Microbiome learning path · Chapter 14
Gut bacteria and conditions of the brain
Parkinson’s disease, Alzheimer’s disease, multiple sclerosis and autism are four very different conditions. Each has been linked with the gut microbiome in headlines. This chapter takes them one at a time: what is established about each condition, what microbiome research has added, whether that came from mice or from people, and what a gastroenterologist can do for the gut symptoms that come with them. It offers no microbiome treatment for any of them, because none has been shown to work.
In short
- No microbiome test can diagnose or predict any of these four conditions, and no microbiome treatment is established for any of them.
- In Parkinson’s disease the gut is involved early. Constipation can come more than ten years before the tremor.
- For Alzheimer’s disease, the evidence on gut bacteria is from mice and small studies. What is known to matter is blood pressure, diabetes, hearing, activity and the other established risks.
- In multiple sclerosis, careful studies find modest differences in gut bacteria. Medicines explain some of them.
- Autistic children have more gut symptoms and eat a narrower diet. The diet appears to explain most of the difference in their gut bacteria.
- The gut symptoms themselves can be treated, and that is worth doing.
How should I read research on gut bacteria and the brain?
With four questions. They come from chapter 4, and they matter more here than anywhere, because families facing these conditions are offered a great deal.
Four questions. Ask them of any headline that links gut bacteria with a brain condition.
- Was it mice or people? Most of the striking results are in mice bred to develop a version of the condition.
- Who were the comparison group? Unrelated volunteers eat differently and live elsewhere. Housemates and twins are a fairer comparison.
- Could medicines, diet or constipation explain it? Each of these changes gut bacteria, and each changes with the condition.
- Has a trial changed the outcome? A difference in stool is not a treatment. Only a controlled trial shows that changing the bacteria helps.
What is known about the gut in Parkinson’s disease?
What is established first. Parkinson’s becomes commoner with age, and inherited factors play a part. A review of long-term studies found higher risk with exposure to pesticides and with head injury, and lower risk in people who are physically active. Smokers and coffee drinkers also develop it less often. That is an observation and no reason to smoke.
The gut is involved early. This is the best-supported link between the gut and any brain condition.
- Among 6,790 men followed for 24 years, those who had fewer than one bowel movement a day were 2.7 times as likely to develop Parkinson’s as those who had one a day.
- In nine studies pooled, people with constipation were about twice as likely to be diagnosed with Parkinson’s later. The link held when the constipation was recorded more than ten years earlier.
Years before. Constipation can come more than a decade before the movement symptoms. Constipation is common, and Parkinson’s is not.
It is not known whether constipation is an early part of the disease or a marker of something that leads to it. Constipation is very common, and most people who have it will never develop Parkinson’s.
Does the disease travel along the vagus nerve? Two national studies looked at people whose vagus nerve had been cut during stomach surgery. In Denmark, those whose nerve was fully cut had about half the risk of Parkinson’s after 20 years. In Sweden, the operation as a whole made no difference, with a hint of lower risk after a full cut. Both estimates were imprecise. The idea is plausible and not settled.
In mice. Mice bred to overproduce the protein that builds up in Parkinson’s developed movement problems only when they had gut bacteria. Bacteria from people with Parkinson’s made the mice worse than bacteria from healthy people. These are mice.
In people. A study of 72 people with Parkinson’s and 72 without found one family of gut bacteria much reduced. A later re-analysis of ten such studies found that the differences were consistent and small. Two cautions apply. Parkinson’s medicines have their own effect on gut bacteria. And constipation changes stool bacteria by itself.
Two further findings are worth knowing. Gut bacteria can break down the main Parkinson’s medicine before it is absorbed, shown in the laboratory. And in insurance records of 144,018 people with inflammatory bowel disease, Parkinson’s was 28% more common than in people without. Both are leads for research.
Have trials of faecal transplant helped? Three randomised trials gave three answers.
Three trials, three answers. Faecal transplant is not a treatment for Parkinson’s disease.
- In China, 56 people took transplant capsules or placebo. The overall symptom score improved more with the transplant.
- In Finland, 45 people were analysed after one dose into the colon. It made no meaningful difference, and gut side effects were more frequent with the transplant: 53% against 7%.
