Scientists from the University of Warwick, UK, and the University of Bologna, Italy, recently reported on their research to develop a biomarker test that may be used in the future to help diagnose autism in children. Results from initial studies and computer modeling are promising, but additional research on larger, more diverse populations are still needed before the test can be considered for clinical use.
Autism, called autism spectrum disorder (ASD), is a developmental disorder characterized by difficulties with social interaction, communication, and behavior. The Centers for Disease Control and Prevention estimates that in the U.S., ASD affects one in 59 children. Research has shown that ASD is the result of different combinations of environmental and genetic factors.
The signs and symptoms of ASD vary widely from person to person and can include problems with speech, repetitive or compulsive behavior, hyperactivity, and anxiety. Cognitive abilities such as learning, memory, and problem-solving can range from gifted to severely challenged.
Because of the wide range of signs, symptoms, and severity, and because the signs usually don't begin to appear until age 2 or 3, ASD can be difficult to diagnose early. Currently, for children suspected of having ASD, doctors perform a series of time-consuming behavioral tests that may or may not provide a definitive answer.
A test that measures something countable, like the lab test described here, could help diagnose ASD and make it possible for children to get earlier treatment and intervention. Early and intensive behavioral intervention for children with ASD can improve learning, communication, and social skills.
Previous studies have shown changes in a number of genes that affect metabolism of proteins and the way the kidneys handle amino acids, the building blocks of proteins. In the current study, researchers looked at proteins and amino acids in blood and urine samples from a small number of study participants. They included 69 children ages 5 to 12—38 children diagnosed with ASD that varied in severity and 31 without ASD.
The scientists noticed a link between autism and changes in proteins and amino acids found in the blood and amino acids found in urine. Importantly, they found higher levels of two types of biomarkers in the blood of 38 children diagnosed with ASD: dityrosine (an amino acid indicator for oxidation damage) and advanced glycation end-products (proteins modified by sugars). Tests to measure these substances appeared to be the most promising of those evaluated by the researchers.
Using the test results from the 69 children, the researchers used a computer program to build an algorithm that could accurately identify 90% of the children with ASD and 87% of the children without ASD. This means 13% were incorrectly identified as having ASD.
Because of the study's limitations, some experts say enthusiasm should be tempered and note that the test is far from being ready for clinical use. For example, the researchers' analysis is based on a small sample of children who were, on average, 7 or 8 years old. ASD symptoms typically present much earlier. This study provides no evidence to say whether children younger or older than this age group or those with different diets or with complex medical conditions that are common in children with ASD would have the same "metabolic pattern." If the metabolic patterns did not hold true in other children, this test would not work.
The study also compared children with ASD to a set of controls with normal cognitive development; therefore, it is unknown how the test would perform to distinguish ASD from other neurodevelopmental or mental health conditions. The next steps are to conduct larger studies involving more participants.