The study could have special relevance in Utah, which has a high birth rate, large families and a suspected higher rate of autism than the national average.
But whether it will influence family planning is hard to predict, said William M. McMahon, psychiatry chairman at the University of Utah's School of Medicine. "It's a touchy subject. Religion and family values play a big role in deciding how many children you should have. And many individuals with autism are productive members of society and have reasonable quality of life."
McMahon did not contribute to the sibling study, but Ozonoff cites some of his research and she was once on the U.'s faculty.
McMahon said the increased sibling risk "is entirely logical because we're diagnosing autism more liberally."
Infants in the study were enrolled before they showed any signs of autism, such as poor eye contact and little social interaction.
The highest rates, according to the study were in infants who had at least two older siblings with autism 32 percent of them also developed autism. Also, among boys with autistic siblings, 26 percent developed autism versus 9 percent of girls. Autism is already known to be more common in boys.
The study involved 12 U.S. and Canadian sites and was published online Monday in Pediatrics.
Ozonoff said parents of autistic children often ask her, "How likely am I to have another child" with autism? She said her study provides a more up-to-date answer.
However, Ozonoff noted that 80 percent of siblings studied did not develop autism, and that the prevalence rate was an average. It may be different for each family, depending on other risk factors they may face.
Autism has no known cause but experts believe that genetics and external influences are involved. Research is examining whether these could include infections, pollution and other non-inherited problems. Ozonoff noted that siblings often are exposed to similar outside influences, which could partly explain the study results.
McMahon predicts advances in genetic science will yield more precise diagnoses and treatments. "There will probably be many autisms," he said.
But more immediately, Ozonoff's study will empower families and doctors to diagnose autism in siblings earlier when behavioral treatment has a better chance of working.
Kathleen Lanese of Kings Park, N.Y., says having one son with autism didn't make her think twice about trying to have another child, even though she knew there was a chance the second would be affected, too.
"We wanted another child and we were going to take whoever we got," said Lanese, who was not involved in the study. Still, when her younger son was a baby, she says she "watched him like a hawk" for autism signs. He was diagnosed with autism at 16 months, earlier than her older boy.
But Ozonoff said the study should prompt families and their children's doctors to be vigilant with infants whose older siblings have autism. Early diagnosis is important because experts say behavioral treatment has the best chance of working if started early.
"Pediatricians need to listen and make a very focused plan for how to monitor those things, rather than taking a wait-and-see attitude" toward children with autistic siblings, Ozonoff said.
Alycia Halladay, a research director at the advocacy group Autism Speaks, said the study provides a more robust, accurate prevalence estimate than previous studies, and strengthens the idea that family history is a risk factor.
Her group, the National Institutes of Health and the Canadian Institute for Health Research are among those who paid for the study.
Tribune reporter Kirsten Stewart contributed to this report.
On the Web
O For more news and analysis:
American Academy of Pediatrics > aap.org
National Institutes of Health > health.nih.gov/topic/Autism