genes inherited from each parent. In their view, the development of these brain disorders is mediated in part by changes in genomic imprinting. Put simply, if the brain is controlled by father-dominated genes, you become autistic, whereas if the brain is molded by mother-dominated genes, you become paranoid schizophrenic, they argued.

Now, Crespi has another idea for how the two disorders might also be linked: In his talk, he suggested that structural changes in a specific set of brain-related genes lead to either autism or schizophrenia. To test his idea, Crespi examined 18 studies published over the past two years that quantified deletions and duplications in patients with either autism or schizophrenia. Combining these studies, he identified seven genomic regions that were linked to both disorders. Five of them were diametrically expressed in patients of the two conditions -- overwhelmingly deleted in one disease and duplicated in the other.

"I was astonished by just how strong [the effect] was," Crespi told The Scientist. "I didn't expect the results to come out so cleanly because it's not something you expect genetic data to do for you."

This newfound link is an "extremely powerful and conceptual hook" to understand both diseases, Crespi said. For example, schizophrenic patients are thought to produce more dopamine in the brain, so perhaps researchers should now look to test whether autistic brains produce a paucity of the hormone, he said.

Another implication of the study is that children with early symptoms of schizophrenia might be being misdiagnosed as autistic, and vice versa. For several of the loci showing differential expression in the two disorders, there were a small number of patients showing the opposite effect -- for example, an autistic patient with a deletion that is normally associated with schizophrenia. Crespi suspects that these are mostly "false positive" diagnoses, rather than results that run counter to his hypothesis. Such incorrect medical assessments "could make children worse," Crespi said, adding that understanding the genetic underpinnings of these diseases could help avoid the problem.

"I took Crespi's point very seriously that if a youngster is misdiagnosed and mistreated then all hell breaks loose," said Mary-Claire King, a geneticist at the University of Washington in Seattle who was not involved in the study.


Related stories:

New autism loci discovered
[12th November 2008]

Inducing autism
[June 2008]

Copy number linked to autism
[15th March 2007]