The Microbiota–Gut–Brain Axis in Autism Spectrum Disorder: From Pathophysiological Mechanisms to Precision Therapeutics, A Comprehensive Review
Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition characterized by persistent social communication deficits and restricted, repetitive behaviors. Gastrointestinal symptoms are common and often correlate with symptom severity, implicating the microbiota–gut–brain axis as a potential mechanism linking gut dysbiosis with neurodevelopment through neural, immune, endocrine, and metabolic pathways. This review summarizes current evidence on alterations in the gut microbiota in ASD and critically examines whether these changes contribute to disease pathogenesis or represent secondary effects. It highlights recent advances in multi-kingdom microbiome profiling, metabolomics, mechanistic studies of neuroinflammation, and neurotransmitter signaling and considers major confounding factors, including diet, medication, and gastrointestinal comorbidities. Emerging studies emphasize microbial function over taxonomy. In the largest multi-kingdom analysis, 31 microbial and functional markers distinguished children with ASD from neurotypical controls with an area under the curve of 0.91, driven primarily by ubiquinol-7 and thiamine diphosphate biosynthesis pathways rather than individual taxa. Metabolomic and genetic studies suggest that microbial metabolites may mediate behavioral effects. Microbiota transfer therapy and fecal microbiota transplantation have demonstrated sustained improvements in gastrointestinal and behavioral outcomes, whereas probiotics and dietary interventions have produced inconsistent results. Although alterations in the gut microbiome are consistently observed in ASD, specific microbial signatures remain heterogeneous, and causality remains unproven. Functional microbial pathways appear more informative than taxonomic composition for biomarker discovery and therapeutic development. Future progress requires prospective birth cohorts, pre-diagnostic sampling, mechanistic validation, and adequately powered randomized trials before microbiome-based diagnostics and therapies can be translated into clinical practice.