Plasma Metabolite Profiles of Children with Autism Spectrum Disorder

Background/Objectives: Autism spectrum disorder (ASD), a neurodevelopmental condition characterised by social and communication differences, is complex and aetiologically heterogeneous. Untargeted metabolomics is emerging as a tool in screening for biochemical abnormalities. This research was conducted using the Australian Autism Biobank resource and involved analysis of plasma metabolites to characterise metabolite differences between autistic children and controls. Methods: We sought to identify molecular signatures in the plasma of study subjects using mass-spectrometry methods. We included 955 untargeted plasma metabolites from autistic children (n = 491; 2-18 years; 78% male) and control subjects (n = 97; 2-17 years of age; 51% male). Statistical analyses were performed using questionnaire data for both groups, including standardised scores from the Autism Diagnostic Observation Schedule-Second Edition (ADOS-2), which measures the severity of autism-related behaviours. We also evaluated intellectual disability by examining the relationships between metabolites and clinical phenotypes. Results: After controlling the false discovery rate at 5%, we identified significant negative associations between the uncharacterised metabolites X-21383 and X-24970 and ASD status (p = 1.85 × 10^-6 and p = 1.92 × 10^-5 respectively). X-21383 was also found to be significantly reduced in autistic children with coexisting intellectual disability when compared with controls (p = 6.06 × 10^-6). No significant associations were identified between the metabolite data and ADOS-2 scores. However, greater levels of X-16938, N1-methyladenosine, and 2-oxoarginine were found to be suggestively associated with higher ADOS-2 scores (p = 2.95 × 10^-4-9.6 × 10^-5). Conclusion: This metabolomics study in the Australian Autism Biobank has identified several novel metabolites associated with core autism diagnostic behaviours.