- In Belgium, 46 people received donor stool or their own. After 12 months the movement score had improved by 5.8 points with donor stool and by 2.7 with their own.
Faecal transplant is not a treatment for Parkinson’s. See faecal transplant, phages and what is coming.
What a gastroenterologist can help with. Constipation, above all. In a trial of 72 people with Parkinson’s and constipation, four weeks of a multi-strain probiotic gave about one more bowel movement a week than placebo. That is relief of constipation, not treatment of Parkinson’s. Fluids, fibre, movement, a review of medicines and laxatives where needed remain the base. Difficulty swallowing, a slow stomach and weight loss also deserve attention. See constipation and slow stomach emptying.
What is known about gut bacteria and Alzheimer’s disease?
What is established first. Age is the strongest risk, and inherited factors matter. A commission of experts lists fourteen risks that can be changed, and estimates that close to half of dementia might in theory be avoided if all were removed.
Fourteen risks that can be changed. Gut bacteria are not on the list.
A trial has tested this approach. In Finland, 1,260 people aged 60 to 77 who were at higher risk were given either two years of guided diet, exercise, memory training and control of blood pressure and other vascular risks, or general health advice. Scores on tests of thinking improved a little more in the guided group. This was not a microbiome trial.
What microbiome research adds.
- In people. A small study found that people with Alzheimer’s dementia had less varied gut bacteria than people of the same age and sex without it. This is a link, and dementia changes eating, activity and medicines.
- In mice. Mice bred to develop the amyloid deposits of Alzheimer’s developed far fewer when raised without any bacteria. Giving them gut bacteria from other such mice increased the deposits.
No trial has shown that changing gut bacteria prevents dementia or slows it.
What a gastroenterologist can help with. Constipation, poor appetite, weight loss and difficulty swallowing are common as dementia advances, and each can be eased.
What is known about gut bacteria and multiple sclerosis?
What is established first. A review lists infection with Epstein-Barr virus, smoking, low vitamin D from little sun, and obesity in adolescence, together with female sex and inherited immune genes. The virus stands out. Among more than 10 million young adults in military service, 955 developed MS, and the risk rose 32-fold after infection with this virus and not after other viruses.
What microbiome research adds. MS research shows how much the answer depends on the comparison group.
Four designs. The comparison group shapes the answer.
- Patients and unrelated volunteers. A study of 60 people with MS and 43 without found several bacteria raised or lowered. It also found that people on MS treatment differed from those not on treatment. A study of 71 untreated patients and 71 volunteers found no major shift in the community as a whole, and differences in particular bacteria.
- Patients and their housemates. The largest study paired 576 people with MS with someone from the same household. Some species were more common in MS and others less, and treatment changed the picture again.
- Identical twins. In 34 pairs of identical twins of whom only one had MS, the overall bacterial profiles did not differ in any major way.
- Twins’ stool in mice. When stool from those twins was given to mice bred to develop an MS-like disease, more of the mice given the affected twin’s bacteria became ill.
So there is a signal, it is modest, and medicines are part of it. No microbiome treatment for MS has been established.
What a gastroenterologist can help with. Bowel problems are very common in MS. Constipation and leakage of stool can occur in the same person. The causes include the nerve pathways themselves, several medicines taken together, and difficulty in reaching a toilet. A review advises starting with daily habits and a regular bowel routine, and moving on to biofeedback or irrigation of the bowel through the anus if those fail. Treating the bowel well improves quality of life and reduces urinary infections.
What is known about gut bacteria and autism?
Autism is a difference in development, not a disease of the gut.
What is established first. Autism is largely inherited. In a study of 2 million people in five countries, about 80% of the variation was explained by inherited factors. Vaccines are not linked with autism. Five cohort studies covering 1,256,407 children found no association.
Gut symptoms are real and common.
- Across 15 studies, autistic children were about four times as likely as other children to have gut symptoms: constipation, diarrhoea and abdominal pain.
- Across 17 studies, they were about five times as likely to have feeding problems, and they took in less calcium and protein.
What microbiome research adds.
