The strangest fact about being human is that you’re not alone in your own body. By cell count, only about half the cells inside you are human. The rest, roughly 38 trillion bacterial, fungal, archaeal, and viral cells, make up the microbiome, the ecosystem of organisms living on and inside you that has co-evolved with humans for at least the entire history of our species and probably much longer. These tenants digest food you couldn’t otherwise process, produce vitamins you can’t synthesize, train your immune system, manufacture neurotransmitters, regulate inflammation, and influence your mood, behaviour, and disease susceptibility through pathways the field is still mapping. The bacteria in your gut have more genes than you do. Roughly 100 microbial genes for every human gene. In a real sense, you’re a walking ecosystem with a single conscious tenant who thinks the whole show is about them.
What makes the microbiome interesting from a health perspective is that it’s both highly individual and highly modifiable. Your specific microbial community was shaped by how you were born (vaginal versus C-section), what you were fed as an infant (breastmilk versus formula), what you’ve eaten since, what antibiotics you’ve been exposed to, where you’ve lived, who you’ve lived with, what you’ve touched, and the genetic background that shaped which microbes could establish in the first place. It changes meaningfully on the timescale of days when you change what you eat, on the timescale of weeks when you change broader patterns, and on the timescale of years through accumulated lifestyle choices. This makes it one of the most leveraged systems in human biology. Small changes in inputs can produce large changes in microbial composition, which can produce systemic effects throughout the body.
What makes the microbiome difficult to traverse is that the popular discourse has parted ways with the underlying research. Probiotic supplements promise specific health outcomes the evidence doesn’t support. Practitioners recommend interventions (peptide injections, megadose glycine protocols, and daily activated charcoal) that have no real research base. The connection between gut and brain exists but is often overstated; the connection between leaky gut and various chronic diseases is contested but often presented as settled. This page attempts to walk that ground carefully. We’ll address what the primary research supports, flag what pop culture oversells, and name the practitioners whose recommendations sit on unstable ground.
The mechanisms behind macronutrient digestion are in Macronutrient & Hydration Basics. The micronutrient story (including B12 and K2 production by gut bacteria) lives in Micronutrient Basics. The fasting-microbiome interactions get covered in Fasting in Part II. Here we focus on the microbiome itself: what it is, what it does, what shapes it, and what to do (and not do) to support it.
The human microbiome includes microbes living in different ecological niches across the body, with the gut microbiome representing the largest and most studied community. Less discussed but also significant are the skin microbiome, the oral microbiome, the respiratory tract microbiome, the urogenital microbiome, and the breast tissue microbiome. Each ecosystem has its own composition, function, and modifying influences.
This page focuses primarily on the gut microbiome because that’s where the most actionable knowledge currently lives, but the broader picture is more important. Disrupting the skin microbiome with antibacterial soaps affects skin barrier function. Disrupting the oral microbiome with antibacterial mouthwashes affects nitric oxide metabolism (covered in Breathwork Basics where this connects to nasal breathing physiology). The microbiome is a collection of ecosystems, each with its own role.
The gut microbiome is dominated by two phyla (Firmicutes and Bacteroidetes), which together typically account for over 90% of bacterial content. The relative balance between these phyla varies dramatically between individuals and has been studied extensively in relation to metabolic outcomes. Ruth Ley and Jeff Gordon at Washington University in St. Louis published landmark work showing that lean and obese individuals have measurably different Firmicutes/Bacteroidetes ratios, and that microbial communities transferred between mice could induce weight changes in recipients. The popular extrapolation that “fix your microbiome ratio to fix your weight” goes a bit far, but the directional finding that microbial composition influences metabolic outcomes is well-established.
The other major bacterial phyla in the gut include Proteobacteria, Actinobacteria, Verrucomicrobia, and Fusobacteria. Specific genera within these (Bacteroides, Clostridium, Bifidobacterium, Lactobacillus, Prevotella, Akkermansia, Faecalibacterium) have been studied for their specific health effects, with varying levels of evidence quality.
Beyond bacteria, the gut also hosts archaea (particularly methanogens), fungi (the mycobiome – including various Candida species, which become problematic only when their populations expand inappropriately), viruses (the virome – particularly bacteriophages that infect bacteria, regulating bacterial populations), and a small number of protozoa.
The modern microbiome era began with the Human Microbiome Project (HMP), launched by the NIH in 2007, which used next-generation sequencing to map the bacterial communities at multiple body sites across hundreds of people. The HMP established the basic taxonomy of human microbial communities and showed how dramatically composition varies between individuals at the species level even when functional capacity remains similar.
