Fecal microbial marker panel for aiding diagnosis of autism spectrum disorders Original paper

What was studied?

The fecal microbial marker panel for autism diagnosis was investigated using shotgun metagenomics and machine-learning models to determine whether gut microbial signatures can reliably distinguish children with autism spectrum disorder (ASD) from typically developing peers. The study evaluated microbial taxa alongside metagenome-assembled genomes (MAGs), aiming to capture high-resolution microbial variation linked to ASD. By integrating taxonomic and genome-level signals, the research sought to build a clinically useful diagnostic model and assess how microbiome-based features relate to ASD symptom severity.

Who was studied?

A total of 598 Chinese children aged 3–12 years were included, comprising 264 clinically diagnosed ASD cases and 334 typically developing controls, divided into discovery and validation cohorts. Children were matched by age and sex where possible, and those with recent antibiotic or probiotic exposure, special diets, neurological disorders, or other major medical conditions were excluded. Stool samples were collected at home, preserved, and sequenced using Illumina platforms. ASD diagnosis followed DSM-IV or DSM-5 criteria, and symptom severity was measured using the Social Responsiveness Scale–2 (SRS-2). This well-characterized pediatric cohort provided a controlled environment for examining microbiome differences relevant to early-life neurodevelopment.

Most important findings

The study identified significant microbial dysbiosis in ASD, including reduced species richness, lower MAG-level diversity, and distinct community composition compared with controls (notably visible in data summaries and plots on pages 3–5). A machine-learning classifier integrating 5 bacterial taxa and 44 MAGs produced the strongest diagnostic performance, achieving an AUROC of 0.886 in the discovery cohort and 0.734 in an independent validation cohort. ASD-enriched microbes included Clostridium and Bacteroides nordii, while ASD-depleted species included Faecalibacterium prausnitzii, Bacteroides uniformis, Blautia wexlerae, Ruminococcus obeum, Streptococcus salivarius, and Bifidobacterium kashiwanohense. Viral MAGs from Caudoviricetes were also depleted. Many ASD-depleted microbes are known producers of short-chain fatty acids or immunomodulatory metabolites, whereas enriched species such as Clostridium and Ruminococcus gnavus have been implicated in toxin production or neurotransmitter-modulating pathways. Importantly, the microbial biomarkers correlated with SRS-2 social-communication and restricted-behavior subscales, indicating that microbiome signatures tracked with ASD severity.

Key implications

The findings demonstrate meaningful associations between gut microbial features and ASD diagnosis, highlighting the value of high-resolution MAG-based profiling in clinical biomarker development. The microbiome-derived model performed best in younger children (≤6 years), suggesting its utility for early screening when behavioral assessments are less reliable. Microbial signatures were robust across sex, gastrointestinal symptoms, and psychiatric comorbidities, reinforcing their potential generalizability within pediatric populations. The depletion of beneficial commensals and enrichment of taxa associated with neuroactive metabolites suggests mechanistic relevance to gut–brain signaling. While further validation across diverse populations is needed, this work establishes a foundation for microbiome-assisted ASD detection and supports deeper exploration of microbial contributions to symptom variability.

Citation

Wan Y, Wong OWH, Tun HM, Su Q, Xu Z, Tang W, Ma SL, Chan S, Chan FKL, Ng SC. Fecal microbial marker panel for aiding diagnosis of autism spectrum disorders. Gut Microbes. 2024;16(1):2418984. doi:10.1080/19490976.2024.2418984