Overview
- An opinion paper by Kevin Mitchell, Dorothy Bishop and colleagues argues that decades of autism–microbiome studies do not establish causality.
- The authors report that highly cited human studies often included only 7–43 participants per group, with apparent differences disappearing after accounting for diet or in sibling comparisons.
- Animal work is deemed unpersuasive because mouse ‘autistic-like’ behaviors lack relevance to human autism and key experiments had methodological and statistical flaws.
- Clinical trials of fecal microbiota transplants and probiotics have generally been small or poorly controlled, and well-designed studies have not shown consistent benefits.
- The paper urges either ending the research line or raising standards, as media interest and funding persist, including a reported $50m Wellcome Leap program, while the National Autistic Society welcomes the conclusion.