Surprising autism finding

Dec 13 2005

New brain imaging research at the University of North Carolina at Chapel Hill indicates that when people with autism look at a face, activity in the brain area that responds is similar to that of people without autism.

The finding is surprising, as it is widely known that autistic individuals tend to avoid looking directly at faces. The research also counters previous published reports that the face-processing area at the back of the brain is under-responsive in people with autism, and it suggests that specific behavioral interventions may help people with autism improve their ability to interact socially.

The study was conducted by Dr. Aysenil Belger, associate professor of psychiatry in UNC's School of Medicine and of psychology in UNC's College of Arts and Sciences; and Dr. Gabriel Dichter, postdoctoral research fellow within UNC's Neurodevelopmental Disorders Research Center.

The study involved functional magnetic resonance imaging, or fMRI. Unlike standard MRI scans that show anatomical structures in black and white, fMRI offers digitally enhanced color images of brain function, depicting localized changes in blood flow and oxygenation.

When particular regions of the brain increase their neural activity in association with various actions or thought processes, they emit enhanced blood oxygen level dependent signals. The signals can be localized in the brain and translated into digital images that portray neural activity level as a ratio of oxygenated to de-oxygenated hemoglobin, the iron-containing pigment in red blood cells. Researchers then can quantify these signals to generate maps of various brain functions.

"The brain regions 'specialized' for face processing, the fusiform gyrus, activated almost identically in our autistic study participants and the control group of individuals without autism. This is one very simple and clear-cut finding," Belger said.

Previous brain imaging findings of under-responsiveness in the facial area have led some experts to consider this a key cause of the social impairments observed in autism. But Belger and Dichter said they thought the problem might have more to do with a deficit in "executive function" in portions of the brain's frontal lobes than with a selective deficit in the brain's processing of information from faces.

These portions of the frontal lobes are where "executive" tasks, such as sifting through complex information, selecting task-appropriate responses or inhibiting inappropriate ones, take place. "In other words, these tasks require high levels of cognitive control," Belger said.

"We wanted to find out if these executive functions of particular brain regions that are critical for the appropriate generation of actions show deficits in autistic individuals. In addition, we wanted to see if these deficits were further compounded by the presence of social cues on which executive decisions had to be based," Belger added.

The study compared fMRI scans of people with autism and those without the disorder as they performed a one-hour task that required them to attend to certain items in the visual environment while disregarding other items. Specifically, participants wore special goggles through which were displayed pictures of arrows (non-social items) and faces (social items) and were asked to report the direction of the central arrow or central face by pushing a left or right button.

Participants were shown horizontal rows of five arrows or five faces. Sometimes all the arrows and faces were oriented in the same direction. At other times, the central arrow or face was in the opposite direction.

The test with arrows alone had been used previously in studies elsewhere, but the addition of faces was unique to this new study.

"We wanted to know if previous findings of under-responsiveness in the 'face area' of the brain in autism could be simply attentional, or related to an inability to willingly direct attention to the pertinent feature of a face, gaze direction in this instance," Dichter said, adding that most such studies did not account for where the participants fixed their gaze during brain scanning.

Because individuals with autism typically choose to look away from faces, the previous studies could not disentangle the effects of a preference to look away from faces from actual brain deficits in the ability to process faces. "We required participants to indicate whether faces were looking left or right, thereby requiring that they look at the picture of the face," Dichter said.

In addition, Belger and Dichter pointed to research that correlated the amount of time autistic people looked at features of a face with activation of the amygdala, the "anxiety center" of the brain.

They propose that the use of behavioral interventions aimed, perhaps, at de-sensitizing autistic individuals to the anxiety triggered by looking at faces may help to improve the quality of social interactions, at least from the perspective of caregivers and others involved with them. "The idea is that maybe behavioral methods can help a child with autism approach facial stimuli without anxiety. This is strictly conjecture, but it is a tantalizing possibility," Dichter said.

Imaging allows researchers to begin looking at underlying mechanisms in brain function, Belger said. "Still, we must be careful in our interpretation of the results. Our findings need to be replicated and the research also broadened to include longitudinal studies," he added.