"A hearing test is being hailed as a revolutionary technique to spot autism years earlier than current methods can," the Mail Online reports. The test is based on measuring how the inner ear reacts to sound.
But while the test shows promise, the headline is premature. The study the report is based on only looked at boys aged 6 to 17 years old and was not used to diagnose autism spectrum disorder.
In the study, 35 boys with autism and 42 boys the same age without autism had a range of hearing tests.
The first tests measured their ability to detect sounds at different levels and frequencies. All boys had the normal range of hearing.
But other tests used to measure the ear's ability to process and distinguish between similar sounds showed boys with autism had a 25% smaller processing response to sounds in the mid-range.
The researchers say this could make it hard for them to discriminate between sounds – for example, similar vowel sounds in speech.
The processing tests – using a measure called otoacoustic emissions – are regularly used to screen newborn babies.
The hope is they could also be used to look for difficulties in sound processing in line with those found in these boys with autism.
But we don't know whether babies with autism have the same sound processing difficulties, so more work needs to be done before it is confirmed (or not) that such a technique can be used to "diagnose autism" in babies.
The study was carried out by researchers from the University of Rochester and was funded by the US National Institutes of Health.
It was published in the peer-reviewed journal, Autism Research.
The Mail Online reporting oversimplified the research, making it sound as if the study simply measured children's ability to hear.
The news story said children with autism "struggled to detect sounds at a frequency of 1-2 kHz".
But, as the study makes clear, all the children could detect a normal range of sounds – it was the ability to process sounds and differentiate between different tones that differed among boys with autism.
The headline also gives the misleading impression that the test had actually been carried out in children with autism, which was not the case.
This case-control study measured the hearing and sound processing abilities of a group of boys with autism and a group of boys with normal development, matched for age.
Case control studies can show links between factors in one group compared with another – in this case, whether autism is linked to different sound processing ability – but it can't show whether one causes the other.
Researchers selected 35 boys aged 6 to 17 with high-functioning autism, and 42 boys matched for age without autism.
Each boy underwent a range of hearing tests – both standard audiometry and tests of cochlear function, which tests how well the ear processes sound.
The researchers looked for differences between the groups. They also looked at whether the results matched up with the boys' verbal or cognitive abilities and the symptoms of boys with autism.
Audiometry screening tests for the ability of the ear to detect sounds at different frequencies, at different decibel levels.
All the boys in the study had to reach a standard level of hearing to be sure any differences were not down to conductive or nervous system hearing loss.
The tests of interest measured the changes to sounds that happen in the ear, which are amplified by the hair cells of the inner ear (cochlear) and can be measured in the ear canal.
These sounds are called otoacoustic emissions (OAEs). An abnormal OAE response can suggest problems with sound processing.
Two types of test were used: one that used two tones close together, and another which used a series of clicks. Boys were tested in both ears.
They were excluded from the study if they had nerve damage or disorders, frequent or persistent ear infections, or other conditions that could affect hearing.
All the boys were tested to check their autism status and IQ before the study.
Researchers looked for differences in OAE results between the groups at different frequencies.
In the first test of OAEs, boys with autism had a smaller response to discriminating between two sounds in both ears, compared with the boys without autism, but only at the 1 kilohertz (kHz) frequency, which is in the mid-range of sound.
In the second test of OAEs, boys showed significantly reduced OAE responses to a series of clicks in the right ear, but not in the left ear, across a range of frequencies. When looking at the mid-range 1kHz frequency, both right and left ears showed reduced responses.
The researchers found no correlation between OAE results and either verbal or cognitive abilities.
However, the first OAE test results were related to severity of symptoms in the autism group, with boys with more reduced responses showing more severe symptom scores.
The researchers said that, "The observed decreases in OAE amplitudes at 1kHz mid-frequency range could cause reduced ability to discriminate between two sounds or impair auditory tuning."
This means children with this hearing problem could have impaired perception and understanding of speech, especially when there is background noise.
The researchers say more work is needed to test younger children and non-speaking children with autism to better understand the role of these hearing tests in autism.
But they suggest these tests might help diagnose children with autism at a much younger stage in the future, allowing them to begin treatment earlier.
Autism is a developmental disorder that affects behaviour and social communication. It's usually diagnosed in children aged around two to four years old.
We know there's a link between the condition and the ability to hear and process sounds – for example, some children with autism are very sensitive to sounds, while others don't respond to them at all.
However, hearing problems seem to be part of autism, rather than a cause of it. This study doesn't mean, for example, that deaf people have autism.
This study is interesting because it found a particular part of the ear, the cochlear, produces different effects in sound processing in children with autism, compared with those without the condition.
It may help us understand how autism starts – for example, whether it happens before birth, when the baby's ears and other organs are still forming.
The study has important limitations, however, meaning the hearing tests used in the study should not be seen as a diagnostic test for autism:
This technique is certainly worth following up, possibly with a cohort study to see if a positive response predicted by the test is actually confirmed in later life by a diagnosis of autism.
Until such research is carried out it is impossible to state with any certainty whether the test will be of practical use.