The Human Microbiome: Gut Bacteria, Health, and Disease

The human body hosts roughly 38 trillion microbial cells — a figure that rivals the count of human cells themselves, according to a landmark 2016 recalculation published in Cell by Sender, Fuchs, and Milo. Most of those microbes live in the gut, and their collective genome encodes approximately 150 times more genes than the human genome. This page covers what the microbiome is, how it functions, what disrupts it, where scientific consensus gets genuinely complicated, and what the research actually supports versus what the supplement aisle would prefer people believe.

• Definition and scope
• Core mechanics or structure
• Causal relationships or drivers
• Classification boundaries
• Tradeoffs and tensions
• Common misconceptions
• Checklist or steps (non-advisory)
• Reference table or matrix

Definition and scope

The microbiome refers to the complete collection of microorganisms — bacteria, archaea, fungi, viruses, and protozoa — along with their genes and metabolic products, inhabiting a defined environment. In the human context, that environment is the body itself, with distinct microbial communities occupying the gut, skin, oral cavity, respiratory tract, and urogenital system.

The gut microbiome gets the most research attention for a straightforward anatomical reason: the large intestine alone harbors the densest microbial community on Earth, with bacterial concentrations reaching 10¹¹ cells per milliliter of colonic content (NIH Human Microbiome Project). The Human Microbiome Project, launched by the National Institutes of Health in 2007, formally characterized baseline microbial diversity across 242 healthy adult participants, establishing reference ranges across 18 body sites.

The term microbiome technically encompasses both the microorganisms (the microbiota) and their genetic material. In practice, the two terms are used interchangeably in most published literature, which causes occasional confusion when researchers are comparing community composition studies with functional metagenomic analyses. Worth knowing, but not worth losing sleep over.

Core mechanics or structure

Gut bacteria don't simply occupy space — they actively metabolize substrates that human digestive enzymes cannot process. Dietary fiber, for instance, passes intact into the large intestine, where anaerobic bacteria ferment it into short-chain fatty acids (SCFAs): primarily butyrate, propionate, and acetate. Butyrate serves as the primary energy source for colonocytes (the epithelial cells lining the colon) and plays a documented role in regulating intestinal inflammation (Ríos-Covián et al., 2016, Frontiers in Microbiology).

The gut microbiome is structured in layers of ecological complexity. At the phylum level, the adult gut is dominated by four phyla: Firmicutes, Bacteroidetes, Actinobacteria, and Proteobacteria, with Firmicutes and Bacteroidetes typically accounting for over 90% of bacterial abundance (NIH Human Microbiome Project). Below that, genus and species-level composition varies enormously between individuals — two healthy adults may share fewer than 30% of their gut bacterial species.

The microbiome also interacts with the immune system through pattern recognition. Toll-like receptors on intestinal epithelial and immune cells detect microbial-associated molecular patterns, calibrating inflammatory responses in ways that research has linked to systemic immune tone, not just local gut immunity. The gut-associated lymphoid tissue (GALT) contains roughly 70% of the body's immune cells, which puts the gut microbiome at the center of conversations about immunity that extend well beyond digestion.

The how-it-works section of this site provides broader context on mechanistic biological processes for those approaching this material without a background in cell biology.

Causal relationships or drivers

Diet is the single most documented driver of gut microbiome composition. A 2015 study in Nature by Dahl et al. and related work demonstrated that a high-fiber diet reliably increases SCFA-producing bacteria like Roseburia and Faecalibacterium prausnitzii within weeks. Antibiotic use produces the opposite effect: a 5–7 day course of broad-spectrum antibiotics can reduce bacterial diversity by more than 25%, with some species requiring 6 months or longer to recover (Jernberg et al., ISME Journal, 2010).

Other well-supported drivers include:

• Delivery mode at birth. Vaginally delivered infants are colonized primarily by maternal vaginal and fecal bacteria (Lactobacillus, Bifidobacterium); cesarean-delivered infants show higher proportions of skin and environmental bacteria (Staphylococcus, Clostridium), a difference documented by Dominguez-Bello et al. in PNAS, 2010.
• Breastfeeding. Human milk oligosaccharides (HMOs) selectively feed Bifidobacterium infantis, an infant-specific strain that degrades these sugars and produces beneficial metabolites.
• Geographic and ethnic variation. The Hadza hunter-gatherers of Tanzania carry gut microbiomes roughly 40% more diverse than Western urban populations, based on data from Sonnenburg et al. at Stanford, illustrating how industrialized diets have narrowed microbial range.
• Host genetics. Twin studies suggest heritability accounts for roughly 20–40% of microbiome composition variability, with the remainder explained by environmental and dietary factors (Goodrich et al., Cell, 2014).

