Many children diagnosed with autism spectrum disorder have difficulty communicating and establishing social bonds with others. A good deal of brain research seeking to explain why this might be the case has focused on the visual system and the processing of visual signals, as manifested, for example, in the interpretation of facial expressions or the ability to maintain eye contact.

Now, a research team that has turned its attention to sounds and the processing of vocal signals has reported its results. They sought to discover the neural basis for another source of difficulty in social communication: the fact that many children with autism have difficulty recognizing the emotional inflection in voices—something called emotional prosody. An example would be whether a voice is expressing happiness, sadness, or emotionally neutral content. Being able to do this accurately is a vital part of interacting and bonding with others.

The team’s data, drawn from comparative study of 43 children aged 7-12, 22 of whom were diagnosed with autism, revealed that portions of the cortex that process sounds are not the source of the problem in discerning spoken emotions. Rather, their evidence suggests, parts of the brain involved in interpreting these sounds appear to be responsible, particularly the function and connectivity of a social processing hub called the bilateral temporoparietal junction (TPJ). Potentially, it could be a target of future therapies to improve social communication in people with autism.

Daniel A. Abrams, Ph.D., a 2018 BBRF Young Investigator, was co-first author of the team’s paper, which appeared in Biological Psychiatry: Cognitive Neuroscience and Neuroimaging. Vinod Menon, Ph.D., a 1998 BBRF Young Investigator, was the team’s senior member. Simon Leipold, Ph.D. was also a co-first author.

“Children typically learn to map certain sounds in people’s voices to particular emotions,” Dr. Abrams explained. “If mom or dad is unhappy, a young child will know that before understanding all the words. But kids with autism often have trouble mapping vocal features onto emotion. Before this study we didn’t know why the brains of people with autism have a roadblock for identifying and recognizing these vocal cues.”

In the study, the 22 children diagnosed with autism and the 21 controls who were matched socio-demographically but were “neurotypical,” went through two kinds of tests. In one, the children performed a standardized test for recognizing vocal emotions. Results from this emotional recognition test indicated that children with autism were less able than neurotypical children to correctly identify emotions from vocal recordings.

A second test was conducted while the children were receiving functional MRI brain scans. During the fMRI task, children heard two recorded sentences spoken in happy, sad, and neutral emotional tones as well as recordings of common non-speech sounds, such as a running dishwasher. The goal was to compare the activity and connectivity of brain areas involved in processing happy and sad emotional vocal signals relative to neutral voices.

Results from the study revealed that areas of cortex that are associated with basic sound processing (auditory cortex) showed similar function in both groups of children. “What was atypical in kids with autism was the way this signal is getting to a crucial social brain region known as the temporoparietal junction, or TPJ,” said Dr. Abrams. In children with autism, during emotional voice processing, the TPJ showed atypical patterns of connectivity with portions of the auditory cortex that handle vocal signals. Moreover, those children with the greatest communication and social deficits had the lowest activity levels in the TPJ during emotional voice processing—the levels could in fact predict in any given child the severity of his or her deficit. Lower activity in the TPJ was especially associated with sad-inflected speech.

The evidence provides support for one of the two main theories of why people with autism have difficulty discerning verbal emotion. Results do not support the “sensory deficit model,” which proposes that the problem is rooted in the auditory cortex and involves faulty auditory processing of the vocal-emotional signals. Rather the new data supports the “social cognition model,” which posits that difficulties in vocal emotion processing are rooted in deficits to social brain regions, such as the TPJ, that are associated with interpreting the emotional content of vocal signals. Importantly, the TPJ and other parts of the “social brain” are thought to underlie, in part, what scientists call “theory of mind”: the ability of an individual to understand other people’s mental states and emotions.

The study’s results, if replicated, have the potential to inform future treatment approaches. It might be possible, for example, to develop a treatment that selectively boosts activity in the TPJ, an approach Dr. Menon says is “a very real possibility.”

Dr. Abrams also notes that the findings could support efforts to teach all children a lesson about neurodiversity, the notion that people with different kinds of brain wiring experience the world differently. “If we’re really going to improve communications with people with autism,” he said, “everybody needs to learn things. Parents could say to their neurotypical child, ‘That other child may have trouble recognizing how you’re feeling, but they really want to know and connect with you.’ That extra knowledge may help keep people with autism in the conversation.”