Scans find autistic brain changes

Friday December 3 2010

“A foolproof test for autism in adults and children is “a major step” closer,” The Daily Telegraph has reported. The newspaper says that the new brain scan “can detect the condition with almost 100 per cent accuracy”.

The news is based on research that compared brain scans of 30 male children and adults with high-functioning autism with 30 matched males with the same IQ but typical development. Scientists found that using magnetic resonance imaging (MRI) scans to examine enhanced details of brain cell fibre patterns allowed them to detect subtle structural differences in the brains of the two groups, specifically within two areas associated with language and social cognition. When they concentrated on six characteristic differences the scientists were able correctly to identify a person with autism 93.6% of the time and to identify individuals without autism correctly 89.6% of the time.

Although this research is a useful preliminary step in finding a biological measure that can distinguish between autistic individuals and typically-developing individuals, further research is needed to see whether these findings apply to a broader group of people with autism, such as those with more severe autism, women and younger children.

Where did the story come from?

The study was carried out by researchers from Harvard Medical School and was funded by The National Institutes of Health in the US. It was published in the peer-reviewed medical journal Autism Research.

The research was covered accurately by The Daily Telegraph , although further validation is required before this method could be used as a clinical test. The Daily Mail reported that a brain scan has been developed that could “diagnose autism in children in just ten minutes”, allowing the condition to be detected at a much younger stage so children could begin therapy and treatment before they start school. This research only conducted MRI scans on individuals over seven years of age, including fully grown adults. Therefore, it is not possible to say whether this technique would be able to detect autism in younger individuals as successfully as conventional methods

What kind of research was this?

This was a laboratory study that looked at whether it was possible to use MRI scanning to find differences between the brain structures of people with autism and unaffected individuals. The researchers said that, unlike psychiatric or psychological assessment, biological measures are not yet clinically adequate to determine whether a person has autism.

The researchers were particularly interested in looking at the structure of the brain’s white matter. This contains the connecting fibres of brain cells and shows up as white on MRI images. The researchers said that they were particularly interested in two areas of the brain, called the superior temporal gyrus (STG) and temporal stem (TS). These contain the white matter fibres critically involved in language and social cognition. They had previously found differences between control subjects and people with autism in these areas and wanted to see whether it was possible to use only these differences to discriminate between individuals with autism and typically-developing individuals.

What did the research involve?

The researchers made measurements of the white matter of 30 high-functioning, right-handed males with autism (who met standard diagnostic criteria and had a performance IQ of more than 85) plus 30 typically-developing males who were matched for age, IQ, right-handedness and head circumference. The individuals were between 7 and 28 years of age.

Some of the individuals with autism also had depression (56%), attention deficit disorder (31%), obsessive compulsive disorder (25%) and anxiety disorder (19%). Sixty-three per cent of individuals with these conditions were taking one or more psychotropic medications, such as antidepressants or stimulants.

To focus on the white matter microstructure, the researchers used an MRI technique called diffusion tensor imaging (DTI), which measures the direction of local water diffusion in brain tissue (water diffuses more rapidly in the direction of the brain cell fibres). They made various measurements of the diffusion properties of water in these areas in order to tease out structural differences between the individuals.

What were the basic results?

The researchers found that individuals with autism showed differences in the diffusion of water in the STG areas found on both sides of the brain, and in the TS on the right-hand side only, compared with typically-developing individuals. In total, there were six differences in water diffusion patterns.

When the researchers used these six differences to discriminate between individuals with autism and typically-developing individuals, they found that they were able correctly to identify a person with autism 93.6% of the time (the sensitivity of the test). They could correctly identify individuals without autism (i.e. rule out autism) 89.6% of the time (the specificity of the test). Overall, they found that the test was 83.3% reliable.

The researchers found that the volume of the white matter area did not aid classification.

How did the researchers interpret the results?

The researchers say that analysing the properties of the white matter microstructure in the STG and TS has a high ability to separate individuals with autism from typically-developing individuals, and that this provides evidence that the STG and TS are involved in the neurobiology of autism. They say that extension of their study to larger sample sizes and to individuals with higher severity of autism, younger children and females is now necessary.


This preliminary research in a small group of 30 individuals with high-functioning autism shows that there may be subtle differences in the microstructure of white matter in two brain areas in individuals with high-functioning autism compared with typically-developing individuals. However, follow-up in a much larger group is needed to see how well these differences actually predict whether a person has autism or not. The researchers acknowledged some limitations of this work:

  • The study only looked at individuals with high-functioning autism, only one type of developmental disorder, and further research is needed to see whether the differences they saw are specific to high-functioning autism or whether individuals with other developmental disorders may also have these subtle differences in brain structure.
  • The study only looked at right-handed males aged seven years and above. Further studies need to assess whether the findings apply to infants and young children and women, and also to individuals who have more severe forms of autism.
  • The researchers say that, alongside larger follow-up studies, longitudinal studies are also needed to see how these brain areas may change with time.
  • Presently, looking at water diffusion using MRI can provide some information about brain microstructure, but the resolution is still relatively low. There may be a variety of brain cell types within an area that cannot be anatomically distinguished by this technique as yet.
  • Some of the autistic groups were taking medications for psychiatric conditions, and it is not clear whether psychotropic medication could affect white matter microstructure, so this needs to be tested.
  • While such a test might eventually be used to diagnose autism, the process is likely still to require psychological and behavioural assessments to identify the best support and intervention strategies for individuals with autism.

This study also indicates that two brain areas called the superior temporal gyrus and temporal stem may be associated with autism, but the nature of this association requires further research. The imaging technique used in this study suggests that there may be subtle differences in the microstructure of cells in the white matter, but further work is also needed to determine any fine anatomical differences in autism.

All in all, this research forms a well-conducted, preliminary step towards finding a biological measure that could distinguish between a specific group of autistic individuals and typically-developing individuals. Further research is now needed to see whether these findings apply to a broader group of individuals with autism.

Analysis by Bazian
Edited by NHS Choices