Skip navigation

Calls to Action

Infant Brain Activity Predicts Language Development in Autism

Brain scans of sleeping babies and toddlers at risk of autism may guide earlier, personalized interventions and new medical treatments
April 09, 2015

Researchers have found they can go far in predicting which children with autism will eventually develop language and which will not – based on differences in brain activity when the children are infants and toddlers.

The findings represent a potential breakthrough in tailoring future early intervention programs to improve communication and language even before a diagnosis of autism is possible at 18 to 24 months.

Further research on these early brain differences might also guide the development of new medicines that target the underlying brain biology, the authors conclude.

The study appears today in the journal Neuron.

“Why some toddlers with autism spectrum disorder get better and develop good language and others don't has been a mystery that is of the utmost importance to solve,” says senior author Eric Courchesne. Dr. Courchesne co-directs the University of California-San Diego Autism Center, where the study was conducted.

For example, some children with autism and language delays catch up with their peers in terms of speech even if their intonation and conversation skills are unusual. Others fail to develop language or even lose early speech abilities.

"We designed our study to gain insight into the brain bases for these different developmental outcomes," Dr. Courchesne says.

“The findings open the possibility of identifying the need for language-focused therapies as early as infancy,” comments Autism Speaks Head of Medical Research Paul Wang, who was not involved in the study. Research has shown that earlier intervention improves long-term outcomes for many children with autism. However, autism’s broad spectrum has made it difficult to select the best therapies for any one child.

"By identifying how subsets of individuals with autism differ in early brain and behavioral development, we can hopefully move steps closer towards more personalizing treatments for individuals with autism,” says lead author Michael Lombardo, of the University of Cambridge and the University of Cyprus.

Monitoring response to speech as babies sleep
The investigators enrolled 103 infants and toddlers, ages 12 to 24 months. The children were being seen at the UCSD autism clinic for developmental delays or other red flags of possible autism.

Using functional magnetic resonance imaging (fMRI), the researchers took a snapshot of each young child’s brain activity in response to speech as the child slept. The children then returned for behavioral and language assessments every six months.

By age four, 60 of the participants had a diagnosis of autism. Of these, 36 eventually developed strong language, though some had early speech delays. The other 24 remained nonverbal or minimally verbal.

Looking back at the children’s early fMRI scans, the researchers found that most of the babies and toddlers who had strong brain activation in response to speech went on to develop good language skills.

By contrast, most of the children who developed autism with poor or absent language had shown no significant activation in brain areas associated with speech in their earlier brain scans.

Combined with behavioral tests, these differences in early brain activity predicted later language development with 80 percent accuracy. Either measure alone (fMRI or behavioral assessments) predicted language development with 68 percent accuracy.

“MRI is one of the most commonly used research methods for studying the brain in autism,” comments Dr. Wang. “Now, by helping to identify which children are at greatest risk for poor language, this test may help guide parents, therapists and educators on how best to tailor a child’s intervention program.”


Subscribe to Autism Speaks Science Digest for autism research news, perspective and expert advice delivered biweekly to your inbox.