Health & Family

Study targets autism in infants

When her toddler son seemed not to notice a door slamming nearby during his checkup, Jo James thought nothing of it. A check of Ben's hearing after a nurse's prompt found nothing amiss. It wasn't until two years and one perceptive Montessori teacher later that his parents finally learned the cause of Ben's obliviousness: autism.

Autism typically isn't diagnosed until after age 2. Yet it may be detectable even in infancy – before a baby is old enough to display telltale traits such as social ineptitude and compulsive preoccupations. Now University of Washington researchers aim to decipher those early clues in hopes of short-circuiting autism before it becomes full-blown.

In January, they began an $11.3 million trial to identify latent signs of autism in infants for intensive behavioral therapy. It is the nation's first attempt to test a hypothesis that early intervention might prevent autism in high-risk infants by rewiring their brains.

“We know the brain has a lot of potential to respond” to the right stimulation, said Sara Webb, the principal investigator for the study. The goal is “to teach parents to give the child that missing piece that he's not getting on his own.”

The study's premise – that autism may not be destiny – has stirred unease and skepticism among some parents.

But for University of Washington researchers the goal is akin to averting diabetes through vigilance in a person with a family history of the disease, said Annette Estes, a study investigator. “If you are at risk for diabetes, you look for signs,” Estes said. With autism, the genetic “risk factors are present at birth. What we are doing is heightening the parents' awareness.”

Though some parents report concerns early on, tiny babies by definition don't have autism. That's because they can't manifest such diagnostic symptoms as language deficits and repetitive rituals.

Yet researchers suspect that babies exhibit subtle clues that precede overt symptoms. For instance, healthy babies react to changes in a person's expression. A baby who doesn't seem to register the change may warrant watching, Estes said. “At 6 months, a baby has a limited repertoire of signs” of autism, she said. “The question is, ‘What are the real early signs?'”

To answer that, researchers will track 200 infant siblings, 6months or younger, of kids with autism. The disorder's strong hereditary nature means that 10 or more of those babies will develop autism themselves. The odds for a typical American child are a little greater than 1 in 200.

The infants will be randomly separated into two groups. Mothers in one group will be coached on engaging with their babies. Later, the children will receive up to 25 hours a week of a type of developmental intervention called the Denver Model that uses play to teach appropriate behaviors.

The other group will be monitored but treated.