Read Original Story

Guardian trumpets a test to diagnose autism, but where’s the evidence?


1 Star


New test can detect autism in children, scientists say

Our Review Summary

A small European study found increased biomarkers in the blood of 38 children with autism spectrum disorder (ASD) when compared to 31 healthy children in the control group.

This story jumps the gun by implying the results suggest these biomarkers cannot only “detect autism in children” but also “could lead to earlier diagnosis.” This speculative language appears to come directly from the lead author of the study who’s also quoted as saying “we hope the tests will also reveal new causative factors.”

What would have helped this story considerably would be interviewing sources not affiliated with the study to clarify the limitations of these biomarkers, explain how ASD is currently diagnosed, and provide some much-needed context about the novelty, availability, and clinical relevance of the biomarkers studied.


Why This Matters

Autism spectrum disorder includes a wide range of problems with social communication and interaction, as well as behavioral issues that can be repetitive, compulsive, and/or disruptive in many social settings.

The condition cannot be predicted or cured, so any news coverage about a potential screening method needs to be extremely cautious and pay special attention to the limitations of the study at hand, as well as include critique from independent sources.

According to 2016 statistics published by the CDC, around 1.1 million American children under the age of 21 have been diagnosed with autism, with a prevalence of roughly 1 in 68 kids.


Does the story adequately discuss the costs of the intervention?

Not Satisfactory

The costs for the blood and urine tests mentioned in the story are not included. A cost estimate might be speculative at this point, but it seems reasonable to expect some discussion of what specialized tests like this might cost.

Does the story adequately quantify the benefits of the treatment/test/product/procedure?

Not Satisfactory

The story says the study found “a link between ASD and damage to proteins in blood plasma … the most reliable of the tests they developed was examining protein in blood plasma, which found children with ASD had higher levels of the oxidation marker ‘dityrosine’ (DT) and certain sugar-modified components called ‘advanced glycation end-products (AGEs).”

However, the story doesn’t include actual data, so the reader has no way of knowing how reliable these tests might be.

According to the study, a computer was used to analyze several combinations of various biomarkers, and the best diagnostic association was found in children with ASD who had higher levels of DT and AGEs.

The sensitivity of this model (ie. how many children with ASD were correctly identified) was 92 percent. The specificity of the model (ie. how many people without ASD were correctly identified) was 84 percent.

It means that 8 percent of the children with ASD were missed by the model, and 16 percent of the children that were diagnosed with ASD didn’t actually have the condition.

Does the story adequately explain/quantify the harms of the intervention?

Not Satisfactory

Although potential harms from blood or urine tests are quite minimal, the story suggests the research findings could lead to routine screening for autism.

Screening tests are never completely benign; in particular — in this case — if a ‘false positive’ result led parents to believe their child had ASD when they did not … or … if a ‘false negative’ result offered false assurance that there was no ASD present.

Does the story seem to grasp the quality of the evidence?

Not Satisfactory

This is quite a small study which relies upon biomarkers in the blood. It can only show an association between those biomarkers and the 38 children already diagnosed with ASD. It can not prove those biomarkers cause autism. Nor can it prove — as suggested in the story — that these biomarkers can actually predict a child developing autism, or reliably diagnose a child with the condition.

These limitations are not mentioned in the Guardian story.

This sentence also caught our eye: “The outcome was a diagnostic test better than any existing method.” This is incorrect on two counts. First, as just noted, the study results do not support these blood and/or urine tests as diagnostic. Second, there is no existing diagnostic ‘test’ to compare to. There is no single diagnostic test, just as there is no cure, for autism.

Does the story commit disease-mongering?


The story does not disease monger and does well to provide this context: “It is estimated that about one in every 100 people in the UK has ASD, with more boys diagnosed with the condition than girls.”

Does the story use independent sources and identify conflicts of interest?

Not Satisfactory

There are no independent sources. Including independent experts could have given the story much-needed context, including whether there are other studies looking for potential biomarkers for ASD.

Conflicts of interest are not mentioned, and we could find none.

Does the story compare the new approach with existing alternatives?

Not Satisfactory

It’s never mentioned that there is no known blood, urine, or imaging test for predicting or diagnosing autism.

Diagnosis, as stated in this Mayo Clinic summary, is based solely on behavioral observations and assessments.

Does the story establish the availability of the treatment/test/product/procedure?

Not Satisfactory

The story mentions two damaged proteins, in particular, which researchers found in higher concentrations of the blood of children with ASD, than in blood of children in the control group. We’re not told if the tests for these proteins are routine or highly specialized. Exactly what was found in the urine isn’t mentioned. Therefore, the availability of both the blood and urine tests is unclear.

Does the story establish the true novelty of the approach?

Not Satisfactory

We’re told the study identifies blood and urine tests “that can detect autism in children” and the tests are believed to be “the first of its [sic] kind.”

However, as mentioned above, the study does not prove these blood and urine tests can predict or diagnose autism. So suggesting novelty in this way is, at the very least, premature if not sensationalistic.

Does the story appear to rely solely or largely on a news release?

Not Satisfactory

The story lifts quotes and almost entire sentences from a news release with no apparent original reporting.

Total Score: 1 of 10 Satisfactory


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.