Jeff Gordon at Washington University in St. Louis has been the foundational figure in modern microbiome research, with his lab producing decades of work on microbial community assembly, metabolism, and disease relationships. Rob Knight at UC San Diego founded the American Gut Project (now part of The Microsetta Initiative), the largest crowdsourced microbiome study in history. Justin and Erica Sonnenburg at Stanford have produced the most accessible body of work for general audiences alongside their primary research, including their 2021 Cell paper with Christopher Gardner showing that fermented food consumption increased microbiome diversity and reduced inflammatory markers more than fibre alone. Patrice Cani at UCLouvain pioneered the metabolic endotoxemia research that anchors much of the gut-systemic-inflammation framework. Sarkis Mazmanian at Caltech has done foundational work on microbiome-immune system interactions, including the role of specific bacterial molecules (polysaccharide A from Bacteroides fragilis) in immune development. Tim Spector at King’s College London has translated microbiome research into the consumer-facing ZOE personalized nutrition project.
These researchers are the actual primary source for most of what’s written about the microbiome in popular contexts. When practitioners cite “research shows” without naming specific studies or researchers, the underlying claims often trace back to one of these labs. Usually with simplifications or extrapolations that the original researchers wouldn’t endorse.
The list of microbiome functions has expanded substantially over the last two decades as research methods have improved.
Human digestive enzymes can’t break down most fibrous plant material, such as cellulose, hemicellulose, resistant starch, inulin, pectin, and various oligosaccharides. Gut bacteria can. They ferment these substrates and produce short-chain fatty acids (SCFAs) (primarily butyrate, acetate, and propionate) that get absorbed back into the body and serve multiple functions:
The Sonnenburgs’ work on microbiota-accessible carbohydrates (MACs), the technical term for the fibres your bacteria can ferment, has shown that MAC-deprived diets produce measurable reductions in microbial diversity over generations, with potentially permanent effects. This is the actual research base behind “feed your microbiome with fibre” advice.
Several vitamins are produced by gut bacteria, contributing meaningfully (though not entirely) to host nutrition:
Antibiotics that disrupt gut bacterial populations can produce subclinical deficiencies in these microbial-derived nutrients, particularly with long-term or repeated exposure.
The gut-associated lymphoid tissue (GALT), the immune cells embedded in and around the intestinal wall, is the largest concentration of immune cells in the body. Estimates that “60% of the immune system” lives in the gut circulate widely; the figure traces partly to comparisons of lymphocyte density and partly to functional importance, but the specific percentage is contested. The defensible claim is that the gut hosts the body’s largest population of immune cells and that gut bacterial populations directly shape immune function from infancy onward.
The mechanism: gut bacteria continuously interact with immune cells through the intestinal lining, producing molecular signals that train the immune system to distinguish between threat and non-threat, regulating inflammatory responses, and maintaining the balance between defensive activity and tolerance. Sarkis Mazmanian’s work on polysaccharide A (PSA) from Bacteroides fragilis showed that specific bacterial molecules can direct immune cell development toward regulatory rather than inflammatory phenotypes. Disruption of normal microbial communities through C-section delivery, antibiotic exposure, low-fibre diets, or other factors has been associated with increased rates of autoimmune disease, allergies, and inflammatory conditions, though the causal pathways are still being mapped.
The connection between gut microbiome and brain function is one of the most-discussed and most-overstated areas of current research.
What’s well-established:
What’s less established:
The 2020 work from John Cryan and Ted Dinan at University College Cork, particularly their book The Psychobiotic Revolution, represents the most accessible serious treatment of this territory, written by primary researchers in the field. Their framing is appropriately measured, the mechanisms are partly mapped, and the clinical applications are largely still emerging.
A 2018 study by researchers at the University of Alabama identified bacterial cells in human brain tissue, suggesting the brain may not be as sterile as previously thought. The implications of this finding remain unclear; the broader research program on bacterial signaling reaching the brain continues to develop.
The microbiome also influences:
Each of these is a substantial area of research in its own right; the summary above just establishes the breadth of microbiome influence on human physiology.
The microbiome’s composition is shaped by inputs accumulated across the lifespan, with the early years being particularly consequential.
Babies born vaginally acquire bacteria from the mother’s vaginal canal during birth, establishing initial microbial communities dominated by Lactobacillus and other organisms that match the maternal vaginal microbiome. Babies born by cesarean section instead acquire skin-associated bacteria from the mother and the hospital environment, with measurably different initial microbial profiles. The differences may persist for years and have been associated with elevated risk of allergic conditions, asthma, and metabolic disease in C-section-born children, though the magnitude and reversibility of these effects remain active areas of research.
