Differential Metabolites in Chinese Autistic Children: A Multi-Center Study Based on Urinary 1H-NMR Metabolomics Analysis

Background: Autism spectrum disorder (ASD) is a group of early-onset neurodevelopmental disorders. However, there no valuable biomarker for the early diagnosis ASD. Our large-scale and multi-center study aims to identify metabolic variations between ASD healthy children investigate differential metabolites associated pathogenic mechanisms. Methods: One hundred seventeen autistic 119 were recruited from research centers 7 cities. Urine samples assayed by 1 H-NMR metabolomics analysis detect variations. Multivariate statistical analysis, including principal component (PCA), orthogonal projection latent structure discriminant (OPLS-DA), as well univariate used assess control groups. The further analyzed receiver operating characteristics (ROC) curve pathways analysis. Results: Compared with group, showed higher levels glycine, guanidinoacetic acid, creatine, hydroxyphenylacetylglycine, phenylacetylglycine, formate lower 3-aminoisobutanoic alanine, taurine, creatinine, hypoxanthine, N-methylnicotinamide. ROC relatively significant diagnostic values hypoxanthine [area under (AUC) = 0.657, 95% CI 0.588 0.726], creatinine (AUC 0.639, 0.569 0.709), creatine 0.623, 0.552 0.694), N-methylnicotinamide 0.595, 0.523 0.668), acid 0.574, 0.501 0.647) in group. Combining AUC reached 0.720 (95% 0.659 0.777). Significantly altered metabolite serine threonine metabolism, arginine proline taurine hypotaurine metabolism. Conclusions: Urinary amino significantly Amino might play important roles mechanisms