Overview
- The authors—Kevin Mitchell, Dorothy Bishop and Darren Dahly—review observational studies, mouse experiments and human trials and conclude there is no credible causal evidence.
- Highly cited studies often relied on 7–43 participants per group and inconsistent assay methods, yielding contradictory findings that disappear when diet or sibling controls are considered.
- The researchers argue mouse models labeled with “autistic-like” behaviors are not relevant to human autism and say properly run probiotic or fecal transplant trials show no benefit.
- They contend observed microbiome differences are better explained by reverse causality, with autism-related dietary patterns shaping gut bacteria.
- The paper urges abandoning low-quality work or conducting large, rigorous studies; the UK’s National Autistic Society welcomed the conclusion even as a separate funder recently pledged $50 million for microbiome–autism research.