health

a Parkinson's disease feed — research, treatments, lived experience, in plain language

Translate

Reader summary

by xavier.grehant on 2026-05-20

Faecal microbiota transplantation Constipation Gut-brain axis Trial readouts

This paper is a systematic review — meaning the authors gathered and analysed all the available human studies on one question — published in npj Parkinson's Disease in May 2026. The question: can faecal microbiota transplantation (FMT) help people with Parkinson's? FMT means taking stool from a carefully screened healthy donor, processing it, and delivering it into a patient's gut (by colonoscopy, enema, or capsule). The idea is that Parkinson's patients have an altered gut microbiome (the trillions of bacteria living in the intestine), and that correcting it might ease symptoms. The reviewers pooled 11 studies — five randomised controlled trials (RCTs, the gold standard), three observational cohort studies, one case series, and two single case reports — covering 236 people in total.

Here is what they found. Safety looked acceptable: the most common side effects were mild, short-lived gut complaints (bloating, diarrhoea, nausea). No alarming serious adverse events stood out. Constipation improved consistently across studies — a meaningful win, since it is one of the most troublesome non-motor symptoms in Parkinson's. Motor and other non-motor outcomes were mixed: one RCT showed motor benefit at 8–12 weeks, another only after a year, and others found no meaningful change. Crucially, most improvements in the RCTs hovered around or just barely exceeded the minimal clinically important difference (MCID) — the threshold at which a change is noticeable to a patient in daily life. Five studies also found shifts in which bacterial species were present after FMT, most notably an increase in bacteria from the Firmicutes group, though what that means for Parkinson's remains unclear.

For someone living with Parkinson's, the honest message is this: FMT is not ready to be prescribed for Parkinson's. The evidence base is small and inconsistent, and the procedures vary so much across studies (different donors, different delivery methods, different preparation) that results cannot easily be compared. The consistent improvement in constipation is real and worth discussing with a gastroenterologist. The motor signals are intriguing but fragile. What is needed now — and what the authors call for explicitly — are larger, well-designed RCTs with standardised protocols. That work is underway (see the related trials linked on the journal page), so a clearer answer may come within the next few years.

What this article adds

Faecal microbiota transplantation (topic pending review)
This systematic review of 11 human studies (5 RCTs, 236 subjects) found FMT to be broadly safe in Parkinson's, with only mild transient gut side effects reported. Motor outcomes were inconsistent across RCTs — one showed benefit at 8–12 weeks, another only at one year, others none — and most changes hovered around the threshold of what patients would notice in daily life, leaving the procedure investigational rather than ready for clinical use.
Constipation
Across the 11 studies reviewed, constipation was the one outcome that improved consistently after FMT — standing out against otherwise heterogeneous results for motor and other non-motor symptoms. This makes constipation relief the most robust signal from the FMT evidence base to date, though standardised, larger trials are still needed before FMT can be recommended for this purpose.
Gut-brain axis
The review provides the most comprehensive human-trial summary to date of attempts to intervene on the gut-brain axis in Parkinson's via FMT, finding that gut microbiome composition does shift after transplant (most consistently an increase in Firmicutes), but that these shifts have not yet translated into reliable clinical benefit for motor or most non-motor symptoms in controlled trials.
Trial readouts
This 2026 review synthesises results from five FMT randomised controlled trials in Parkinson's, highlighting a critical methodological problem: donor selection, delivery route, pre-treatment, and follow-up duration varied so widely that cross-trial comparison is unreliable. The authors argue this heterogeneity — not a lack of biological signal — is the main obstacle to a definitive answer, and call for larger, protocol-standardised RCTs.

Contribute an improved analysis →