A 9-year-old looks out the window at the Autism Training Center in Illinois.… (Chuck Berman, Chicago Tribune )
If parents have a child with some form of autism, there is a 19% chance that their next child will have autism too, according to a new report.
Experts said the study, published Monday in the journal Pediatrics, is the most comprehensive of its kind and should be used to counsel families.
"Parents are concerned — could this happen again?" said study leader Sally Ozonoff, a professor of psychiatry and behavioral sciences at the UC Davis M.I.N.D. Institute in Sacramento.
Although researchers have long known that autism runs in families, most previous estimates of the risk of recurrence range from 3% to 10%. Those estimates have been based on small numbers of families and an older, narrower definition of autism than the one used today.
In the new study, researchers in 12 locations across the U.S. and Canada followed 664 infants who had at least one older sibling with autism.
The infants, on average, were 8 months old when their parents enrolled them in the study — before the signs of autism are usually apparent. At age 3, each child was evaluated.
Among the boys, 26% were found to have some form of autism. The rate for girls was 9%. The rates were the same regardless of the gender of the older sibling and the severity of that child's case.
Having more than one older sibling with autism further increased the chance of diagnosis to 32%.
The calculations are averages, and the risk for any given family could be higher or lower, experts said.
"I have seen families with five kids with autism," said Dr. Fred Volkmar, a child psychiatrist and autism expert at Yale University who was not part of the study.
The scientists have collected DNA samples from many of the children in the study to look for genetic differences between the sibling pairs who had autism and those who did not. The aim is to develop a clearer understanding of the genes involved in autism — and potentially develop tests that would measure risk in individual families.
To a limited extent, scientists can already predict risk from DNA, said Dr. Daniel Geschwind, an expert on autism genetics at UCLA who was not involved in the study.
Up to 15% of autism cases have been linked to specific genetic mutations. If one of those mutations is present in both an autistic child and a parent, the risk to the next child is considerably higher than the rate found in the study. But if the mutation occurs only in the autistic child and not in a parent, that means it arose spontaneously, and future siblings are probably in the clear.
Geschwind and other outside experts said the study provided further evidence for the strong role of genetics in autism.
Scientists already know that the chance of both twins having autism is far higher when twins are identical than when they are fraternal. The new study found the autism risk for ordinary siblings to be on par with the rates reported for fraternal twins. This is what scientists would expect for a genetic disorder, because fraternal twins and ordinary siblings share half their DNA, on average.