An integrative multi-omics approach identifies microbiome alterations linked to pathological and behavioral features in autism spectrum disorder

This study employs a multi-omics approach to investigate the gut-brain axis in 326 children with autism spectrum disorder (ASD) and 169 typically developing (TD) controls, aged 0-10 years. By analyzing neuroimaging, gut microbiome, and plasma metabolome data, we find that microbial features could accurately distinguish ASD from TD children. A key finding is that gut microbial abundance, particularly an increase in Clostridioides difficile, serves as the strongest predictor of both ASD symptom severity and brain structural variations. Crucially, these gut and brain differences are age dependent, diminishing as children with ASD aged and converging toward TD patterns. A mediation model suggests a potential pathway where specific microbes influence brain structure and behavior via metabolites. The findings establish the gut microbiota as a robust predictor of brain and behavioral phenotypes in pediatric ASD, underscoring the necessity for early, age-stratified therapeutic strategies via modulating the composition of the gut microbiome.