Thirty-eight trillion neighbours
Your gut holds roughly as many bacterial cells as your body has human ones. For most of medical history we treated them as enemies. We now know they digest, train, signal and defend, and that you can shape them.
What it is
Your microbiome is the community of bacteria, archaea, fungi and viruses that live on and in you. The densest population by far is in the large intestine, where hundreds of species share a warm, dark, oxygen-poor space and live off whatever your own digestion leaves behind, especially fibre. Together their genes outnumber yours many times over, which is part of why researchers describe the microbiome as an extra organ, one that you inherit partly, acquire mostly, and can change.
How it came to be
Animals evolved in a world that had already belonged to microbes for billions of years, so every animal has lived with them from the start. Studies of great apes suggest some gut bacteria have split into new lineages alongside their hosts, so human, chimpanzee and gorilla gut bacteria have been diverging for millions of years in step with us.
Each person's microbiome is rebuilt from scratch in early life. A baby's gut is seeded at birth, from the mother during a vaginal delivery, and more from skin and the hospital environment after a caesarean section. The differences between those two starts mostly fade over the first years. Then breast milk does something I find remarkable. It contains around two hundred kinds of complex sugars that the baby can't digest at all. They're there to feed a particular bacterium, Bifidobacterium infantis, which thrives on them and crowds out less helpful microbes. A mother's milk is, in part, food for her child's bacteria.
Many people in Japan carry a gut bacterium with a gene for digesting the sugars in red seaweed, the kind wrapped around sushi. The gene appears to have jumped from ocean bacteria living on the seaweed into gut bacteria, carried in on generations of meals. Most North Americans' gut bacteria don't have it.
What we used to think
- 1683Antonie van Leeuwenhoek scrapes plaque from his own teeth and describes "very many little living animalcules" under his microscope.
- 1860s–80sPasteur and Koch establish that germs cause disease. For a century after, medicine's view of microbes is a story of enemies.
- ~1900"Autointoxication" is fashionable, the idea that the colon poisons the body. A few surgeons remove healthy colons to treat fatigue and vague illness.
- 1907Élie Metchnikoff, a Nobel laureate in immunology, credits soured milk bacteria for the long lives of Bulgarian villagers. The probiotic industry traces its origin to his idea.
- 1970sA long-repeated estimate appears: bacteria outnumber human cells ten to one. It stays in textbooks for forty years.
- 1984Barry Marshall drinks a culture of Helicobacter pylori to prove it causes stomach ulcers, which doctors had blamed on stress and acid. He wins a Nobel Prize in 2005.
- 2007The Human Microbiome Project begins. DNA sequencing lets researchers count microbes that had never grown in a lab dish, which turns out to be most of them.
- 2016A careful recount revises ten-to-one to roughly one-to-one.
What we think now
The microbiome is now understood as a working partner with several jobs:
- Digesting what you can't. Bacteria ferment fibre into short-chain fatty acids. One of them, butyrate, is the main fuel for the cells lining your colon and helps keep that lining sealed.
- Making nutrients. They produce vitamin K and several B vitamins, and transform bile acids, which affects how you absorb fat.
- Training the immune system. Early-life exposure helps the immune system learn what to tolerate, and the community keeps calming or provoking inflammation throughout life.
- Defending territory. A full, diverse community leaves little room for invaders. Antibiotics that clear it out are how C. difficile infections usually start.
- Talking to the brain. Gut microbes influence the nervous system through the vagus nerve, immune signals and the chemicals they make. Most of the body's serotonin is made in the gut, partly under microbial influence, though that serotonin doesn't reach the brain directly.
- Processing your medicines. Gut bacteria can activate, deactivate or alter drugs. One species, Eggerthella lenta, inactivates the heart drug digoxin, which helps explain why the same dose works differently in different people.
Two cautions shape current thinking. First, there's no single "healthy" microbiome. Healthy people differ widely, and what seems to matter is diversity, resilience and what the community does, more than which names are on the list. Second, a lot of the evidence is association. Mouse experiments can be striking: germ-free mice given gut bacteria from the heavier of two human twins gain more fat than mice given bacteria from the leaner twin. Proving the same causes and effects in people is much harder.
The one treatment with undeniable results is the most literal one: replacing a broken community with a healthy donor's.
Faecal microbiota transplantation cures most cases of recurrent C. difficile infection where antibiotics keep failing, and regulators have now approved standardised versions, including one taken as capsules. For most other conditions it's still experimental.
When it goes well
A diverse, well-fed microbiome is associated with steadier blood sugar after meals, a stronger gut lining, lower background inflammation, better resistance to gut infections, and more regular bowels. People who eat more kinds of plants tend to have more diverse microbes, and greater diversity tends to travel with better metabolic health.
The Amish and the Hutterites in the United States share ancestry and a similar way of life. Amish children, raised on traditional farms among animals, have far lower rates of asthma than Hutterite children on industrialised farms. When researchers exposed mice to dust from Amish homes, the mice were protected from allergic airway disease. Early contact with a rich microbial world seems to teach the immune system restraint.
When it goes badly
A disrupted microbiome, sometimes called dysbiosis, shows up alongside a lot of conditions. In most of them, whether the microbes are cause, consequence or both is still open.
- Gut disease: inflammatory bowel disease (Crohn's and ulcerative colitis), irritable bowel syndrome, and recurrent C. difficile, where the causal link is clearest.
- Metabolism: obesity, insulin resistance and type 2 diabetes. Metformin, a common diabetes drug, appears to work partly by changing gut bacteria.
- Heart: some bacteria turn choline and carnitine (plentiful in red meat and eggs) into a compound the liver converts to TMAO, which is linked to cardiovascular risk.
