NEW PATTERN OF RISK: A new theory explaining the genetic underpinnings of autism states that in some families, a male child could have as high as a 50 percent risk of developing the disorder.
© TONY KURDZUK/STAR LEDGER/CORBIS

In work that may one day lead to earlier detection of children at risk of developing autism, a team of scientists has devised a genetic model for the enigmatic disorder. The two-tiered theory integrates families with one or more autistic children.

An estimated one in every 150 children born in the U.S. develops autism, according to the Centers for Disease Control and Prevention (CDC); it is four times more prevalent in boys than in girls. The condition is characterized by cognitive deficiencies and symptoms ranging from antisocial (not responding to one's name and / or avoiding eye contact) to obsessive, repetitive behavior. The most popular theory about its genesis is that there are flaws in several genes passed down through generations of a family that culminate to predispose a child to the disorder, especially if exposed to certain environmental factors such as toxic chemicals or a lack of oxygen at birth.

"People thought there was this uniform risk—if you have an autistic child, then there's some uniform, but fairly low, risk that you'll have another one," says Michael Wigler, a professor of genomics at Cold Spring Harbor Laboratory (CSHL) in Long Island, N.Y., and senior author of the new model described in Proceedings of the National Academy of Sciences USA. "None of the population geneticists, in my experience, had thought that there might be two classes of families: low risk and high risk."

Wigler's rethinking of autism's cause stems from an exhaustive analysis of risk based on a database of families with more than one autistic child. (The Autism Genetic Resource Exchange, or AGRE, manages the database.) The team determined that most cases of autism arise from novel, spontaneous mutations passed down from one or both parents, resulting in large gaps in a person's genome often encompassing several genes, which are then disrupted or inactivated. (This loss of genetic code—known as copy number variation—results in an offspring receiving only one of the standard two copies of a gene, which could cause an insufficient amount of protein to be produced by those genes.) In most instances, this mutation will result in an autistic child. However, in some cases—more likely in girls than boys—the recipient of this mutation will not produce any symptoms.

"When that child matures and becomes a parent, they have a 50 percent chance of transmitting … [their mutation] … to a child that might not be as lucky as they were, especially if … [its] … a boy," Wigler says. "So, they will be transmitting this with close to a 50 percent frequency—and that is the source of the high-risk families."

Wigler says that the team will continue to update its model as new figures are added to the AGRE database and try to gain new insight into the mechanism that gives girls greater resistance than boys. "To understand that [disparity] at a molecular or genetic level would be very important, because you could theoretically treat kids … you could detect something early and intervene," Wigler says. "I view it as the most important thing to understand."

Maja Bucan, a professor of genetics at the University of Pennsylvania, says that the new autism model is a creative way to interpret the familial data. "It's important to come up with new theories and then just test them once we have more data," she explains. "I don't think we have enough data [yet] to say whether this theory is right or wrong."

According to Wigler, the new model "certainly changes the way you think about autism. The paradigm shift is … something can be genetic without being heritable. The field has ignored the contribution of spontaneous mutation for a whole range of things that matter a lot to society," which, he adds, includes schizophrenia and morbid childhood obesity.