From the Labs
Houston, Texas
Volume 7, Issue 2
March 2008

Autistics' sense of self impairs broader understanding

By Ruth SoRelle, M.P.H.

P. Read Montague, Jr., Ph.D.
P. Read Montague, Jr., Ph.D.

The common perception of people with autism is that they cannot relate to others.

However, that perception may stem from their difficulties in recognizing "self," said researchers from Baylor College of Medicine in a report that appeared in a recent issue of the journal Neuron.

"When an individual cannot model self, he or she has an impaired ability to understand the world itself," said P. Read Montague, Jr., Ph.D., professor of neuroscience, and director of the Human Neuroimaging Lab and the Computational Psychiatry Unit at BCM. "It's an interesting disconnect."

Tying brain activity to symptoms

Using a functional magnetic resonance imaging (fMRI) scanner, Montague and his colleagues scanned the brains of people considered "high functioning" autistics because they have normal or high normal intelligence quotients but many of the symptoms of people with autism. During this procedure, the researchers identified a pattern of activity or "signature" in the brain that identified those with autism. The level of activity correlates with the severity of the autistic symptoms. The less activity there is, the more serious the symptoms. The finding could lead to a test to speed diagnosis.

Pearl H. Chiu, Ph.D.
Pearl H. Chiu, Ph.D.

"We are very excited about the usefulness of the hyperscanning technology and economic games as new tools to probe autism. Our hope is that these same approaches can be used to probe a wide range of psychopathologies," said Dr. Pearl Chiu, first author on the study and an assistant professor in the departments of neuroscience and psychiatry and behavioral sciences at BCM.

Hyperscanning

To understand the behavioral patterns of people with autism spectrum disorder, Montague and his colleagues used a technique called hyperscanning, which enabled them to scan two brains simultaneously while the research subjects played a trust game. Hyperscanning was developed in Montague's laboratory.

In the trust game, one player receives an amount of money and then sends whatever amount he or she wants to the other player via computer message. The amount sent is tripled and the player at the other end then decides how much of the tripled amount to send back. The game has several rounds.

During this interaction, Montague and his colleagues evaluated the brains' response by watching bright spots in the brain that represent increased blood flow and thus brain activity. Prior work had shown that during the trust game, most of the activity occurs in an area called the cingulate cortex.

Study structure

brain scan from fMRI
Image of the 'self' response in the cingulate cortex

To hone their picture of the "self" response in that part of the brain, Montague and his colleagues had 81 athletes (football, baseball and soccer players as well as members of the Houston Ballet) take part in an imagining task. They watched clips of various athletic activities while in the scanner. They then imagined themselves performing those activities. The pattern of activity in the cingulate cortex during that "imagining" reflected the "self" response.

Later the scientists identified the same "self" response in the cingulate cortices of normal subjects when they decided how much money to send to the other person. The pattern contrasted with the "other" response seen when the actions of their partner in the trust game were revealed to them.

"We would not have really known what we were looking at had we not scanned these athletes," said Montague.

In a previous study published in the journal Science, Montague and his colleagues showed that outside factors are unlikely to affect this response.

"The amount of money, whether you are male or female, whether you are playing the game personally or impersonally does not matter. This tissue and this response pattern in this tissue is activated when you are playing a human being or at least an agent that plays enough like a human that your brain believes you are playing a human," said Montague. "This tissue seems to be devoted to that, at least in humans."

The researchers then brought in 18 adolescent males with high functioning autism to play the game. Montague and his colleagues determined that the subjects understood the game and helped them adjust to the scanning procedure.

"It was the first time an autistic kid had been scanned in a social exchange," said Montague.

Cingulate cortex differences

The adolescents did not play the game differently from their partners, who were taken from a population of similar teens who did not have autism. They made similar amounts of money overall and round by round.

However, when the researchers scanned the brains of the youngsters with autism during the trust game, they found that the youngsters' "self" responses were dim compared to those of normal subjects. Not only that, but the more serious the subject's autistic symptoms, the dimmer the response.

The response occurred in the cingulate cortex. In a normal "self" response there, the brightest area was in the middle of that area of the brain. That response was significantly less in the brains of the youngsters with autism.

"They cognitively understood the game," said Montague. "It's not that they don't understand the game. It's that there is a very low level of 'self' response. It's impaired in them and the degree to which it is missing correlates with their symptom severity. The more you are missing the self response, the more autistic you are.

"To have a good self concept, you have to be able to decide if the shared outcome is due to the other person or due to you," said Montague. "If people can't see themselves as distinct entities at deeper levels, there is a disconnect."

Benefits of early diagnosis

Would having a way to recognize this problem early help? Montague asked himself that question, but the response of one mother assured him that there was benefit. She was interviewed about the benefits of the test.

"She told this long story. Because her son was so high functioning and smart, he could handle school tasks, but he could not handle everything else. He did these repetitive behaviors. He did know that the other kids would pick on him. He would overreact. First he was a behavior problem. Experts told the family, 'We think he has this.' Then it was another thing.

"She said if she could have a definitive diagnosis early on, it would have changed a whole lot. He was not diagnosed until he was 11. Then there were things they could do to structure his life and environment to help him have a more normal life."

He believes that the problem occurs at an unconscious level. He hopes to use the imaging technique in the future to scan the brains of people with autism whose intelligence is less than that of those in this first experiment.

Marrying information from the trust game with information gleaned from the athlete studies indicates that maybe the "self" response can be measured outside of the interactive activity, said Montague. "You can turn this thing on by just looking at a picture of yourself or other people who are familiar to you. You get the same phenotype (response that can help determine the severity of symptoms)."

"The mantra has been that people with autism don't respond to others right," said Montague. "That means they can't build appropriate emotional models of others. But if they can't cut themselves off as being distinct from others, then that would never even start. That a little bit of surprise in this finding and that's why you do science."

"The genius of this study was to recognize the primary deficit with this spectrum of disorders," said Dr. Michael Friedlander, chair of neuroscience at BCM. "Then they took that information dealing with the deficit social cognition and invented a new kind of experiment and technology to probe it."

Others who took part in this study included Drs. Pearl H. Chiu, M. Amin Kayali, Kenneth T. Kishida and Damon Tomlin of BCM and Drs. Laura G. Klinger and Mark R. Klinger of the University of Alabama at Tuscaloosa.

Funding for this work came from The Kane Family Foundation, The Dana Foundation and Autism Speaks, the National Institute on Drug Abuse, the National Institute of Neurological Disorders and Stroke, The Angel Williamson Imaging Center, and the American Psychological Association.