- In mice. Mice given gut bacteria from autistic donors showed behaviour that researchers regard as autism-like in a mouse. Mice given bacteria from other donors did not.
- In children. The largest careful study, of 247 children in Australia, found almost no direct link between an autism diagnosis and gut bacteria. It found a chain instead: restricted interests led to a narrower diet, and the narrower diet to less varied gut bacteria and looser stools. Its authors cautioned against claims that gut bacteria drive autism.
Which way the arrow points. In the largest careful study, a narrow diet explained the difference in gut bacteria.
What trials have tried.
- A transfer of gut bacteria. In 18 autistic children, a course of antibiotics, bowel preparation, an acid-lowering medicine and then donor gut bacteria for seven to eight weeks was followed by an 80% fall in gut symptoms and better behaviour ratings. Everyone knew what was given and there was no comparison group, so the result cannot be separated from expectation. The researchers themselves called for a placebo-controlled trial.
- A probiotic. In a small placebo-controlled trial, a probiotic did not change the overall severity of autism or the repetitive behaviours. One measure of social functioning improved. The authors asked for larger trials.
- Gluten-free, casein-free diets. Six trials with 143 participants found no effect on the core features of autism, and a hint of more gut side effects. In a careful trial in 14 children already on the diet, giving gluten or casein in disguised snacks made no difference to behaviour.
A restrictive diet has a cost in a child who already eats few foods. Calcium and protein are the nutrients most at risk.
What a gastroenterologist can help with. The gut problem deserves assessment and treatment in its own right. Constipation, reflux and abdominal pain are treated as in any child. A child who cannot describe pain may show it as a change in behaviour or sleep. Coeliac disease should be tested for before gluten is removed, because the test is unreliable afterwards. A dietitian can widen a narrow diet gradually. A urine organic acids test does not diagnose autism.
What do the four conditions have in common?
- Gut symptoms come with them, and sometimes before them. Those symptoms are treatable.
- Stool differences exist, and they are small. Diet, medicines and bowel habit explain a good part of them.
- Mouse experiments are strong. They show what is possible in a mouse bred for the purpose.
- Trials in people are few. Where there are several, as with transplant in Parkinson’s, they disagree.
Which claims should I doubt?
- A stool test that claims to diagnose, predict or grade any of these conditions
- A programme that promises to undo dementia, Parkinson’s or MS through the gut
- Antibiotics or antifungal medicines for autism
- Faecal transplant for any of these conditions outside a clinical trial, in India or abroad
- Strict elimination diets for a child without a diagnosis that calls for one
- A long list of supplements chosen from a microbiome report
What is not known?
- Whether the changes in gut bacteria come before these conditions or follow them
- Whether Parkinson’s truly begins in the gut in some people
- Whether any change to gut bacteria alters the course of any of the four
- Which of the mouse findings will hold in people
From my practice
I see people with Parkinson’s disease and multiple sclerosis for constipation, swallowing and nutrition, and I work alongside their neurologist. Relieving a bowel that has been troublesome for years is worth doing for its own sake.
I do not offer faecal transplant, and I do not advise it for these conditions outside a clinical trial. I do not order stool microbiome tests to judge anyone’s risk of a brain condition. For autistic children with gut symptoms, I advise that a paediatric team assesses and treats the gut problem as it would in any child.
| What this page says | How it is known |
|---|---|
| Infrequent bowel movements went with 2.7 times the later risk of Parkinson’s | 6,790 men followed for 24 years |
| Constipation about doubled the later risk | Nine studies pooled |
| A fully cut vagus nerve went with lower risk | Two national register studies, which partly disagree |
| Gut bacteria were needed for movement problems | Mice |
| Stool differences in Parkinson’s are consistent and small | A re-analysis of ten studies |
| Faecal transplant in Parkinson’s gave three different results | Three randomised trials of 45 to 56 people |
| A probiotic eased constipation in Parkinson’s | A randomised trial in 72 people |
| Fourteen risks for dementia can be changed | An expert commission |
| Diet, exercise and vascular care improved thinking scores a little | A randomised trial in 1,260 people |
| Without bacteria, amyloid deposits were far fewer | Mice |
| Epstein-Barr virus raised the risk of MS 32-fold | A study of over 10 million young adults |
| Identical twins with and without MS had similar gut bacteria | 34 twin pairs |
| A narrow diet explained gut bacteria differences in autism | A study of 247 children |
| A transfer of gut bacteria was followed by improvement | 18 children, with no comparison group |
| Gluten-free, casein-free diets did not change autism | Six randomised trials, 143 participants |
When should I see a doctor?
