Babies, Babbles and ASD


Autism spectrum disorder (ASD)** is a developmental disability that affects an individual’s social interaction and communication, among other behaviors. Although language impairment is not a core feature of ASD, many individuals on the autism spectrum struggle with using language appropriately in social contexts, and parents of autistic children report atypical patterns of language development. Researchers know there is much that can be learned about ASD by taking a close look at trajectories of language development. Understanding the earliest linguistic signals of ASD, as well as understanding sex-based communication differences across the lifespan may lead to improved diagnostic methods and novel therapies to boost communication skills.

The language-based challenges experienced by individuals with ASD manifest in several different ways. Some individuals with ASD may never develop spoken language, and use varying vocal sounds or assistive technology to communicate. Others have above-average verbal abilities. Children with ASD often do not develop language skills at the same pace as their peers without ASD, and their progress is not always linear. Some individuals develop an extensive vocabulary centered around a particular subject of interest, but their choice of words may be unusual. During conversations, individuals with ASD may repeat a recently-heard word, phrase, or sentence (called echolalia).   

Verbal language makes up only a part of how people communicate. Nonverbal communication - eye contact, gestures, and body language - provide important context during social interaction, and facilitate or impede social flow. Research has shown that individuals with ASD often have difficulty using and understanding nonverbal communication signals as well.

The Quantitative Linguistics Lab (QL2) in the Center for Autism Research (CAR), led by Julia Parish-Morris, PhD, is dedicated to understanding more about language development across the autism spectrum and in related neurodevelopmental and behavioral conditions. Using data collected as part of the NIH-funded Infant Brain Imaging Study (IBIS), Parish-Morris and her team are analyzing vocalization samples from infants and very young children who are later diagnosed with ASD. The team aims to better understand different developmental trajectories, with the long-term goal of identifying vocal signatures of ASD that can aid in early detection of symptoms.  “Currently, we do not have a good grasp of how babies [who go on to be diagnosed with autism] vocalize compared with those who do not. Our research is focused on vocalizations during the first two years of life, because early coos and cries could eventually provide clinicians with important clues for very early diagnosis of ASD,” says Dr. Parish-Morris, a developmental psychologist. “We also want to understand what these vocalizations are communicating – are they directed towards people? Are they well-formed babbles? Are they just cries and what to do these cries sound like compared to the cries of babies who do not develop ASD?”

One QL2 study examined the fusses and cries of babies at low genetic risk for ASD, compared with those of babies at high genetic risk. Dr. Parish-Morris and CAR research assistant Samantha Plate found that between 6 and 24 months, all infants produced fewer cries over time.  However, infants who went on to be diagnosed with ASD at age 2 produced more and longer cries than other groups of infants. When distinguishing between cries and fusses, the infants without ASD produced acoustically distinct sounds for each, whereas the high risk group with ASD did not. “This study represents a first step toward harnessing the potential power of early vocalizations as behavioral biomarkers of ASD; although our sample is small and bears replication, we are eager to see where this line of research leads.” 

Infants with ASD also produce fewer “socially directed” vocalizations than infants without ASD. The QL2 team’s research, presented at the 2019 International Society for Autism Research conference, showed that toddlers with ASD produce a smaller proportion of socially directed vocalizations than toddlers without ASD, with different growth patterns over time. “These findings provide early evidence that when toddlers with ASD vocalize, their vocalizations are less likely to be directed toward others,” explained Dr. Parish-Morris. 

The recent findings also complement research presented last year by Samantha Plate, showing that toddlers with ASD use fewer gestures to communicate, but do not supplement decreased gestures with additional speech. Instead, they produce more non-speech distress vocalizations. “By studying this important developmental period [of language acquisition] and incorporating verbal and non-verbal communication, we hope to understand ASD emergence and identify bio-behavioral markers that indicate the need for early intervention,” explained Parish-Morris.  QL2 has received an NIH grant to team up with machine-learning experts at CAR to test whether combining this full range of information about early vocalizations, including verbal, nonverbal, and acoustic data, can predict later ASD diagnosis at 2 years.

In an exciting extension of this work, PhD candidate Lisa Yankowitz, MA, has teamed up with Dr. Parish-Morris to combine behavioral and acoustic data from early vocalizations with brain imaging data. She was recently awarded a prestigious grant from the Autism Science Foundation to support the project. “We hope this research will allow us to not only identify early vocalizations that distinguish ASD, but also understand how the developing brain supports the emergence of these behaviors, both in ASD and typical development,” explains Yankowitz. This research is part of a larger effort at CAR to build a “digital phenotype” of ASD, combining linguistic data with behavioral, genetic and brain imaging data, ultimately linking symptoms and their underlying brain and genetic mechanisms.

Editor's Note: This project utilizes data collected as part of the Infant Brain Imaging Study (IBIS) funded by the National Institutes of Health (R01 HD055741, PI: Piven)

** Please note, CAR uses both person-first and identity- first language. Recent research has demonstrated that “autistic” is the preferred term in the autism community, with many others preferring “on the autism spectrum” or “person with autism” (Kenny et al., 2016, "Which terms should be used to describe autism? Perspectives from the UK autism community" Autism).