Home > News & Opinion > News > 2013 > Controversial study uncovers hearing glitch in autism

Controversial study uncovers hearing glitch in autism

Popularity tracker
Virginia Hughes
29 July 2013

Listen up: A simple hearing test may help researchers better understand the brain circuits driving sensory sensitivities in children with autism.

An ear muscle is more sensitive to loud sounds in children with autism than in controls, according to a study published 3 July in Autism Research1. The researchers say this measure could serve as a simple clinical biomarker of the disorder, but others fiercely disagree.

The so-called stapedial reflex causes the tiny stapedius muscle inside the ear to contract in response to a loud sound. The reflex pulls the stapes bone away from the inner ear, which dampens the sound’s effect and protects the inner ear from too much vibration.

In children with autism, the reflex is a split second slower and triggers at sounds a few decibels lower than in typical controls, the study found.

Doctors routinely test the stapedial reflex in babies, but usually only assess its existence. The precise timing and sensitivity of the reflex could be a biomarker that points to autism years before abnormal behaviors are apparent, the researchers say.

“We hope that this is something we could use for screening in really young children,” even newborns, says lead investigator Randy Kulesza, Jr., associate professor of anatomy at the Lake Erie College of Osteopathic Medicine in Erie, Pennsylvania.

A few groups are finding other physiological differences in children with autism, such as a high heart rate and slow pupillary response to light.

These sorts of tests are especially appealing because of their simplicity: They are inexpensive, fast and can be done in a doctor's office. In contrast, most autism biomarkers depend on sophisticated and expensive technologies, such as brain scans.

Some scientists, though, are skeptical of the new test, saying the study relies on a small group of children and teenagers with autism who had problems with auditory attention or sound sensitivity.

“The idea that you could screen neonates based on this measure is just wild speculation,” says Gordon Ramsay, director of the Spoken Communication Laboratory at the Marcus Autism Center in Atlanta, who was not involved in the study.

Reflexive reactions:

The stapedial reflex is controlled by the brainstem, a conduit for nerves that transmit sensory information between the brain and the rest of the body.

In 1996, researchers in New York looking at postmortem brainstem tissue from a young woman with autism reported a “near-complete absence” of cells in the superior olive, a relay station for sound information2.

“That paper led me to the hypothesis that maybe these auditory pathways are disrupted in the autistic brain,” Kulesza says.

Several years ago, Kulesza began analyzing brainstem samples from the Autism Tissue Program, a repository of postmortem samples from individuals with autism and related disorders. As in the earlier study, Kulesza found that children with autism have dramatically fewer neurons in the superior olive than controls do3.

“Normally this structure has about 15,000 neurons. But in autism we consistently find only about 5,000, and sometimes even fewer,” Kulesza says.

Neurons in the superior olive are also involved in the stapedial reflex. In the new study, Kulesza searched 15 years of medical records from a nearby autism clinic, the Barber National Institute in Erie, for children who had been tested for this reflex.

The researchers identified 54 children with autism from the medical records and also gave the reflex test to 29 typically developing controls.

The test is quick and painless. The child wears a headset that simultaneously emits loud beeps and records pressure changes inside the ear.

Sounds ranging from 88 to 91 decibels — about as loud as a passing motorcycle — tend to trigger the reflex in controls. Slightly softer sounds, from 83 to 90 decibels — similar to a blender or blow-dryer — trigger the reflex in children with autism, the study found.

Other groups have used different tests to probe the auditory brainstem response in autism. In a 2000 study, for example, researchers used electroencephalography — a noninvasive technique that measures brain waves through the scalp — to measure the so-called 'brainstem auditory-evoked response,' the pattern of brain waves triggered after hearing clicks or tones. Some children with autism and their unaffected family members show an abnormally slow EEG response, the study found4.

Finding reliable brainstem biomarkers could have big implications for the early identification of children with autism, the researchers say. Most children are not diagnosed with autism until age 3 or 4 years, but studies have shown that the earlier they begin behavioral treatments, the better. “The stapedius reflex can be done the day they’re born,” Kulesza says.

But the finding comes with many caveats. The researchers did not calculate the sensitivity or specificity of the measure, so it's unclear how well it would predict whether a child has autism. They also did not look at the reflex test in children with other developmental disorders.