Breastfeeding further shapes the developing microbiome. Human milk contains over 200 different oligosaccharides that humans can’t digest but that selectively feed beneficial bacteria, particularly Bifidobacterium infantis, which dominates the gut of breastfed infants. Formula-fed infants develop measurably different microbial communities. Within 2-3 years of life, the microbiome stabilizes into something resembling adult composition, with the early-life community shaping which species can establish over time.
Antibiotics are the most disruptive intervention to gut microbiomes that most people experience. A single course of broad-spectrum antibiotics can affect 30% or more of bacterial diversity, with measurable changes lasting 6 months to 2 years. Multiple courses produce cumulative effects; people who have had many antibiotic courses through childhood and adulthood show measurable reductions in microbial diversity that may not fully recover.
This isn’t an argument against antibiotics; they’re life-saving when needed. It’s an argument for using them only when necessary, and recognizing that the consequences of disrupting the microbiome can be substantial. The agricultural antibiotic problem compounds this: livestock raised on antibiotic-treated feed produces meat containing antibiotic residues, and the environmental release of antibiotics from farming and pharmaceutical waste contributes to broader antibiotic resistance and ecosystem-level microbial disruption.
Diet shapes the microbiome on the timescale of days. The 2014 Nature paper by Lawrence David at Duke showed that switching subjects between an animal-based diet and a plant-based diet produced measurable shifts in microbial community composition within 24 hours, with full transition completing within a week. This rapid response is part of what makes diet such a leveraged intervention for microbiome modification.
Specific dietary patterns produce specific microbial signatures:
Chronic psychological stress affects gut microbial composition through HPA axis activation and direct effects on gut function. Sleep disruption affects microbiome composition (this connects to the Sleep & Circadian Rhythm Basics coverage of how circadian disruption affects peripheral systems). Exercise has measurable effects on microbiome composition. Moderate exercise tends to increase diversity, though excessive endurance training can produce counterintuitive disruption through inflammatory mechanisms and altered gut permeability.
People living together share microbial communities to a measurable degree. People living with dogs have distinctly different microbiomes than people without pets. People living in urban versus rural environments differ. People who spend time outdoors versus indoors differ. The over-sanitization of modern environments, such as the routine use of antibacterial soaps, the elimination of soil contact, and the air-conditioning of every space, has been hypothesized to contribute to immune dysfunction by reducing exposure to environmental microbes that the immune system was designed to encounter. This is the hygiene hypothesis: the broader idea arguing that some environmental microbial exposure is protective rather than harmful, and while specific applications remain contested, the directional finding is supported by enough research to take seriously.
Worth specific treatment given the framework’s position established in Micronutrient Basics. Glyphosate has been shown to disrupt microbial communities through multiple mechanisms — its primary mode of action (inhibiting the shikimate pathway) affects bacteria as well as plants, since many gut bacteria use this pathway. Research has documented selective effects on microbiome composition, with reductions in beneficial organisms (Bifidobacterium, Lactobacillus, some Enterococcus species) and relative increases in potentially harmful organisms. The modern glyphosate-saturated food supply is both a cancer and microbiome concern.
This section is where the rebuild needs to be most pointed, because the current popular discourse on microbiome health includes substantial pseudoscience alongside legitimate practice.
The probiotic industry markets billions of dollars worth of supplements annually.
Where probiotics have reasonable evidence:
Where probiotics have weaker evidence:
Quality control problems parallel those covered in Micronutrient Basics for supplements generally. Many commercial probiotic products contain different strains or quantities than labeled, with dead organisms in many cases (probiotics need refrigeration and have limited shelf life). Look for refrigerated products, clearly identified strains (genus + species + strain identifier, not just genus), and reputable third-party tested manufacturers.
Some probiotic strains produce histamine through fermentation, particularly Lactobacillus casei, Lactobacillus reuteri, and Lactobacillus delbrueckii subsp. bulgaricus. People with histamine intolerance (typically due to insufficient diamine oxidase, the enzyme that breaks down histamine) can experience worsened symptoms from probiotic supplementation containing these strains. Symptoms include migraines, sinus issues, skin reactions, and digestive distress.
Histamine-degrading or histamine-neutral strains include Bifidobacterium infantis, Bifidobacterium longum, and Lactobacillus plantarum. People with histamine sensitivity should select probiotics carefully or focus on prebiotic fibre instead.
