Most siblings with a diagnosis of autism do not share the same genetic risk factors for the disorder and are as distinct in their behaviors as any brothers and sisters, scientists reported on Monday in a study that came as a surprise to many doctors, if not to parents.
Scientists analyzed genetic material from 85 families, using a technology called whole-genome sequencing. Unlike other approaches, which illuminate a sample of a person’s genetic material, the whole-genome technique maps out the entire voluminous recipe, every biological typo, every misplaced comma or transposed letter. Each of the families had two children with a diagnosis of autism.
The researchers focused their analysis on about 100 genetic glitches linked to the development of autism. They found that about 30 percent of the 85 sibling pairs in the study shared the same mutation, and about 70 percent did not. The sibling pairs who shared a genetic glitch were more similar to each other, in their habits and social skills, than those pairs whose genetic risks were different, the study found.
The finding drives home the exasperating diversity of autism, even in the most closely related individuals. And it suggests that scientists will need to analyze tens of thousands of people, perhaps more, to tell any meaningful story about its biological basis.
Experts said the report, in the journal Nature Medicine, would most likely encourage changes in clinical practice. Some hospitals analyze the genetic profile of the eldest affected sibling to try to understand an infant, or to advise parents of the odds of having another child with the same disorder. That approach is not informative in most cases, the study authors said.
“This is very important work for people like me,” said Valerie South, an emergency nurse in Oakville, Ontario, outside Toronto. Her sons Cameron, 20, and Thomas, 14, have severe autism. In a family of four or more, the odds of having two children with autism is about one in 10,000.
In 1998, she and her husband consulted doctors about the risks of having another child with autism. At the time, they had Cameron and an older son, Mitchell, who has no developmental problems.
“We wanted Mitchell, who is a great brother, not to have to carry the burden all by himself,” Ms. South said. “They told us that the odds of having another Cameron were tiny. And even if it happened, they said, it wouldn’t be a severe case.”
They did not get another Cameron — or another Mitchell. They got Thomas, who shares Cameron’s autism diagnosis, but who is very different. Thomas will walk right up to strangers, Ms. South said, while Cameron draws back. Thomas loves his iPad, whereas Cameron has no interest in computers. Thomas is an “escape artist,” continually on the move, while Cameron prefers to park himself in one place.
The new study provides a good biological explanation for those differences — and should put to rest the sort of predictions the Souths got before having Thomas.
“This study makes us step back and realize we’re not necessarily going to get as much predictive value out of genetic mapping as we thought,” said Helen Tager-Flusberg, a developmental neuroscientist at Boston University, who was not involved in the research.
The report is the latest twist in a genetic plot that seems only to thicken. In recent years, scientists have isolated gene mutations that steeply raise the risk of autism. But those glitches account for only a tiny fraction of cases, and the number of them continues to increase — to about 100 now, and counting.
Scientists have not been able to tell a coherent story about causation. It could be that common variants — that is, gene variations that many people carry, which cause no apparent problems — increase the dormant possibility for the disorder, in some combinations.
The investigation focused on 85 families with two children on the autism spectrum. “We anticipated that, more often than not, there would be shared inheritance” in siblings, said the project’s research director, Dr. Stephen Scherer, a professor of medicine at the University of Toronto. Dr. Scherer is also director of the Center for Applied Genomics at the Hospital for Sick Children. “That wasn’t the case.”
Experts not involved in the study said the finding was convincing and important. “The study is very well designed, the end result is somewhat surprising, and it reiterates the complexity of the underlying genetics of autism,” said Dr. Yong-hui Jiang, an associate professor in the department of pediatrics and neurobiology at the Duke University School of Medicine.
The report is the first major finding from a large-scale collaboration financed by Autism Speaks, an advocacy group, and based at the University of Toronto and the Hospital for Sick Children. The researchers are analyzing genetic material from a registry of more than 2,800 families who have at least one child with a diagnosis. The registry, funded by both the National Institutes of Health and Autism Speaks and managed by Autism Speaks, includes comprehensive medical and behavioral histories and is the largest of its kind.
In a first, the study coordinators, working with Google, published the data and analysis tools online, in a cloud-based format available to any user. In previous genetic studies, it has typically taken researchers months or years to make their data available, if they do so at all, Dr. Scherer said. Autism specialists said that the study findings were a welcome confirmation of a common observation. “After you get over the surprise and think about it, we all know that the kids are different in these families,” Dr. Tager-Flusberg said.
The ratio of siblings with autism who have a common genetic basis for the disorder versus those who do not — now 30 percent to 70 percent — may be less lopsided as researchers analyze more such families, some experts said. The study “used a quite liberal approach for variants relevant to autism, and the finding, while solid, may be less general than implied,” said Dr. David B. Goldstein, director of the Institute for Genomic Medicine at Columbia University.