Metabolomic analysis of plasma biomarkers in children with autism spectrum disorders

Abstract

Abstract Autism spectrum disorder (ASD) presents a significant risk to human well‐being and has emerged as a worldwide public health concern. Twenty‐eight children with ASD and 33 healthy children (HC) were selected for the quantitative determination of their plasma metabolites using an ultraperformance liquid chromatography‐tandem mass spectrometry (UPLC‐MS/MS) platform. A total of 1997 metabolites were detected in the study cohort, from which 116 metabolites were found to be differentially expressed between the ASD and HC groups. Through analytical algorithms such as least absolute shrinkage selection operator (LASSO), support vector machine (SVM), and random forest (RF), three potential metabolic markers were identified as FAHFA (18:1(9Z)/9‐O‐18:0), DL‐2‐hydroxystearic acid, and 7(S),17(S)‐dihydroxy‐8(E),10(Z),13(Z),15(E),19(Z)‐docosapentaenoic acid. These metabolites demonstrated superior performance in distinguishing the ASD group from the HC group, as indicated by the area under curves (AUCs) of 0.935, 0.897, and 0.963 for the three candidate biomarkers, respectively. The samples were divided into training and validation sets according to 7:3. Diagnostic models were constructed using logistic regression (LR), SVM, and RF. The constructed three‐biomarker diagnostic model also exhibited strong discriminatory efficacy. These findings contribute to advancing our understanding of the underlying mechanisms involved in the occurrence of ASD and provide a valuable reference for clinical diagnosis.

Keywords

• autism spectrum disorder
• biomarkers
• machine learning
• metabolomic
• UPLC‐MS/MS