Charlie Lamb was barely 2 years old when he was diagnosed with autism. His mother Susan had been convinced for months that "something was not right" with her second child. He wouldn't stand in line like the other kids in gymnastics class, she recalls, and he spoke fewer words. He was more captivated by spinning wheels than Teletubbies. His father Tom noticed that his blond, blue-eyed son would always walk in circles around the kitchen table and that he would do the equivalent at their local park in Seattle walking along the perimeter fence rather than crossing into the play area.
Ten years ago, autism was rarely detected before ages 3 or 4. Now, thanks to growing awareness and widespread screening at 18 and 24 months, as recommended by the American Academy of Pediatrics, more autistic children like Charlie are being identified when they are toddlers. But for all the emphasis on early detection, very little research exists on how to intervene effectively for children so young.
A report in the current issue of Pediatrics helps fill in the gap, providing the first randomized, controlled trial the most rigorous kind of study of a comprehensive autism treatment that appears to work well for children as young as 18 months. While none of the children in the study were "cured" of autism, those receiving two years of intensive therapy achieved major leaps in IQ score, big improvements in their use of language and significant gains in their ability to handle the kinds of everyday tasks necessary for a child to function at school and at play.
"We were extremely pleased to see how effective this was," says Geraldine Dawson, lead author of the study and a professor of psychiatry at the University of North Carolina at Chapel Hill. "This really underscores the importance of early detection and very early intervention for autism," says Dawson, who is also the chief science officer for the advocacy group Autism Speaks.
The study, conducted at the University of Washington and funded by the National Institute of Mental Health, involved 48 children ages 18 to 30 months. Half were randomly assigned to receive an intensive intervention called the Early Start Denver Model, which involved 15 hours a week of one-on-one work with trained therapists and another 16 hours a week with parents, who were taught how to continue the treatment during everyday activities such as eating, bathing and getting dressed.
The other 24 children were assigned to a control group and referred to a variety of therapists in the greater Seattle area. Although they received less intensive therapy than the intervention group, they still got an average of nine hours a week of one-on-one therapy and another nine hours a week in a specialized preschool or other group setting.
After two years, children in the Denver Model group were way ahead of the control group. Their IQ scores had jumped an average of 17.6 points, to a mean of 78.6, which is just within range of normal intelligence. Much of the gain came in their ability to understand and use language. The control group, by contrast, gained just seven points, remaining in the zone of intellectual disability. Children who received the intervention also improved dramatically in what psychologists call "adaptive behavior" which includes such everyday behaviors and skills as getting dressed, brushing teeth and participating in family meals. Children in the control group improved much more slowly, falling further behind normal peers.
Also notable, seven out of 24 children in the treatment group improved enough to move out of the autism disorder diagnosis and into a milder part of the autism spectrum (Pervasive Developmental Disability, Not Otherwise Specified). Only one child in the control group did so.
One of the peculiarities of autism treatment is that there are many competing brands, including some therapies that are promoted as cures, but few have been subjected to rigorous study, and virtually none have been tested head to head. The Pediatrics report "brings the methodological rigor that's often been lacking," says psychologist Tony Charman, an autism researcher and professor at the Institute of Education in London, who was not involved in the study. Charman was further impressed with two features of the Early Start Denver Model: it deeply involves parents in their children's treatment an approach "for which there is good evidence" and it combines the two kinds of autism therapy for which there is the most research data.
Specifically, the Denver Model uses key elements of Applied Behavior Analysis (ABA), a rigorous system for shaping behavior by parsing desired behaviors into small, measurable and teachable units and using rewards to reinforce them. It also incorporates a more naturalistic, relationship-based approach that draws heavily on decades of research on normal child development. "We follow the sequence of normal development in everything we teach," explains psychologist Sally Rogers of the MIND Institute at the University of California, Davis, who developed the Denver Model while at the University of Colorado.
Thus, for example, rather than teaching children to speak by drilling sounds and words, Denver Model therapists begin with what they call "talking bodies" the nonverbal communication of smiles, gestures and eye contact that normally precedes speech but which toddlers with autism have missed. While therapists use ABA techniques to chart progress toward specific goals, the therapy itself "looks like play," says Rogers, a co-author of the study. "If you saw it, you would say, 'That's what I do with my own baby.' "
Charlie Lamb was among the 24 children in the treatment group. Though the first few sessions were hard ("He would scream and cry and pound on the door of his room," his father recalls), Charlie soon began to enjoy the playful therapy and made steady progress in speech and behavior. Now 5½, he attends a special preschool and continues to work with therapists on social skills and language. The Lambs expect that Charlie will ultimately attend a regular school. "His autism is subtle," says Susan Lamb. "Most people say they can't tell." But like most children with autism, Charlie suffers from anxiety and is especially vulnerable in unfamiliar situations.
He and the other children in the study will be followed by researchers for several years. In addition, Rogers and Dawson have begun an expanded version of the study involving 120 toddlers at the three sites. It's one of a number of trials involving very young children that should over the next few years bring greater clarity to autism treatment.
Whether the Early Start Denver Model will prove to be more effective than other therapies remains to be seen. Leading autism researcher Tristram Smith, an ABA expert at the University of Rochester, who lauds the new study for its methodological rigor, notes that the gains made by children in the intervention group were similar to those reported in studies of ABA models. "I do think there is a need for head-to-head studies," says Smith. Also needed is high-quality research on how to match individual children with the therapy that suits them best.
Claudia Wallis was a 2008 Spencer Education Journalism Fellow at Columbia University's Graduate School of Journalism.