The Sonnenburg/Gardner 2021 Cell paper is particularly important here. Their randomized trial compared two interventions (a high-fibre diet and a fermented-foods diet) and found that fermented foods produced clearer benefits for microbiome diversity and reduced inflammation. The participants ate roughly six servings per day of fermented foods including yogurt, kefir, kombucha, kimchi, sauerkraut, and brine. Far more than typical Western consumption.
The implication: fermented food consumption may be more reliably effective than probiotic supplementation for general microbiome health. The mechanism is partly diversity (fermented foods contain hundreds of different bacterial species, versus the handful in most supplements) and partly the food matrix (other compounds in fermented foods may support microbial establishment).
Practical fermented foods worth incorporating:
The modern microbiome optimization protocols associated with Ben Greenfield, Olli Sovijärvi and the Biohacker’s Handbook, Andrew Huberman, and Dave Asprey have substantial overlap and substantial divergence. Some of their recommendations are well-grounded; others are concerning.
The well-grounded recommendations:
The reasonable but evidence-limited recommendations:
The recommendations that warrant scepticism:
The Greenfield protocol’s recommendation of LL-37 peptide subcutaneous injection for autoimmune and gut inflammation conditions deserves direct treatment. LL-37 is a real antimicrobial peptide produced by the human body; subcutaneous injection of synthetic LL-37 is not an established or regulated intervention. The peptide is sold through unregulated grey-market channels, isn’t FDA-approved for any indication, and is being recommended by practitioners for self-administration in protocols with no clinical trial evidence supporting them. It’s recommending unregulated injectable peptide use for self-treatment based on theoretical mechanism rather than demonstrated benefit. If you have a condition severe enough to warrant peptide therapy, work with a clinician operating under established medical practice.
The 5g glycine three to four times daily protocol (totaling 15–20g daily) for blocking LPS-induced inflammation has a partial research basis, but the dose recommendations exceed what the research supports. The Patrice Cani research on LPS and metabolic endotoxemia is important, but the specific high-dose glycine intervention as routine self-care isn’t well-supported. Modest glycine supplementation (3–5g daily) for sleep and as part of methionine balancing has reasonable support; daily 15–20g is in the territory where benefits are speculative, and side effects (nausea, GI distress) are documented.
The daily activated charcoal recommendation for “buffering immunity” or “removing endotoxins” is concerning. Activated charcoal binds compounds indiscriminately. It can be useful in acute poisoning emergencies, but daily consumption produces deficiencies in nutrients, medications, and beneficial compounds, along with whatever toxins it might bind. The “binders” framework presented in some practitioner protocols isn’t supported by evidence for routine daily use; activated charcoal, in particular, shouldn’t be a daily supplement.
The raw carrot daily protocol for reducing endotoxin absorption traces back to Ray Peat’s longstanding nutritional framework, which has been influential in some wellness circles. Peat’s underlying framework has interesting elements, but the specific raw carrot intervention has weak primary research support. There’s nothing harmful about eating raw carrots; treating them as a daily endotoxin-binding intervention overstates what the evidence supports.
The helminth therapy mentioned (parasite supplementation for autoimmune disease treatment) deserves clear framing. There’s a genuine research base. The Old Friends hypothesis suggests that the absence of parasitic worms in modern populations may contribute to autoimmune conditions, and clinical trials of helminth therapy for IBD and autoimmune conditions have shown some promise. But “helminth therapy” in the practical sense ranges from regulated clinical trials with specific organisms to unregulated commercial products of unknown quality. This is an experimental medicine that should be approached only through clinical research contexts.
The Huberman Lab podcast has produced substantial accessible content on microbiome health, much of which is a reasonable synthesis of primary research. The recommendations to eat fermented foods (drawing directly on the Sonnenburg/Gardner research), to prioritize sleep, to avoid processed foods, and to allow some environmental microbial exposure are well-grounded.
Where Huberman’s content runs slightly ahead of the underlying research is in the specificity of some recommendations, particularly prebiotic doses, specific timing protocols, and the precision with which gut-brain effects are characterized. The popularizer’s challenge of producing actionable advice from inconclusive research is a real difficulty; treating Huberman’s specific protocols as more precise than the underlying evidence supports is a recurring pattern. Engage critically rather than treating his protocols as gospel.
The intestinal lining is a single layer of cells (enterocytes) that separates the contents of your gut from your body. Tight junctions between these cells normally prevent unwanted material from passing through. When this barrier function is compromised, larger molecules (food particles, bacterial fragments, lipopolysaccharides) can enter the bloodstream and trigger immune responses.
