A Hearing Test May Help Detect Autism Risk in Newborns

A standard hearing test may one day be used to detect autism in newborns.

The findings, published in November in the journal Autism Research, shed new light on the connections autism spectrum disorder (ASD) may have with hearing and other sensory systems.

Previous studies already found that children with autism have slow brain responses to sounds. This new research observed that newborns who were later diagnosed with autism also had slower brain responses to sounds. That suggests that a hearing test already administered to millions of babies worldwide could potentially identify newborns with higher autism risk months or years before typical diagnosis. 

“Even though [autism spectrum disorder] may not manifest clinically until age 2, 3, or 4, at birth there's already a difference for some of these patients,” co-author Isaac Kohane, MD, PhD, a pediatric endocrinologist, tells Verywell “A lot of parents, understandably, see it happening at some point and say, ‘Oh my God, what happened at age 2?’ Well, this tells us for some of the patients, it's already happening at birth—and maybe before.”

Researchers Used a Common Newborn Hearing Test

Researchers from Harvard Medical School and the University of Miami examined the results of newborns’ auditory brainstem response (ABR), a test that determines whether they have any hearing loss or impairments. Pediatrix Medical Group, which screens 850,000 newborns a year for hearing impairment in the United States, performed the ABR tests.

The ABR is a simple test, part of the Universal Newborn Hearing Screening, that measures how well a baby’s inner ear and brain respond to sound. The newborn’s ears are covered with earphones that emit a series of soft clicks. Electrodes on their forehead and neck then measure the resulting brainwave activity. A computer records the responses and then compares them against predetermined response ranges.

These sounds are soft enough that the tests can be performed while the babies are sleeping, around 35 decibels normal hearing level. This is a significantly lower sound level than what is typically used during ASD screening. 

“Although the purpose of this is to check for hearing disorders, it in fact allows you to have this cheap neural physiologic experiment,” says Kohane, who is the inaugural chair of the Department of Biomedical Informatics at Harvard Medical School and associate professor of medicine at Brigham and Women’s Hospital in Boston. 

“The ABR is a poor man’s [electroencephalogram] EEG. It was surprising to me that it could be picked up on a simple, 3-lead EEG as opposed to a 12- or 24-lead EEG [conducted in] controlled conditions," he says.

Slower Brain Responses to Sounds

Researchers analyzed nearly 140,000 ABR tests from babies born in the state of Florida between 2009 and 2015. The hearing test was performed in the hospital within the first week of birth, on average between one and two days after, unless it was postponed because of a medical condition.

The data were then checked against the Florida Department of Education’s records of children with ASD. Of the 139,154 children in the data set, 321 were diagnosed with ASD by age 3 to age 5. 

Researchers found newborns who were later diagnosed with ASD had slower brain responses to sounds during their ABR tests, suggesting “a systematic difference between these kids who ended up having autism and those who don't,” Kohane says. 

He says their findings are significant because they could:

• Allow researchers to use the ABR as a prognostic test or a screen for autism spectrum disorder
• Provide researchers a biomarker that can be measured to determine the effectiveness of interventions or other therapies
• Give researchers a physiological difference that may offer insights into the cause of ASD, at least for some patients

Long-term Implications

Approximately one in 59 children have autism. Behavioral signs of autism are often present by 18 months, but diagnosis is not typically made before age 3-4. Earlier identification—and intervention—can have significant impacts on children. There are long waiting lists, sometimes up to a year, for children to be seen at specialty clinics, and this delay occurs during a critical time in childhood development. 

“We know that the brain has a much higher plasticity at younger ages,” Kohane says. “If you're able to, with some degree of certainty, identify patients who are at risk for being on the spectrum, [then early interventions are] much more likely to be effective. That sometimes makes the difference between being able to live independently or not.” 

The ABR can remove uncertainty by providing an objective likelihood of the child having ASD, alerting clinicians to follow up, and giving parents—who likely aren’t trained clinicians and who may be willing to make excuses when faced with signs of autism—a framework for how to proceed. 

Will the Test Be Used In the Future?

Researchers aren’t sure whether an ABR test performed at the lower sound intensity can accurately identify infants who will develop autism. They therefore need to replicate their findings to determine whether the test can provide clinically useful data. 

Even if future studies prove the ABR can accurately detect at-risk children, clinicians will still need to rule out other diagnoses before referring infants for a behavioral autism diagnosis. In addition to hearing impairments, it’s possible the ABR test can help detect speech and language impediments and sudden infant death syndrome (SIDS), among other developmental disabilities. It’s also possible the ABR test can be performed repeatedly to examine changes over time, rather than a once and done test on newborns. 

“Given the fact that we essentially have this free data being generated every day, [it] is an opportunity, without major expenditures, to study a way to both detect autism early and also to further characterize subgroups of autism," Kohane says. “It's a great opportunity to move forward.”

Researchers aren’t yet ready to recommend clinicians use the standard hearing test to diagnose autism. But Kohane is encouraged by the findings thus far and the potential for earlier autism detection and the research to come.

“I just want to be clear: Our findings are not ready for prime time,” he says. “In other words, if I had a child, I would not use the results of this test right now to inform me of the risk of autism. It is encouraging, and … we're thinking about how this test can evolve for the future.”