Skip navigation

Calls to Action

Improving autism therapies by exploring the roots of social avoidance

Post by Katherine Stavropoulos, a  2012-13 Autism Speaks Dennis Weatherstone Predoctoral Fellow. Today, Dr. Stavropoulos is a clinical psychologist and neuroscientist at the University of California-Riverside. Her recently published research findings grew out of her Weatherstone research project on “Motivation and Autism.

In my work with children and families affected by autism, I see the behavioral challenges that bring them to our clinic for help – and how these challenges affect their daily lives. Some of the most common involve social interaction. Children with autism tend to be less socially communicative than other children. For example, many of them show little interest in initiating conversations, shared play and other types of social interaction. Similarly, they may fail to respond to social overtures from others.

What discourages social interaction?
We use behavioral interventions to encourage social behavior, and we base these approaches on different ideas about why people who have autism tend to struggle with social interactions. My Weatherstone fellowship project launched my research into two seemingly different ideas about why many people with autism avoid social interactions – with the goal of improving and personalizing how we help them.

The social-motivation hypothesis
Most of us get a hit of dopamine – a chemical that affects the brain's reward and pleasure centers – when we interact with other people, whether it’s making eye contact or sharing some good news. In other words, it feels good to be social. The “social-motivation hypothesis” proposes that people with autism simply don’t experience this rewarding pleasure. As a result, they don’t go out of their way to engage with other people.

The overly intense world hypothesis
Another idea focuses on sensory experiences. We know that many people with autism are hypersensitive to everyday stimuli like sights and sounds. The “overly intense world hypothesis” proposes that many people with autism avoid social interactions because they find them overwhelming. After all, most of us wouldn’t want to talk to someone who seems to be screaming, especially in a room that already felt overwhelming bright.

As a practicing therapist and a neuroscientist, I wanted to see if these two explanations might both apply to children on the autism spectrum. In our most-recent study, we used noninvasive brain monitoring (electroencephalography, or EEG) to measure brain activity in 43 participants, ages 7 to 10. Twenty had autism, and 23 were developing typically.

A game with a hidden purpose
We asked each child to wear an EEG cap with 33 electrodes. (See image at right.)

This allowed us to monitor activity in different parts of the brain while the children played a computerized guessing game along the lines of “pick a hand.” We told them to guess which box in a pair of boxes was the right one, while the computer randomly assigned which box was “correct.”

During the “social cue” segment of the game, a correct guess produced a smiling face and a wrong guess produced a frowning face. During the “non-social cue” segment of the game, a correct guess produced an upward pointing arrow that the computer created from the scrambled image of a face. The wrong guess produced a downward arrow.

Additionally, we structured the game so that once the participants picked a box, there was a brief pause before they learned whether they guessed correctly. This allowed us to measure brain activity while the kids begin to wonder, “Did I get it right?” We call this “reward anticipation.” The more rewarding something is to a person, the greater the associated brain activity while they are anticipating it.

We also wanted to measure brain activity associated with learning that they got the right answer. We call this “reward processing.”

What did we find?

The results provided some real insights. Compared to the children who had autism, the typically developing children showed stronger brain activity while anticipating the social cue (smiling face) that they’d guessed correctly.

And within the autism group, the children with severe symptoms showed greater brain activity when anticipating the non-social cue (upward arrow) than they did the smiling face cue.

Support for the social-motivation hypothesis
We see these findings as confirming and extending the social motivation hypothesis. It may be that children with autism not only feel less rewarded by social interactions. They may also find nonsocial information more rewarding than social information. This has important real-life implications if the nonsocial cues in their environment are competing for their attention – that is, distracting them from their social interactions.

Evidence of sensory overload?
As I mentioned, we also looked at reward processing – the brain activity related to seeing the cue that they had guessed correctly. We found that the children with autism showed comparatively less brain activity indicating “reward-processing” – that is, in response to the feedback that they had guessed correctly – than did the typically developing kids. At the same time, the children with autism experienced more of the brain activity we associate with paying attention to incoming information generally and reacting to it. This might suggest that they were feeling relatively overwhelmed by the sensory aspects of what they saw.

In conclusion, our study suggests that the social motivation hypothesis and the overly intense world hypothesis may both be correct. People with autism may struggle with social interactions because they find them both less rewarding and overwhelming.

Translating research into personalized interventions
Our findings may suggest that behavioral interventions for autism should aim to both increase the reward value of social interactions and be sensitive to a child’s hypersensitivities. It does no good if we have an excellent social skills intervention, but a child can’t respond to it because he or she finds the lights too bright, voices too loud or the truck braking outside too high pitched.

Along these lines, perhaps we should rethink the current practice of working on social communication and sensory issues separately. For example, a child with autism may work on social skills with a behavioral therapist on Mondays and work on sensory issues with an occupational therapist on Wednesdays. It might be better to have joint therapy sessions or at least combine and coordinate the two approaches.

Next steps in research and practice
I’m now pursuing several studies to extend these findings and help improve behavioral interventions for autism.

In one study, I’m using noninvasive brain monitoring before and after a social skills intervention to gauge how brain activity changes in response to the therapy. I hope this will help us understand why our current behavioral interventions work well for some children but not others. This understanding, in turn, could help us shape more personalized approaches to maximize benefits for each child.

In another study, I’m exploring the role that anxiety plays in autism-related social challenges. I suspect that some children may avoid social interactions due to anxiety, while others are simply less interested in social interactions. If true, this may present another opportunity to personalize our interventions to make them more effective.  

Also see “Managing anxiety in children with autism” and
Research yields tips for easing anxiety in nonverbal kids with autism,”
in the Autism Speaks “
Got Questions?” advice column.

I want to thank Autism Speaks and the Stavros Niarchos Foundation for the Weatherstone fellowship that allowed me to begin this research, further my training as a therapist and merge my passions for neuroscience and clinical practice. 


The Autism Speaks blog features opinions from people throughout the autism community. Each blog represents the point of view of the author and does not necessarily reflect Autism Speaks' beliefs or point of view.