Can a blood test predict if a sibling will also have autism?

For parents who already have a child with autism, there’s usually an anxious waiting period to see if the next child will develop autism, one where every developmental milestone is tracked and the often-curious behaviours of infants and toddlers scrutinized.

That’s because the average risk for autism in the general population is about 1.5 percent, but that jumps to an estimated 7 percent to 20 percent for babies who have an older sibling with autism.

Canadian researchers are hoping a genetic test could help predict whether the child will go on to develop autism or not.

“We hear from parents all the time how challenging it can be to access a diagnosis and access intervention because sometimes the early signs of autism can be quite subtle,” says Lonnie Zwaigenbaum, a professor of paediatrics at the University of Alberta.

In a study published earlier this month led by Zwaigenbaum and Stephen Scherer, director of the Centre for Applied Genomics at the Hospital for Sick Children in Toronto, researchers analyzed the DNA of 288 infants with older siblings who had already been diagnosed with autism. They looked for genetic differences and narrowed down 15 variants that they considered to be associated with the condition.

Out of the 288 kids, 4.5 percent, or 13 infants, carried these genetic variations. Six of those were diagnosed with autism by age three, while five others were developing atypically. However, the genetic variations weren't present in all kids who were later diagnosed with autism—36 percent of the infants were later formally diagnosed—and so the test can't reliably predict the condition generally.

“The sample size is too small to investigate how well a test can predict autism,” says Cecile Janssens, an epidemiologist at Emory University in Atlanta, Georgia, who researches the application of genomics research to health care. “If they replicate this study in a bigger group of kids, like a thousand kids or two thousand kids, the test could easily end up being more or less predictive,” she says.

Still, Zwaigenbaum says the test, which is actually the standard genetic test that’s already being offered for children who are being assessed for developmental delays or autism, “would help identify that small percentage of kids who do have a genetic marker that would signal a very high risk of autism. That would allow the parents to more carefully monitor those younger infants and be ready to start intervention as soon as any differences emerge,” says Zwaigenbaum.

Scherer and Zwaigenbaum are continuing to investigate genetic predictors of autism, and are currently leading a study that goes beyond the 15 variants and involves analyzing the whole genome in younger siblings of kids with autism. This will give a much bigger piece of the picture, as it’s estimated that hundreds of genes are involved in autism, and could increase the test’s accuracy.

Their reason for wanting to crack the genetic code for autism is simple: Research shows the earlier someone gets access to autism therapies, the more successful those therapies will be. By identifying infants and starting behavioural and social interventions at younger ages, “there’s tremendous potential to lessen the challenges associated with autism as they get older,” says Zwaigenbaum.

In general, though, Janssens believes a reliable test for autism is a long way off, if even possible. “For psychiatric diseases, the genetic predisposition is extremely complex. Your environment plays a role and there is such a variety in how these diseases present in people. The link between the genes and symptoms is very fuzzy,” she says. “I’m not very optimistic about genetic prediction tests for psychiatric diseases because of the complexity.”

Read more:
My son has autism, but I refuse to let it limit our family
Does my child have autism?: 10 red flags to watch out for