Gut-Derived Metabolic Imbalance in Autism Spectrum Disorder: Toward the Concept of a Metabolic Subtype

Autism spectrum disorder (ASD) is highly heterogeneous in symptom onset and severity, comorbidities, and treatment responsiveness, challenging a single “brain-centered” pathogenic model. Increasing evidence indicates that a subset of individuals with ASD exhibits prominent peripheral physiological alterations, including gastrointestinal (GI) dysfunction, gut microbial dysbiosis, immune imbalance, oxidative stress, and mito-chondrial/energy metabolic vulnerability. In this context, gut-derived metabo-lites—particularly short-chain fatty acids (SCFAs)—have emerged as plausible modula-tors of the neurodevelopmental milieu through the expanded gut–immune–metabolic–brain axis. This review synthesizes: (i) SCFA biogenesis and core physiological functions (GPCR-mediated signaling and epigenetic regulation), (ii) context- and developmental stage–dependent bidirectional effects shaped by dose, exposure duration, and tissue spec-ificity, (iii) the clinical heterogeneity of reported microbiome and SCFA alterations in ASD, and (iv) propionate as a frequently discussed candidate signal and the interpretive boundaries of preclinical evidence. Human studies show inconsistent directions and magnitudes of SCFA changes (increases, decreases, or no differences), driven by major sources of variability such as sample type (stool vs. blood, reflecting distinct physiological layers), GI symptom stratification, diet and medication/antibiotic exposure, and non-standardized analytical pipelines and reporting units. Accordingly, SCFAs should not be treated as universal ASD biomarkers; rather, they are better interpreted as con-text-dependent signals that may become salient under specific clinical-biological condi-tions. Building on this premise, we propose the conceptual framework of “metabolic ASD,” defined as a subtype in which peripheral metabolic–immune perturbations plausibly contribute to neurodevelopmental vulnerability. To avoid premature causal claims, we outline design requirements for future research, including developmentally informed lon-gitudinal cohorts, rigorous phenotypic stratification, standardized metabolomics, and multi-layer endpoints integrating barrier integrity, systemic inflammation, and metabolic stress. Ultimately, metabolic ASD should be positioned as a testable precision-medicine research frame rather than a universal etiological model.