See a neurologist for a new tremor, slowness, stiffness, loss of memory that affects daily life, loss of vision, numbness or weakness. See a paediatrician for concerns about a child’s development. See a gastroenterologist for constipation that does not settle, difficulty swallowing, weight loss, blood in the stool or leakage of stool, whatever other condition is present.
Questions patients ask me
I have constipation. Will I get Parkinson’s disease?
Almost certainly not. Constipation is very common and Parkinson’s is not. The studies show a higher risk across large groups. They do not predict what will happen to one person.
Can a stool test show my risk of dementia?
No. No stool test can diagnose or predict dementia, Parkinson’s disease, multiple sclerosis or autism.
Can a probiotic slow Parkinson’s or Alzheimer’s disease?
That has not been shown. In one trial a probiotic eased constipation in people with Parkinson’s, which is useful and is a different thing.
Should my autistic child go gluten-free and casein-free?
The trials found no benefit for autism, and the diet can narrow an already limited range of foods. If your child has gut symptoms, have them assessed first, including a test for coeliac disease before any gluten is removed.
Is faecal transplant available for these conditions?
Not as treatment. Trials in Parkinson’s disagree, and for the other three conditions there is no controlled evidence that it helps. I do not perform it at present.
Is there anything I can do for my brain through my gut?
Eat well, stay active and treat constipation. These are sound for many reasons. None has been proven to prevent a brain condition through gut bacteria.
Sources
This page is written for patients and is based on the guidelines and studies below. It is for education and does not replace a consultation.
- Review of risk factors for Parkinson’s disease. Ascherio A, Schwarzschild MA. Lancet Neurol. 2016. doi:10.1016/S1474-4422(16)30230-7
- Study of bowel movement frequency and later Parkinson’s disease in 6,790 men over 24 years. Abbott RD et al. Neurology. 2001. doi:10.1212/wnl.57.3.456
- Meta-analysis of constipation before a diagnosis of Parkinson’s disease. Adams-Carr KL et al. J Neurol Neurosurg Psychiatry. 2016. doi:10.1136/jnnp-2015-311680
- Danish register study of vagotomy and later Parkinson’s disease. Svensson E et al. Ann Neurol. 2015. doi:10.1002/ana.24448
- Swedish register study of vagotomy and later Parkinson’s disease. Liu B et al. Neurology. 2017. doi:10.1212/WNL.0000000000003961
- Mouse study showing gut bacteria are needed for movement problems in a model of Parkinson’s disease. Sampson TR et al. Cell. 2016. doi:10.1016/j.cell.2016.11.018
- Study of gut bacteria in 72 people with Parkinson’s disease and 72 without. Scheperjans F et al. Mov Disord. 2015. doi:10.1002/mds.26069
- Re-analysis of ten studies of gut bacteria in Parkinson’s disease. Romano S et al. NPJ Parkinsons Dis. 2021. doi:10.1038/s41531-021-00156-z
- Study separating the effects of Parkinson’s disease and of its medicines on gut bacteria. Hill-Burns EM et al. Mov Disord. 2017. doi:10.1002/mds.26942
- Laboratory study of gut bacteria that break down the main Parkinson’s medicine. Maini Rekdal V et al. Science. 2019. doi:10.1126/science.aau6323
- Cohort study of inflammatory bowel disease and later Parkinson’s disease. Peter I et al. JAMA Neurol. 2018. doi:10.1001/jamaneurol.2018.0605
- Randomised placebo-controlled trial of oral faecal transplant in 56 people with Parkinson’s disease. Cheng Y et al. Gut Microbes. 2023. doi:10.1080/19490976.2023.2284247
