Disrupted levels of molecular compounds in maternal blood and cord blood are linked to later diagnosis of autism spectrum disorder (ASD) according to a new study. The study was led by researchers at Columbia University Mailman School of Public Health and the Norwegian Institute of Public Health. Identification of these compounds sheds light on the biological processes that give rise to ASD and paves a way for earlier diagnosis and treatment for ASD.
The findings are published in Molecular Psychiatry in an article titled, “Metabolomic analysis of maternal mid-gestation plasma and cord blood in autism spectrum disorders.”
“The discovery of prenatal and neonatal molecular biomarkers has the potential to yield insights into ASD and facilitate early diagnosis,” wrote the researchers.
The researchers analyzed levels of 1,208 different chemical compounds in plasma samples collected from 408 mothers at mid-gestation (17–21 weeks) and in cord blood taken from 418 children at birth through the Norwegian Autism Birth Cohort (ABC). They used chromatography/mass spectrometry-based metabolomics assays to measure levels of chemical compounds. They used machine learning to assess the predictive value of the compounds as biomarkers for ASD.
The researchers found 12 chemical compounds in maternal mid-gestation (MMG) samples of ASD girls, three compounds in MMG samples of ASD boys, eight compounds in cord blood (CB) samples of ASD girls, and 12 compounds in CB samples of ASD boys to be linked to autism, including those that involve inflammation, disruption of membrane integrity, and impaired neurotransmission and neurotoxicity. Machine learning analyses suggested the potential utility of the compounds as biomarkers, especially those in cord blood, for early identification of children at risk for ASD.
“Our latest findings add to evidence that chemical compounds can be harnessed as an early biomarker for autism spectrum disorder with rapid developments in machine learning suggesting such a diagnostic test is feasible,” said first author Xiaoyu (Jason) Che, PhD, assistant professor of biostatistics in the Center for Infection and Immunity (CII) at Columbia Mailman School of Public Health.
