Photo by LisaW123, via Flickr
Photo by LisaW123, via Flickr

The Roots of Autism Are in the…Skin?

The skin is our gateway to the physical world. Below its surface are oodles of nerve fibers relaying different types of messages to the brain. At the ends of the fingertips, for example, fat and fast Aβ nerves help you fish for keys at the bottom of a messy purse, or feel the difference between cotton and polyester. Nearby those big nerves are thinner and slower C-fiber nociceptors, which transmit pain, and others that relay itchiness.

What I didn’t know until this week is that there is yet another type of nerve, found only under hairy skin, that carries information about our social interactions. These nerves, known as C-tactile (CT) afferents, respond to slow, gentle stroking — the soft touch you’d give to a baby’s forehead or a lover’s arm. And some researchers believe that these fibers are crucial for the development of the social brain.

“A hand on the shoulder, a pat on the back — these things anchor and cement social relationships in a meaningful way,” says Francis McGlone, a cognitive neuroscientist at Liverpool John Moores University in the U.K.

In today’s issue of Neuron, McGlone and his colleagues published a commentary reviewing what’s known about these nerves, including some provocative studies suggesting they play a role in autism.

“This C-tactile system is not there to sense the physical world, it’s there to feel the physical world,” McGlone says. “It’s coding something very important, particularly during early development.”

View Images
Nerve fibers in hairy skin (left) and smooth skin (right), from McGlone et al., Neuron 2014

CT nerves were first described in 1939. Swedish physiologist Yngve Zotterman found that in a cat’s leg, certain thin nerve fibers would fire electrical signals in response to slow stroking with the edge of a round wooden pin. “The complex response to stroking naturally raises the question whether the different groups of spikes are derived from groups of fibres with different sensory functions,” Zotterman wrote.

Subsequent studies confirmed that these nerves exist in cats (1957), as well as in monkeys (1977) and rats (1993). They weren’t reported in people until the late 1980s and ‘90s, thanks to a technique called microneurography pioneered by two Swedish scientists, Karl-Erik Hagbarth and Ake Vallbo. With this method (which Hagbarth and Vallbo first perfected on their own arms to prove its safety) a metal electrode tungsten microelectrode is poked through the skin of an awake person to record electrical signals from the nerves underneath.

Deciphering the code of these nerves is difficult and takes a lot of patience. “It’s like putting a microphone into a United Nations convention — there’s lots of different languages you’re going to be hearing,” McGlone says. “I think five people on this planet can record from C-tactile afferents.” These trained scientists can hear the language (that is, a certain pattern of electrical waveforms) of the CT afferents only when the skin is gently stroked.

McGlone’s lab has done a series of fascinating studies on CT afferents. His team uses a robotic stimulator to deliver gentle brush strokes in a steady, consistent way. Here’s a quick video of how it works:

The person in the video is rating how pleasant (or unpleasant) the touch feels on different parts of his body. In 2009, McGlone’s team published a study in Nature Neurosciencepublished a study in Nature Neuroscience in which this robotic brush stroked volunteers at different velocities. Turns out that the velocities that volunteers rated as most pleasant are the same ones that activate CT nerves. “They matched up perfectly,” McGlone says.

OK, so there are nerves that selectively respond to a soft touch. The real question is, why? What are they for?

“We’re still asking that question,” McGlone says. “What I hope I’ve done in this new paper is put a few more pieces of the jigsaw puzzle in place.”

Several puzzle pieces come from brain imaging studies. In 2012, for example, McGlone and collaborators in Montreal scanned volunteers’ brains while slowly stroking two skin areas: a hairy patch of the forearm, which holds CT fibers, and the hairless palm, which does not have CT fibers. Stroking CT fibers triggered activity in the posterior insular cortex and mid-anterior orbitofrontal cortex, which are both part of the brain’s limbic system, deep circuits that process emotion. Stroking the palm, in contrast, activated the somatosensory cortex, the outer layers of brain that process our physical sense of touch.

Stroking CT fibers also activated a brain region called the angular gyrus, which is involved in our internal representation of our body. (In studies of epileptic patients, stimulating this region leads to dramatic out-of-body experiences.) This result is intriguing, McGlone says, because it suggests that CT afferents are involved not only in our awareness of others, but in our physical sense of self.

The same brain regions activated by CT afferents — the insular cortex, orbitofrontal cortex, and angular gyrus — have also been implicated in autism and related disorders. Could autism be the result of an impaired touch system?

“I think it’s very believable,” says Kevin Pelphrey, a neuroscientist at Yale who is known for his brain-imaging studies of children with autism. In 2012, Pelphrey’s team scanned the brains of 19 healthy adults while they received either slow or fast strokes on their forearm. The slow touch activated brain regions involved in social behaviors, as shown before. But this brain activation was lowest in participants who scored high on tests of autistic traits.

Pelphrey is most interested in one of those brain regions, the superior temporal sulcus, or STS. This area is sensitive not only to social touch, but to socially meaningful sights and sounds. “We’re working on smells now as well,” Pelphrey says. His earlier work has shown that children with autism have abnormally low activity in the STS.

Pelphrey has also scanned the brains of children with autism while they felt slow or fast arm-stroking. “We wanted to know if the brain response to social versus non-social touch is present in autism or not, and to what degree it’s disrupted,” he says. “We found something,” he says, but wants to keep the results under wraps because the study is currently under review.

There are other reasons to suspect that CT afferents may be involved in autism, Pelphrey says. Many individuals with autism, such as autism advocate Temple Grandin, have sensory sensitivities. Pelphrey notes that some of the earliest descriptions of the disorder mention that babies with autism don’t react to being picked up in way that most babies do. “Touch is the first sensory system to develop,” Pelphrey says. “The brain response to C-tactile afferents should be present well before birth.” So if the system is disrupted in autism, it could become a very early biomarker of the disorder.

The question of why these afferents exist is still open. They could be vestigial, useless leftovers of our evolutionary past, the skin’s appendix. But McGlone doesn’t think so. He believes that affective touch is crucial for our brain development, and worries about what will happen as we transfer more and more of our social lives online.

“We live in a touch-deprived world,” he says. “You can see sort of an Armageddon scenario, where the affective touch system may well become vestigial. And what would the consequences be for the social brain?”


The text has been slightly modified: The electrode used in microneurography is made of tungsten, not metal.