- Randomised placebo-controlled trial of faecal transplant by colonoscopy in Parkinson’s disease. Scheperjans F et al. JAMA Neurol. 2024. doi:10.1001/jamaneurol.2024.2305
- Randomised trial of donor against own stool in 46 people with early Parkinson’s disease. Bruggeman A et al. EClinicalMedicine. 2024. doi:10.1016/j.eclinm.2024.102563
- Randomised placebo-controlled trial of a probiotic for constipation in 72 people with Parkinson’s disease. Tan AH et al. Neurology. 2021. doi:10.1212/WNL.0000000000010998
- Report of an expert commission on dementia prevention, intervention and care. Livingston G et al. Lancet. 2024. doi:10.1016/S0140-6736(24)01296-0
- Randomised trial of diet, exercise, cognitive training and vascular care in 1,260 older people at risk. Ngandu T et al. Lancet. 2015. doi:10.1016/S0140-6736(15)60461-5
- Study of gut bacteria in people with and without Alzheimer’s dementia. Vogt NM et al. Sci Rep. 2017. doi:10.1038/s41598-017-13601-y
- Study of amyloid deposits in Alzheimer’s-model mice raised without bacteria. Harach T et al. Sci Rep. 2017. doi:10.1038/srep41802
- Review of genetic, lifestyle and environmental risk factors for multiple sclerosis. Olsson T et al. Nat Rev Neurol. 2017. doi:10.1038/nrneurol.2016.187
- Study of Epstein-Barr virus infection and multiple sclerosis in over 10 million young adults. Bjornevik K et al. Science. 2022. doi:10.1126/science.abj8222
- Study of gut bacteria in 60 people with multiple sclerosis and 43 without. Jangi S et al. Nat Commun. 2016. doi:10.1038/ncomms12015
- Study of gut bacteria from 71 untreated people with multiple sclerosis and 71 controls, with mouse experiments. Cekanaviciute E et al. Proc Natl Acad Sci U S A. 2017. doi:10.1073/pnas.1711235114
- Study of gut bacteria in 34 identical twin pairs of whom one had multiple sclerosis, with mouse experiments. Berer K et al. Proc Natl Acad Sci U S A. 2017. doi:10.1073/pnas.1711233114
- Study of gut bacteria in 576 people with multiple sclerosis and household controls. iMSMS Consortium. Cell. 2022. doi:10.1016/j.cell.2022.08.021
- Review of bowel problems in multiple sclerosis and their management. Preziosi G et al. Degener Neurol Neuromuscul Dis. 2018. doi:10.2147/DNND.S138835
- Study of inherited and environmental factors in autism in 2 million people in five countries. Bai D et al. JAMA Psychiatry. 2019. doi:10.1001/jamapsychiatry.2019.1411
- Meta-analysis of vaccines and autism in cohort and case-control studies. Taylor LE et al. Vaccine. 2014. doi:10.1016/j.vaccine.2014.04.085
- Meta-analysis of gut symptoms in autistic children. McElhanon BO et al. Pediatrics. 2014. doi:10.1542/peds.2013-3995
- Meta-analysis of feeding problems and nutrient intake in autistic children. Sharp WG et al. J Autism Dev Disord. 2013. doi:10.1007/s10803-013-1771-5
- Study of autism, diet and gut bacteria in 247 children. Yap CX et al. Cell. 2021. doi:10.1016/j.cell.2021.10.015
- Study giving gut bacteria from autistic donors to mice raised without bacteria. Sharon G et al. Cell. 2019. doi:10.1016/j.cell.2019.05.004
- Open-label study of a transfer of gut bacteria in 18 autistic children. Kang DW et al. Microbiome. 2017. doi:10.1186/s40168-016-0225-7
- Pilot randomised placebo-controlled trial of a probiotic in autistic children. Mazzone L et al. Cell Host Microbe. 2024. doi:10.1016/j.chom.2023.11.021
- Meta-analysis of six randomised trials of a gluten-free, casein-free diet in autistic children. Keller A et al. Nutrients. 2021. doi:10.3390/nu13020470
- Double-blind challenge trial of gluten and casein in 14 autistic children. Hyman SL et al. J Autism Dev Disord. 2016. doi:10.1007/s10803-015-2564-9