Autistic children are less sensitive to the movements of living things

Blogging on Peer-Reviewed ResearchFor any animal, it pays to be able to spot other animals in order to find mates and companions and to avoid predators. Fortunately, many animals move in a distinct way, combining great flexibility with the constraints of a rigid skeleton - that sets them apart from inanimate objects like speeding trains or flying balls. The ability to detect this "biological motion" is incredibly important. Chicks have it. Cats have it. Even two-day-old babies have it. But autistic children do not.

Ami Klim from Yale has found that two-year-old children with autism lack normal preferences for natural movements. This difference could explain many of the problems that they face in interacting with other people because the ability to perceive biological motion - from gestures to facial expressions - is very important for our social lives.

Indeed, the parts of the brain involved in spotting them overlap with those that are involved in understanding the expressions on people's faces or noticing where they are looking. Even the sounds of human motion can activate parts of the brain that usually only fire in response to sights.

You can appreciate the importance of this "biological motion" by looking at "point-light" animations, where a few points of light placed at key joints can simulate a moving animal. Just fifteen dots can simulate a human walker. They can even depict someone male or female, happy or sad, nervous or relaxed. Movement is the key - any single frame looks like a random collection of dots but once they move in time, the brain amazingly extracts an image from them.

But Klim found that autistic children don't have any inclination toward point-light animations depicting natural movement. Instead, they were attracted to those where sounds and movements were synchronised - a feature that normal children tend to ignore. Again, this may explain why autistic children tend to avoid looking at people's eyes, preferring instead to focus on their mouths.

Alim created a series of point-light animations used the type of motion-capture technology used by special effects technicians and video game designers. He filmed adults playing children's games like "peek-a-boo" and "pat-a-cake" and converted their bodies into mere spots of light. He then showed two animations side-by-side to 76 children, of whom 21 had autism, 16 were developing slowly but were not autistic, and 39 were developing normally.

Of the two animations, one was a normal upright version and the other was flipped upside-down and ran backwards (previous studies have found that these tweaks disrupt a child's ability to spot biological motion). Using an eye-tracking camera, Alim could see which of the two frames the children were most drawn to.

Both the normal and the developmentally delayed children  showed small but significant preferences for the upright animations, spending 63% and 59% of their time looking at it respectively. On the other hand, the autistic children had no such biases and spent just as long looking at both animations. The fact that the developmentally delayed children behaved normally suggests that the autistic children were behaving differently because of something specific to autism, rather than a general feature of delayed mental development.

But one of the animations produced the opposite effect. When the actor in the video played pat-a-cake, the toddlers spent 66% of their time watching the upright version. For this animation and this one alone, all three groups of children behaved in the same way. Why?

The difference lay in the audio. The other animations were just accompanied by a human voice but the pat-a-cake one also included a clap, that coincided with two lights - representing the actor's hands - coming together. In the other videos, sounds and movements may coincide, but in this one, there's a strong sense that the colliding lights are creating the sound.

And it seems that autistic children are very sensitive to that. Reverse and flip the animation, and the actions no longer match the sounds, which is why the children tended to ignore the inverted animation in favour of the upright one.

A fine theory, but one that needed support. To get that, Alim actually measured the extent to which sound and visuals marched in sync in the animations, by multiplying the speed at which the point-lights moved with how loud the associated sounds were. This measure, which  Alim called "audiovisual synchrony" or AVS had no sway over the non-autistic kids but it strongly affected the attention of the autistic ones. The higher it was for an animation, the longer they spent looking at it and, in fact, AVS accounted for 90% of the toddlers' preferences.

This fondness for matching sounds and sights could explain a recent discovery by Alim's group - that autistic children spend unusually long staring at a person's mouth rather than their eyes. And after all, of all the parts of the face, surely our mouths produce the greatest synchrony between movements and noise, between moving lips and spoken sounds.

The results also show that autistic children start down a very different route of development very early on in life. We know that autism is strongly influenced by genes, but Alim this that these inherited influences are exacerbated because from a very early age, autistic children experience the world in a very different way.

They develop a strong attraction to matches in light and sound, but at the expense of a preference for biological motion that normal babies have two days after birth. Without this preference, they aren't privy to the world of social information that children are typically drawn to.  Little wonder then that they face social problems when they get older.

Reference:  Klin, A., Lin, D., Gorrindo, P., Ramsay, G., & Jones, W. (2009). Two-year-olds with autism orient to non-social contingencies rather than biological motion Nature DOI: 10.1038/nature07868

More on autism:

Subscribe to the feed

More like this

That's fascinating. I wonder if that discovery might have practical implications in helping autistic children by synchronizing sound and movement in engaging them.

Could this be used as an early screening method?

By Tsu Dho Nimh (not verified) on 30 Mar 2009 #permalink

To even take what Lilian said a bit further, could this discovery lead to teaching materials aimed at our ever increasing autistic populations?

By LostInDaJungle (not verified) on 31 Mar 2009 #permalink

Interesting. Doesn't really surprise me, what with the general problems with sensory perception. With visual processing being poor to begin with, it's not surprising that such highly-simplified representations of humans would be less clear to autistics.

Very interesting experiment. I wonder what kind of results the research would show if the tests were administerd to chimps. Hope to find out more.

Very interesting and clear, thank you.
I wonder if even using simple gestures rhythmically in time with speech would assist students with autism in focussing. Will be interested to see where this research goes.

This is terrible science: technically, all the study proves is that autistic children aren't fooled by the point-light animations.
Not responding to point-light animations as real animal movement doesn't prove anything at all about how one might respond to real animal movement. It merely demonstrates how one responds to point-light animations. And how often do we encounter those in our everyday life?

To a non-autistic person, sure, maybe that's close enough to the same thing, and a few points of light placed at key joints can simulate a moving animal... but this is because their brain is taking a lot of shortcuts in processing the visual data. Non-autistic brains are prone to very efficiently leaping to conclusions based on limited evidence; this improves processing speed but reduces accuracy, often significantly. So their brains see animal movement when what is actually in front of them is animated dots. Compared to an autistic person, non-autistics are sensorially numb, process information slowly, and are terrible at spotting patterns. Pointing out that a particular visual illusion fools non-autistics but doesn't fool autistics doesn't prove that there's something wrong with the autistic group, it just proves that they're less prone to being taken in by illusions.

If you look at what the test is actually testing â the ability to see what's really there rather than what you expect to see â then the autistic children are out-performing the non-autistics. They're not fooled by the illusion. That's interesting in its own right, even if it's less flattering to the vanities and prejudices of non-autistics.