The new study is one of many on the sensory issues that frequently crop up in individuals with autism. Some children with the disorder are hypersensitive to sounds, for example, while others have trouble discriminating between different kinds of sounds.

If verified in other studies, the reflex test may give researchers clues about the brain circuits driving these sensory sensitivities, notes Tim Roberts, vice chair of radiology research at the Children's Hospital of Philadelphia, who was not involved in the new study.

“Perhaps most encouraging is their notion that auditory deficits are a cardinal feature of autism,” Roberts says. His team has used magneto-encephalography and found that children with autism show delays during a later stage of sound processing, in the cortex, or outer layers of the brain.

But other experts point out that these cortical circuits are likely to be much more relevant to autism than the lower-level mechanisms in the brainstem. Studies dating back decades have not found significant problems with hearing and vision in people with autism. At least two of these studies found the stapedial reflex to be normal in children with autism5, 6.

"In autism, it’s probably not the physiological mechanisms themselves that go wrong," Ramsay says. "It's the way those mechanisms are deployed in the world to actively look for information in the environment."

News and Opinion articles on SFARI.org are editorially independent of the Simons Foundation.

References:

1. Lukose R. et al. Autism Res. Epub ahead of print (2013) PubMed

2. Rodier P.M. et al. J. Comp. Neurol. 370, 247-261 (1996) PubMed

3. Kulesza R.J. Jr. et al. Brain Res. 1367, 360-371 (2011) PubMed

4. Maziade M. et al. Arch. Gen. Psychiatry 57, 1077-1083 (2000) PubMed

5. Gravel J.S. et al. Ear Hear. 27, 299-312 (2006) PubMed

6. Tharpe A.M. et al. Ear Hear. 27, 430-441 (2006) PubMed

Comments

Name: Eileen Nicole Simon
30 July 2013 - 11:57PM

Auditory system impairments should have been a primary focus of autism research decades ago. Complications at birth have often been reported as part of the background of children with autism, but are then usually dismissed as "non-specific." However, oxygen insufficiency is the great worry for any difficulty during birth.

William F. Windle investigated the effects of asphyxia at birth in monkeys, and found selective severe damage of brainstem auditory nuclei including the superior olives, and especially the inferior colliculi [1]. Blood flow and metabolism are higher in nuclei of the brainstem auditory pathway than anywhere else in the brain [2].

Kulesza et al. reported malformation of the superior olive in brains from people with autism [3]. Lukose et al. then found the same malformation in laboratory rats exposed to valproic acid (Depakote) during gestation [4]. Prenatal exposure to valproic acid is associated with development of autism. This most recent paper by Lukose et al. continues their investigations of auditory disturbance in autism.

Language development is the greatest concern, and auditory system defects are likely to prevent normal language development.

Reference
[1] Windle WF. Scientific American. 1969 Oct;221(4):76-84.
[2] Kety SS. Bull N Y Acad Med. 1962 Dec;38:799-812.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1804882/?tool=pubmed
[3] Kulesza RJ et al. Brain Res. 2011 Jan 7;1367:360-71.
[4] Lukose R et al. Brain Res. 2011 Jun 29;1398:102-12.

Name: Marinel
31 March 2014 - 8:46PM

I am a father of an autistic daughter. I am convinced that autism starts with a hearing deficiency in many cases. This deficiency can be genetic or induced by medication. My daughter received Gentamicin antibiotic soon after birth.
This antibiotic in particular can cause total deafness. My daughter is not deaf but definitely has sensitivity to noises.

Name: Sue Gerrard
31 July 2013 - 5:38PM

This finding is controversial (or contradictory) *only* if you're working on the assumption that 'the disorder' refers to autism as a discrete disorder with a homogeneous aetiology.

If you see autistic characteristics as an outcome of a number of possible causes, an abnormal stapedial acoustic reflex could be a biomarker for autism in *some* people.

Incidentally, it could also be a biomarker for auditory hyperacusis (e.g. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC539655/)
or even sub-clinical hypothyroidism (e.g. http://www.ncbi.nlm.nih.gov/pubmed/9709911), but sadly so-called co-morbidities that might provide important clues to the causes of autism always seem to be overshadowed by a pre-occupation with autistic characteristics per se.

