Research on neurodivergent kids: what the gut microbiome studies say, and what they do not

If you are the parent of an autistic child or a child with ADHD, you have probably seen headlines about the gut. Here is what the studies behind those headlines actually found. The short version: real associations, mostly from early life, a handful of small trials, and no food that treats a diagnosis.

The big Swedish birth cohort

A study that followed 16,440 Swedish children for more than 20 years found that 1,197 went on to a neurodevelopmental diagnosis. Early-life factors linked to a later diagnosis included infections, antibiotics and stress, and the infant gut microbiome and stool metabolome differed in children who later received a diagnosis of ADHD, autism, intellectual disability or speech disorder. The authors describe these as associations that point to immune and metabolic pathways, and as a foundation for prediction, not as causes. (Ahrens et al., Cell, 2024, PMID 38574728.)

The six-month microbiome study

In Detroit, 314 children had stool samples at 1 and 6 months and were assessed for ADHD at age 10; 59 had ADHD. The diversity and composition of the gut bacteria at 6 months were associated with later ADHD, and 51 bacterial groups at 6 months were linked to it, 14 of them in the order Lactobacillales. The authors asked for replication. (Cassidy-Bushrow et al., Pediatric Research, 2023, PMID 35440767.)

The Bifidobacterium finding

This one cuts against the easy story. In the Danish COPSAC2010 cohort, a higher level of Bifidobacterium in the gut at one week of age was associated with ADHD at age 10, after adjusting for genetics and early-life factors, and a Bifidobacterium-derived metabolite, indole-3-lactic acid, in the newborn blood spot mediated the link and replicated in two other cohorts. The authors' reading is that the timing of Bifidobacterium colonisation in the first week matters, not that Bifidobacterium is bad. (Widdowson et al., Molecular Psychiatry, 2026, PMID 41673465.)

The twin study

In a Swedish twin cohort enriched for ADHD (33 with ADHD, 79 without), the urine of males with ADHD carried more hippurate, a compound made jointly by gut microbes and the body that can cross into the brain; stool metabolites also differed. The twin design suggested many of these differences had a stronger genetic than environmental influence. (Swann et al., Neuropharmacology, 2023, PMID 37100381.)

The machine-learning signature

In Hong Kong, researchers sequenced the stool of 1,627 children aged 1 to 13 with or without autism. They found 14 archaea, 51 bacteria, 7 fungi, 18 viruses, 27 microbial genes and 12 metabolic pathways that differed, and a panel of 31 markers distinguished autistic from neurotypical children with an area under the curve of 0.91, driven mostly by lower biosynthesis of ubiquinol-7 and thiamine diphosphate. The authors present it as a possible non-invasive diagnostic aid, not a treatment. (Su et al., Nature Microbiology, 2024, PMID 38977906.)

The trials so far

• Boys aged 7 to 15 with autism, four weeks of Lactobacillus plantarum PS128 or placebo: better opposition and defiance scores, and better SNAP-IV totals in the 7 to 12 year olds. (Liu et al., Nutrients, 2019, PMID 30979038.)
• 84 children with ADHD, two months of a three-strain probiotic or placebo: parent-rated executive function scores improved against placebo (151.5 against 190.7, p < 0.001). (Parhiz et al., Neuropsychopharmacology Reports, 2026, PMID 41450035.)
• 53 UK children aged 8 to 13 with ADHD, six weeks of daily kefir or placebo: no overall effect on ADHD symptoms; fewer minutes awake at night on actigraphy (70 against 90, p = 0.04) but more self-reported sleep problems; more Bifidobacterium species in the kefir group. (Lawrence et al., BMC Psychiatry, 2025, PMID 41286799.)
• 30 autistic children (mean age 8), a 12-week open-label pilot of a synbiotic: improvements in anxiety, sensory hyperresponsiveness and abdominal pain, with no placebo group. (Wong et al., npj Biofilms and Microbiomes, 2026, PMID 41535300.)

What to take from it

The gut is part of the picture for neurodivergent kids, and the science is moving fast. But every study above is an association, a pilot, or a single small trial. None of it says a food or a probiotic treats autism or ADHD, and anyone who tells you otherwise is ahead of the evidence. Support plans, therapies and medication decisions belong with your child's clinicians.

This article reports published research for parents and adults who want to read it for themselves. It is not medical advice, it does not describe or recommend any product, and it never replaces a conversation with your GP or your child's clinicians.