- Immune and allergic conditions: asthma, eczema and food allergies, especially when early-life exposure is limited.
- Colorectal cancer: certain bacteria are found more often in and around tumours.
- Mood and brain: links with depression, anxiety and Parkinson's disease are an active research area, with intriguing but early human evidence.
What helps it
I've labelled each option below for how good the human evidence is.
Fibre, and many kinds of plants Strong
Fibre is what your gut bacteria eat, and most people in industrialised countries eat far less of it than their microbes would like. Variety seems to matter as much as amount. In one large citizen-science project, people who ate thirty or more different plants a week (vegetables, fruit, whole grains, legumes, nuts, seeds, herbs) had more diverse gut bacteria than those who ate ten or fewer.
Fermented foods Moderate
In a Stanford study, healthy adults who ate several servings of fermented foods a day (yogurt, kefir, kimchi, sauerkraut, kombucha) for ten weeks had more diverse microbiomes and lower levels of several inflammatory markers. A high-fibre group in the same study didn't gain diversity in that time, which suggests the two approaches work differently and may complement each other.
Using antibiotics only when needed Strong
Antibiotics save lives, and you should take them when they're prescribed. They also reduce diversity, and while most of the community recovers within weeks, some species can stay missing for months. Avoiding antibiotics you don't need, such as for viral colds, protects your microbiome.
Exercise Moderate
Active people tend to have more diverse gut bacteria, and in studies where sedentary people started exercising, their microbiomes shifted, then shifted back when they stopped. In one study of marathon runners, a bacterium that feeds on lactate became more common after races, and gave mice more endurance when transplanted into them.
Less alcohol and less ultra-processed food Moderate
Heavy drinking damages the gut lining and shifts the bacterial mix. Diets high in ultra-processed food tend to be low in fibre. Some additives raise questions of their own. Certain emulsifiers caused gut inflammation in mice and altered the microbiome in a small human trial, and some artificial sweeteners changed blood sugar responses in some people in a controlled study. Both are early findings.
Sleep, stress, and getting outside Early
Disrupted sleep and chronic stress alter gut bacteria in human and animal studies. Contact with nature, soil and animals (dog owners share skin microbes with their dogs) may broaden microbial exposure, especially for children.
Supplements and peptides
Products here range from proven drugs to grey-market compounds tested mainly in rats. I'm not giving doses in this guide.
Probiotics Moderate
Effects depend on the exact strain, not the species name on the label. The best evidence is for specific uses. Certain strains (such as Saccharomyces boulardii and Lactobacillus rhamnosus GG) reduce diarrhoea caused by antibiotics, and probiotics help protect premature babies from a serious gut disease. For healthy adults the benefits are unproven, and most probiotic bacteria pass through without settling. One study even found that taking probiotics after antibiotics slowed the return of a person's own microbiome compared with doing nothing.
Prebiotic fibres Moderate
Specific fibres such as inulin, galacto-oligosaccharides and resistant starch reliably feed bifidobacteria and other helpful groups. Psyllium has strong evidence for regularity and for lowering LDL cholesterol. Bloating is common at first. Whole foods deliver many fibres at once, and supplements don't.
Akkermansia Early
A bacterium that lives in the gut's mucus layer and is more common in leaner, metabolically healthier people. In one small human trial, a pasteurised (killed) form modestly improved insulin sensitivity. It's now sold as a supplement, ahead of larger trials.
Polyphenols and urolithin A Early
Plant compounds in berries, tea, cocoa, olive oil and pomegranate are mostly transformed by gut bacteria before the body uses them. Only some people's bacteria can turn pomegranate compounds into urolithin A, a molecule that supports the clean-up of worn-out mitochondria. Everyone else doesn't make it at all, which is one reason urolithin A is now sold directly.
Glutamine and collagen Early
Often marketed for "leaky gut". Glutamine is a fuel for gut lining cells and helps in some clinical settings such as critical illness, but the evidence in healthy people is weak. Evidence for collagen's gut effects is weaker still.
Teduglutide (a GLP-2 peptide) Strong
A real, approved peptide drug, a version of a natural gut hormone that grows the intestinal lining. It's used for short bowel syndrome, where much of the intestine has been removed. It shows what a properly tested gut peptide looks like.
BPC-157 Mostly animal
A synthetic peptide based on a fragment of a protein in stomach juice, widely sold for gut and tissue healing. Rat studies are numerous and positive, but controlled human trials are almost absent. It's banned in sport, and US regulators have flagged safety concerns about it in compounding pharmacies.
KPV and larazotide Mostly animal
KPV is a small fragment of a natural anti-inflammatory hormone that reduces gut inflammation in mice. Larazotide, designed to tighten the gut lining in coeliac disease, reached late-stage human trials, which were stopped when it didn't work well enough. Promising mechanisms often fail the step from animals to people.
Do-it-yourself transplants Strong evidence of harm
Faecal transplants work for recurrent C. difficile under medical supervision, with screened donors. Home versions skip the screening, and the screening is there for a reason. Transplanted material has transmitted dangerous drug-resistant bacteria even in a supervised trial, where one case was fatal. Leave this one to clinics.
The short version
Your gut bacteria are an old partnership, rebuilt in each of us from birth and shaped every day by what we eat. For a century we saw them mainly as threats. Now we see them as partners in digestion, immunity and metabolism, though in most conditions the science is still sorting cause from correlation. The best-supported moves are also the cheapest: many kinds of plants, some fermented food, fewer unnecessary antibiotics, regular exercise, and less alcohol. Supplements help in specific situations, and a lot of popular peptides haven't been tested properly in people yet.
This guide explains the science in general terms. It isn't medical advice and it gives no doses. Talk to your doctor before starting anything here, especially if you have a gut condition, a weakened immune system, or take regular medication.