Scientists have found a link between autism and a set of proteins in the blood. This could be detected through a blood test, facilitating an earlier detection of the disorder.
Autistic spectrum disorder (ASD) is still a poorly understood condition. It affects information processing in the brain by altering how nerve cells and their synapses connect and organize, but the mechanism through which this happens is still unclear. Rather, autism is generally defined as a broad set of developmental disorders which cover a wide spectrum of behavioral problems. These problems can vary wildly in intensity and how they manifest themselves, potentially including speech disturbances, repetitive and/or compulsive behavior, hyperactivity, anxiety, and difficulty to adapt to new environments.
Since there is such a wide range of ASD symptoms, it can be extremely difficult to diagnose autism, especially at the early stages of development. Suspicious behavior of children can often be explained by natural causes, and symptoms can sometimes be quite subtle. This is why a direct, objective physical test would be extremely useful.
Researchers working in Bologna, Italy, locally recruited 38 children (29 boys and nine girls) who were diagnosed with ASD, as well as a control group of 31 healthy children (23 boys and eight girls) between the ages of five and 12. Blood and urine samples were taken from each of them.
The team noted the chemical differences in the samples and then inserted them into an Artificial Intelligence (AI) algorithm. The AI developed a mathematical equation that distinguishes between ASD and healthy controls. The outcome was a diagnostic test better than any method currently available.
Dr. Naila Rabbani at the University of Warwick and lead author of the study said that the discovery could lead to "earlier diagnosis and intervention."
The false positive rate was very low (positive predictive value was 88%), while the overall accuracy was 88%, she told ZME Science in an email. She was also kind enough to detail exactly how the test works.
"The test is based on an optimum combination of markers of damage to protein in blood plasma. The damage is low level and of two main types: oxidative damage - likely linked to low-level inflammation, and damage caused by the reactive carbonyl metabolite, glyoxal - likely linked to increased lipid peroxidation. Similar damage may be occurring in the brain in autism. We also found some disturbance in the handling of the amino acid arginine which supports previous evidence of a genetic association with autism."
She also added that their discovery can lead to a better understanding of the autistic specter, allowing us to understand what causes it and how it manifests throughout the body.
"We hope the tests will also reveal new causative factors. With further testing we may reveal specific plasma and urinary profiles or "fingerprints" of compounds with damaging modifications. This may help us improve the diagnosis of ASD and point the way to new causes of ASD."
So far, the study only analyzed children from age of 5 - 12 years old -- the applicability of the test in younger age groups remains to be assessed in future research. But since the test is objective and doesn't require any psychological evaluation, it could be scaled and implemented in clinics around the world
"The test could be widely implemented and provided by well-equipped clinical centers. Our test is an objective, blood-based clinical chemistry test that does not require psychiatric expertise," Dr. Rabbani told ZME Science.
"With further development, this test could help with the diagnosis, care and treatment of children with autism."
ASD is caused by a combination of genetic and environmental factors. Genetic factors have been found to account for 30-35% of cases of ASD and the remaining 65-70% can be explained by a combination environmental factors, multiple mutations, and rare genetic variants.
This study is reminiscent of a previous effort which found that autism can be detected even in babies by monitoring brain activity. The idea is somewhat similar -- you find the differences in the brains of ASD sufferers and feed them into an algorithm which then predicts autism incidence. The beauty of this approach is that you don't even need to know exactly what you're detecting, you just find enough differences, and that's enough to successfully predict incidence.