The clinical phenomenon of increased intestinal permeability is real and well-characterized in specific conditions:
Alessio Fasano at Harvard has been the most influential researcher on intestinal permeability and zonulin biology, and his research is the actual foundation for much of what gets called “leaky gut” in popular contexts.
The popular “leaky gut” framework that subclinical permeability underlies most chronic disease and that specific protocols can “heal the gut lining” goes beyond what the research currently supports. The directional claim (chronic permeability contributes to chronic inflammation, which contributes to chronic disease) is reasonable. The specific causal claims (your X disease is caused by leaky gut, this protocol will fix it) often aren’t well-supported.
This is the actual research-anchored version of the “leaky gut” framework. Patrice Cani’s group at UCLouvain pioneered the metabolic endotoxemia concept: the idea that lipopolysaccharides (LPS) from gram-negative bacteria can cross a compromised intestinal barrier, enter circulation, and trigger systemic low-grade inflammation that contributes to metabolic disease. The mechanism is well-established at this point; the practical implications and interventions are still being characterized.
What’s clear:
What’s less clear:
The reasonable practical implications: avoid combinations that reliably increase LPS translocation (high-fat plus high-refined-carb meals, chronic alcohol use, very poor diet quality), support gut barrier function through fermented foods and fibre, and address specific symptoms through clinical workup rather than self-supplementation.
The popular practitioner recommendations for LPS reduction include various compounds with varying evidence quality:
Reasonable evidence:
Mixed/limited evidence:
Avoid as routine interventions:
Some specific microbiome-related conditions deserve direct treatment because they’re commonly mismanaged.
The small intestine normally has substantially fewer bacteria than the large intestine. SIBO occurs when bacterial populations expand inappropriately into the small intestine, where they interfere with digestion and absorption. The condition has been increasingly recognized in IBS populations, with some estimates suggesting up to 60% of IBS may involve SIBO.
Symptoms: Bloating, gas, abdominal pain, alternating constipation/diarrhea, food sensitivities, nutrient deficiencies (particularly iron and B12), brain fog, fatigue.
Risk factors: Low stomach acid (often from PPI use), poor migrating motor complex function, certain medications, history of food poisoning (which can damage the migrating motor complex permanently), previous gut surgery, hypothyroidism, and diabetes.
Diagnosis: Hydrogen and methane breath tests, ideally combined with symptom evaluation. False negatives are common.
Treatment: This is where the field has multiple approaches with varying evidence:
SIBO is a legitimate clinical condition that warrants proper diagnosis and treatment rather than self-management with random protocols. People who have suspected SIBO should work with a clinician who actually treats this. The field has moved substantially in the last decade, and approaches that worked five years ago may not be the current best practice.
Candida is a normal commensal yeast in the human microbiome; pathological overgrowth becomes problematic when something disrupts normal microbial competition. Risk factors include antibiotic use, high sugar consumption, immune compromise, and chronic stress.
The popular “Candida overgrowth” diagnosis as the cause of various non-specific symptoms has been overused. Genuine systemic candidiasis is a serious condition typically affecting immunocompromised patients. The widespread “I have a Candida problem” diagnosis often doesn’t reflect actual pathological overgrowth. It can reflect general dysbiosis, food sensitivities, or other issues misattributed.
That said, intestinal Candida overgrowth is a real phenomenon that responds to specific interventions:
A note: Candida can live on either glucose or ketones, so going into ketosis doesn’t “starve” it. The general advice: if you suspect a Candida problem, get a proper diagnosis (stool testing, organic acid testing) rather than self-treating with elaborate protocols.
Many food sensitivities involve interactions between food components and microbiome composition. People with low DAO (diamine oxidase) activity react to histamine-rich foods. People with FODMAP sensitivity often have specific microbial patterns that produce excessive gas with these substrates. Lactose intolerance in adults reflects normal post-weaning loss of lactase (the enzyme that digests lactose) in populations without lactase persistence. The bacterial fermentation of undigested lactose produces the symptoms.
The point: many “food sensitivities” reflect the interaction between specific food components and your specific microbiome composition. Modifying the microbiome (through diet, time, sometimes targeted interventions) can sometimes change tolerance for foods that previously caused issues.
After all the complexity, the practical recommendations are simpler than the field’s popular discourse suggests.
Eat diverse whole foods.
Eat fermented foods regularly.
Feed your microbiome with prebiotic fibre.
Be cautious with antibiotics.
Reduce things that disrupt the microbiome.
Consider supplementation strategically.
Address specific conditions specifically.
Don’t fall for the practitioner protocols that lack evidence.