Classification boundaries

The microbiome field draws several distinctions worth keeping clear.

Microbiome vs. microbiota: Microbiota = the organisms. Microbiome = the organisms plus their genes and metabolic environment. Most research papers use both terms; context clarifies which dimension is under study.

Core microbiome vs. variable microbiome: A "core" microbiome — a universal set of species shared across all healthy humans — has not been conclusively identified at the species level. What has been identified is a shared set of metabolic functions that different species can perform. This functional redundancy is biologically important and is sometimes misrepresented as species-level universality.

Dysbiosis: The term refers to a microbial community state associated with disease, characterized by reduced diversity, loss of beneficial taxa, or expansion of potentially harmful ones. Dysbiosis is a descriptive, not mechanistic, term — it identifies a pattern, not a cause. Whether dysbiosis drives a condition or results from it remains an open question in most disease contexts. The relationship between gut microbiome science and broader bioscience topics covered on this site reflects that same complexity: correlation is often the first finding, causation the harder work.

Tradeoffs and tensions

The gut microbiome field is scientifically productive and commercially over-exploited, which creates real interpretive noise. Three genuine tensions in the research:

1. Mouse models vs. human translation. A disproportionate share of microbiome-disease mechanistic data comes from germ-free mouse models, which lack any microbiome and are then colonized experimentally. These models have generated compelling data on conditions from obesity to anxiety. However, mouse gut physiology differs from human anatomy in transit time, pH, and immune architecture. Results that hold in mice have failed to replicate in human trials at a rate that has produced frank criticism within the field (Nguyen et al., World Journal of Gastroenterology, 2015).

2. Fecal microbiota transplantation (FMT) scope. FMT has demonstrated efficacy for recurrent Clostridioides difficile infection, with clinical success rates exceeding 80% in randomized controlled trials (FDA, C. diff FMT Guidance*, 2022). Claims that FMT treats autism, depression, or autoimmune conditions remain investigational and lack the controlled trial evidence that supports the C. diff indication.

3. Probiotic generalizability. The FDA does not evaluate probiotics as drugs unless specific health claims are made. The clinical evidence for commercially available probiotic strains is strain-specific, condition-specific, and dose-specific — meaning that data supporting Lactobacillus rhamnosus GG for antibiotic-associated diarrhea does not extend to other strains or other conditions without independent evidence.

Common misconceptions

"More bacteria is always better." Diversity is generally associated with health, but sheer abundance of any single taxon is not. Helicobacter pylori colonizes roughly 44% of the global population and is associated with gastric ulcer disease and gastric cancer — its presence is decidedly not beneficial.

"Fermented foods always add bacteria to the gut." Many commercially pasteurized fermented products (most supermarket yogurts, for instance) contain heat-killed organisms by the time of consumption. Whether live cultures survive gastric acid to reach the large intestine in functional numbers is not guaranteed.

"The gut-brain axis proves probiotics treat depression." The gut-brain axis — the bidirectional communication network linking enteric nervous system, vagus nerve, immune signaling, and central nervous system — is a real and documented structure. The jump from "this axis exists" to "probiotics reliably treat mood disorders in humans" outpaces the clinical evidence available as of the most recent systematic reviews (Dinan & Cryan, Nature Reviews Gastroenterology & Hepatology, 2017).

"Antibiotics permanently destroy the microbiome." Most studies show substantial recovery within 1–2 months post-antibiotic, though full return to baseline composition can take longer and some taxa may remain reduced for 6–12 months (Jernberg et al., ISME Journal, 2010). The damage is real; the permanence is overstated.

Checklist or steps (non-advisory)

Key domains evaluated in microbiome clinical research studies:

Reference table or matrix

LGG = Lactobacillus rhamnosus GG; UC = ulcerative colitis; CD = Crohn's disease; RCT = randomized controlled trial