Name: Eileen Nicole Simon
31 July 2013 - 11:40PM

Autism has many causes. Prenatal exposure to valproic acid is one, and (before dietary treatment was worked out) so was PKU in the past. All causes of autism must affect a “common final pathway” in the brain responsible for autistic traits.

Oxygen insufficiency at birth is brushed off with disdain, but needs to be faced up to. According to the Neonatal Resuscitation Program, one of every ten children needs some degree of assistance to begin breathing after birth. One reason for this may be the current protocol for clamping the umbilical cord immediately after birth. This dangerous practice was adopted in the mid-1980s. Obstetric textbooks up to that time taught that the cord should not be clamped until pulsations in it ceased, or at least until after the first breath.

Failure to establish breathing within 5 to 6 minutes after birth is ominous, and is likely to cause damage to auditory nuclei in the brainstem. This was demonstrated by William F. Windle, whose article in the October 1969 issue of the Scientific American merits renewed attention, as does the paper by Seymour Kety I cited above. It is free online. Use the link I provided above.

Louis Sokoloff extended Kety’s method through use of deoxy-glucose, and showed that glucose uptake is highest in the same nuclei of the auditory pathway that have the highest circulation in the brain. Lookup deoxyglucose in PubMed, and add terms like colliculus or auditory, or Sokoloff.

Young children are able to hear syllable and word boundaries before aging during the first decade of life begins to make this more difficult. This is why young children can learn a second language so easily. Trophic transmitters produced in the auditory system also guide postnatal maturation of target structures in the cerebral cortex (the language areas).

The research of Kulesza and Lukose is most important, and I hope the Simons Foundation will consider funding their research.

Name: Anonymous
1 August 2013 - 6:29AM

@Eileen:

Absolutely agree about the risk of oxygen insufficiency. The cord-clamping protocol would explain quite a lot.

However, why should there be a 'common final pathway' in autism? This would be the case if all people diagnosed with autism had the same problems with social interaction, communication and flexible behaviour, but they don't. Suggests there are a bunch of different common final pathways.

Not sure about the 'young children are able to hear syllable and word boundaries' part, either; Patricia Kuhn's model suggests a computational explanation that's more plausible.

Name: Eileen Nicole Simon
1 August 2013 - 11:39PM

Autism is a specific neurological disorder defined (diagnosed) according to specific traits: (1) lack of general awareness, including social obliviousness, (2) deviant language development, and (3) repetitive movements, or activities. These traits suggest injury to specific structures in the brain. Autism has many causes, but all must affect the same set of specific brain structures. This set of brain structures comprise the “final common pathway.”

Language disorder suggests disruption of maturation of the language areas (Broca’s and/or Wernicke’s areas) in the frontal and/or temporal lobes. Repetitive movements imply involvement of the subcortical basal ganglia. Obliviousness to the environment suggest defects in sensory systems.

Roger Brown (A First Language: The Early Stages. Cambridge, MA: Harvard University Press, 1973) determined that speech begins with use of stressed syllables, which a child then uses to put together “telegraphic” utterances. Autistic children by way of contrast begin speaking in phrase fragments, which suggests they may not hear syllabic stress.

I will have to lookup Patricia Kuhl’s computational model. Is this related to Noam Chomsky’s grammatical transformations?

Name: Vicki Stone
5 August 2013 - 6:40AM

Does anyone know of any research on superior canal dehiscence syndrome and autism?

Name: Sue Gerrard
8 August 2013 - 7:56PM

@Eileen:

"Autism is a specific neurological disorder defined (diagnosed) according to specific traits: (1) lack of general awareness, including social obliviousness, (2) deviant language development, and (3) repetitive movements, or activities. "

'Autism' is the three characteristics you've listed: whether or not they result from a *specific* neurological disorder is a moot point.

"These traits suggest injury to specific structures in the brain."

Only if you adopt a lesion-deficit model of brain dysfunction. In children, because the brain doesn't start out fully-formed, structural and functional abnormalities can be caused by anomalies way upstream of the structures where the abnormalities are observed.

