Skip to main content

People with autism were 40 per cent faster at finishing an intelligence test that measures reasoning than volunteers without the disorder, a new Canadian study has found.

Using a brain scanner, the scientists also discovered that their autistic subjects used different parts of their brain to solve problems. They said the work, published in the journal Human Brain Mapping, could lead to new ways to help people with autism learn.

"By studying how they solve this test, we could find a way to teach them and present information in a way they could process. I think it could fundamentally change the way they are taught," said Isabelle Soulières, the lead author of the paper, who is now at Harvard University doing a post-doctoral fellowship.

The non-verbal test that was used in the experiment measures problem-solving and learning skills. In one problem, subjects were given a diagram of dots and lines with a missing section. They had to pick the correct combination of dots and lines from eight options to fill in the blank space.

Dr. Soulières said people with autism relied on visual processing, and found the right answers more quickly than members of the control group, who were more likely to explicitly test potential solutions until they found the right one.

The work is still in the early stages, and a new teaching method would have be developed and tested, Dr. Soulières said. The current practice involves breaking tasks down into small components, she said.

But when teaching someone with autism to read the sound "ea," for example, it might be better to present a page with 100 words like peach, teach and reach to show the visual pattern.

Her co-authors include Laurent Mottron, a respected autism researcher at the University of Montreal, and Michelle Dawson, who has autism, with whom Dr. Soulières has collaborated on several studies that show the intelligence of people with autism is not being accurately measured.

Autism and related conditions, known under the catch-all term autism spectrum disorders, have become increasingly common in recent years and affect communication and social interaction. The brain disorder starts in early childhood and persists into adulthood; research suggests that only about 5 per cent of adults with autism are self-supporting.

Traditionally, about three-quarters of people with autism have been classified after testing as having low intelligence. But Dr. Mottron and his colleagues have found that this depends on the type of test used. People with autism fare poorly on a standard IQ test known as the Wechsler scales, in which even the non-verbal questions require verbal instructions.

But another widely used test, Raven's Standard Progressive Matrices, is a more appropriate way to measure the intelligence of people with autism because it is as independent of language as possible.

Many people with autism - although not all of them - score much higher on the Raven test than on a standard IQ test, the University of Montreal team found. In one case, a research subject deemed mentally retarded scored in the 94th percentile on the Raven test. That's the intelligence level of a university student.

In this experiment, the researchers wanted to know if people with autism use different parts of their brain on the intelligence test. So they asked 15 people with autism and 18 other volunteers to take the Raven test in a brain scanner. Both groups practised in a mock scanner so they would be comfortable during the experiment.

The most striking part of the study, Dr. Mottron said, is that when the questions got harder, the part of the brain involved in visual processing became increasingly active in the people with autism.

"They have a special kind of intelligence for doing this task." On average, the 15 people with autism, who were between the ages of 14 and 36, didn't score any higher than the 18 members of the control group, but finished much more quickly.

Dr. Mottron said it is important to explore new ways of teaching people with autism because there isn't strong scientific evidence that the behavioural therapies currently offered are effective.

That's a controversial idea among families affected by autism. Sam Yassine, an Ottawa father who has a 10-year-old son with autism, is worried that the latest research will be used to support the argument that people with autism are simply wired differently, and don't need specialized therapy. Behavioural therapy helped his son, who as a three-year-old was isolated and spent hours watching a spinning wheel. Now he has friends and is doing well in school.

He said he agrees that people with autism have a different kind of perception, intelligence and visual skills, but adds that behavioural therapy has made an enormous difference in his son's life.

"It is the most effective way to teach children with autism," he said.