Autism Spectrum Disorder encompasses a large group of developmental disorders characterized by deficits in social interaction and communication, and the expression of restrictive and repetitive behaviours and/or interests. An incredible new study highlights the need for early diagnosis and offers a way to test if a child is on the spectrum by looking at the metabolites in their blood.
Currently, 1 in 66 Canadian children have an ASD diagnosis, and the condition is usually noticeable before 2 years of age. It is known that early diagnosis is key for successful treatment, so current research is extremely important. However, screening can be difficult as there isn’t a clear-cut test that doctors can perform, and we mostly rely on behavioural responses.
“Understanding trends and patterns in ASD diagnosis is essential to developing meaningful programs and services to support people living with ASD and their families.”
– Dr. Theresa Tam, Canada’s Chief Public Health Officer.
The goal for years, for ASD and many psychological disorders, has been to try and find a blood test or some medical, physiological way to see if a child does in fact have the condition. About a year ago, a study came out hypothesizing a physiological test for autism, and recently their follow-up paper was released confirming its incredible success.
The team from the Rensselaer Polytechnic Institute used big data techniques to search for patterns in metabolites relevant in two connected pathways with suspected links to Autism Spectrum Disorder – the methionine cycle and the transsulfuration pathway. Metabolites are tiny molecules that circulate throughout the human body in the bloodstream. They are products of metabolic processes or metabolism.
The lead researcher, Juergen Han, created a predictive algorithm using 24 metabolites. In the first study published last year, they were able to identify 96% of typically developing participants and 98% of the ASD participants. In this recent follow-up study, using only 22 of the 24 metabolites, they were able to replicate the results, correctly predicting autism 88% of the time.
Han states that the drop in accuracy is likely due to many factors, one being that 2 of the metabolites were left out of the second trial, as they had been strong indicators in the previous study. Regardless, the accuracy in both papers is very promising.
It is also mentioned that though the results are exciting, future research should focus on using larger cohorts, as these trials had around 150 participants each. Successfully addressing the limitations would help to solidify the legitimacy of these tools to accurately separate autism spectrum disorder from the typically developing individuals.