"Autism has many causes, but all must affect the same set of specific brain structures. This set of brain structures comprise the “final common pathway.” "

That would be a reasonable conclusion if everybody diagnosed with autism had the *same* impairments. Someone with no apparent awareness at all of others and no speech, who spends all day rocking to and fro does not have the same autistic characteristics as someone who responds inappropriately to others, has difficulty with social communication and is obsessed by railway timetables. Both of these individuals clearly have issues with their frontal lobe function, but the frontal areas of the brain carry out a vast range of functions. As far as I'm aware there's no evidence showing that precisely the same structures are affected in everybody diagnosed with autism.

"Language disorder suggests disruption of maturation of the language areas (Broca’s and/or Wernicke’s areas) in the frontal and/or temporal lobes. Repetitive movements imply involvement of the subcortical basal ganglia. Obliviousness to the environment suggest defects in sensory systems."

Agreed, but see above re variations in impairments.

"Roger Brown (A First Language: The Early Stages. Cambridge, MA: Harvard University Press, 1973) determined that speech begins with use of stressed syllables, which a child then uses to put together “telegraphic” utterances. Autistic children by way of contrast begin speaking in phrase fragments, which suggests they may not hear syllabic stress. "

Quite likely, but things have moved on in the last 40 years.

"I will have to lookup Patricia Kuhl’s computational model. Is this related to Noam Chomsky’s grammatical transformations?"

About as far from Chomsky as you can get. Kuhl's model is a developmental, information-processing one. My apologies for typo and for not providing a reference. Try this one http://fonsg3.hum.uva.nl/paola/Kuhl_2004_.pdf

Name: Sue Gerrard
8 August 2013 - 8:02PM

@ Vicki:

We speculatively joined a few dots from different disciplines here http://link.springer.com/article/10.1007/s10803-009-0773-9, but couldn't find a direct connection in the literature, then or now. There's material in the otolaryngology literature that explains some auditory/eye movement/balance issues in autism, but few autism researchers appear to be working in that field.

Name: Eileen Nicole Simon
13 August 2013 - 11:41PM

@Sue Gerrard

Thanks for this discussion!

Neurological disorder should be the focus of research on autism. The Wernicke and Broca language areas of the cortex develop during early childhood and under influence of trophic neurotransmitters in the brainstem auditory pathway. If these brainstem centers have been damaged by asphyxia during birth (or by toxic substances in the neonatal period) maturation of the language areas will not proceed normally.

Children are diagnosed with autism based on distinctive impairments in language, awareness, and repetitive movements. Clearly these impairments are more severe in some children than in others, and depend upon the severity of damage to structures within the brain. Connectivity in the frontal lobes will not develop normally if subcortical structures are damaged (as in the language areas of the cortex)

I believe Roger Brown’s work was seminal. Thank you for the link to Kuhl’s paper. She devotes a good deal of discussion to young children’s ability to distinguish “acoustic changes at the phonetic boundaries.” Injury of the auditory system is a severe handicap for learning to speak.

I think I best presented my point of view at a meeting of the Interagency Autism Coordinating Committee (IACC) 5 years ago
http://iacc.hhs.gov/events/2008/slides_eileen_simon_112108.pdf

The IACC meetings are so frustrating because you are allowed 3 to 5 minutes to say something, but they never discuss the comments presented. Much more back and forth conversations are needed, as we seem to have going here.

Name: A G Gordon
12 January 2014 - 4:55PM

"(1) lack of general awareness, including social obliviousness, (2) deviant language development, and (3) repetitive movements, or activities. These traits suggest injury to specific structures in the brain."

They are also consitent with abnormal inner ear development. A recent paper in Science found abnormal inner ear genes in ADHD.

I reported stapedial reflex abnormalities in autistic children in 1976. These were the same anomalies as occur in audiosensitive non-autistic persons.

Name: A G Gordon
12 January 2014 - 5:32PM

"Does anyone know of any research on superior canal dehiscence syndrome and autism?"

Yes, see here for example:
Eur Arch Otorhinolaryngol. 2013 Aug;270(8):2353-8. doi: 10.1007/s00405-013-2482-4. Epub 2013 Apr 12.
Auditory profile and high resolution CT scan in autism spectrum disorders children with auditory hypersensitivity.
Thabet EM, Zaghloul HS

Incidentally, I proposed that perilymph fistulas were the cause of autism back in 1976!

Add a Comment

You can add a comment by filling out the form below. Plain text formatting.

Question: What is 10 + 4 ?
Your answer: