Testing Baby's Brain

As far as her friends and teachers are concerned, Ashdod is an ordinary, bright, playful 5-year-old. They might be surprised to learn that not long ago therapists were fighting to keep her from succumbing to autism—a brain disorder that afflicts one in 100 children, typically leaving them with lifelong difficulties in communicating, socializing and carrying out many basic tasks. Ashdod was lucky: when she was 10 months old, her parents became alarmed that she had little interest in looking them in the eyes, eating and moving from her back, and whisked her to the Mifne Center in Rosh Pinna, Israel, a clinic that focuses on children 5 months and older who show early warning signs of autism. The results of the Mifne treatment were startling, recalls the girl's mother, Tikva. "Now she goes to a regular school where she is the same sort of articulate, funny, normal child as anyone else," she says.

Despite a big jump in autism awareness in the past decade, parents, schools and doctors still frequently ignore warning signs in very young children. These can be subtle: a child never points at things, shows more interest in objects than people, has delayed speech and develops a fascination with spinning in place or with spinning toys. Many pediatricians dismiss these symptoms as harmless quirks that kids will outgrow. New research and experience in some autism clinics, however, suggests that starting treatment by age 2 is critical to mitigating and in some cases entirely avoiding the disorder.

That's because unlike the brain of an adult or even an older child, a 12- or 18-month-old's brain is, in a sense, highly reprogrammable—that is, it responds well to treatments designed to permanently change basic patterns of thought and behavior. "All the evidence we have suggests that outcomes for these children will be better with an earlier diagnosis, before they reach 18 months, if possible," says Christopher Gillberg, a child psychiatrist at Gothenburg University in Sweden.

Although there are currently no effective treatments for autism symptoms in older children or adults, the prospects are turning out to be entirely different for very young children who get prompt treatment. Psychologists have had remarkable success with behavioral therapy, which involves therapists working intensively with children to get them to do tasks they're having difficulty with. The Mifne Center in Israel applies its own form of intensive therapy, typically lasting about two weeks and focusing on getting the child to make contact with parents and to eat and move normally. Some 200 children have been through the program; about three quarters have remained free of any signs of autism or any other significant developmental disorder, according to Mifne founder and director Hanna Alonim. "If we can get them here as babies, close to 100 percent won't develop autism," she says. "If we don't see them until they're 2½, it's a different story." To bolster Mifne's findings with more-formal research, clinicians at the Tel Aviv Sourasky Medical Center have begun screening and videotaping infants thought to be at risk of pre-autism before the Mifne treatment.

Having a treatment option for infants raises the thorny issue of diagnosis. Autism can be tricky to recognize—it encompasses any or all of a broad range of symptoms, including difficulty with social interactions, language, motor skills and taking in sensory information, as well as repetitive behaviors, eating problems and in some cases unusually high or low levels of activity. A study of nearly 10,000 children in Bergen, Norway, indicated that the number of children who showed "pronounced autistic features" was about five times higher than the number who qualified for a formal diagnosis of autism.

Even children who exhibit only partial or mild versions of autism symptoms are at risk of ending up with lifelong challenges, say researchers, and would benefit from autism therapies. But tagging more very young children as candidates for autism therapy creates another problem. The cost of behavioral therapy is eye-opening—applied behavior analysis, an intensive treatment that requires 15-to-25 hours of sessions a week, costs about $30,000 a year, and even a modest program typically runs about $10,000 a year. That's one reason studies estimate that less than one in 10 very young children with a diagnosis of autism get 25 hours a week of therapy.

Health-care systems are not up to this task. In the United States, where health insurance rarely covers such treatments, the chances of having the government pay for therapy varies wildly from state to state. Children don't necessarily fare much better under national health care. The United Kingdom picks up the tab for treatment—but often only after the parents hire a lawyer and win their case at a regional "tribunal," where more often than not communities will fight to force the parents to settle for the few hours a week of therapy offered in a local special-education program. In Italy, toddlers with disorders who live near large cities in the north-central part of the country can get the attention of a team of therapists, but those in the southern, rural areas tend to get few services. Laurence Robel, a child psychiatrist and autism researcher at the Necker Children's Hospital in Paris, notes that France retains a bias against behavioral therapy, which critics compare to training a dog or programming a robot. "Away from Paris," she says, "children are lucky to receive much treatment at all."

Experts insist that governments are being penny-wise and pound-foolish in refusing to pony up for intensive therapy to infants and toddlers who show early signs of a developmental disorder. Simple, brief screening tests are now designed to flag children at risk as early as 18 months. Earlier diagnosis might be possible by measuring brain activity and recognizing patterns that are unique to autism. Researchers at the Baby Lab in Uppsala University in Sweden are looking for these patterns by placing dozens of soft-foam sensors on infant's heads.

The benefits of early treatment are likely to grow in coming years as new research into developmental disorders continues to pay off.