Fibre and the Microbiome

TL;DR. Fibre is the part of plant food your own enzymes cannot digest, and it is the single most consistently under-eaten thing in modern diets: most populations eat about half the recommended amount. It does mechanical work in the gut and, more importantly, feeds the bacteria in your colon, who ferment it into short-chain fatty acids that nourish your gut lining and signal to your immune system. The evidence linking fibre intake to lower rates of heart disease, type 2 diabetes, colorectal cancer, and total mortality is among the strongest in nutrition. The microbiome field around it is genuinely exciting and, so far, has produced far more press releases than treatments.

Key takeaways

• Recommended intake is around 30 g a day; typical intake in the UK and US is 15 to 20 g. This gap is one of the clearest nutritional deficits in wealthy countries.
• Soluble and insoluble is a useful but crude split. Fermentability and viscosity are the properties that actually predict effects.
• Butyrate, produced by bacteria fermenting fibre, is the main fuel for colon cells. A low-fibre diet starves the gut lining directly.
• Diversity of plant foods predicts microbiome diversity better than quantity of any one fibre. The often-quoted "30 plants a week" target comes from the American Gut Project.
• Most probiotic claims outrun the evidence. A few specific strains for a few specific conditions have real support; most products do not.

What fibre actually is

In short: Plant carbohydrates that reach the colon intact, because human enzymes cannot cut their bonds.

From Chapter 1: starch and cellulose are both chains of glucose, and the only difference is the geometry of the link. Your amylase cuts α bonds and not β bonds. Everything with the wrong bond geometry, plus a few other polymers, arrives in the colon undigested. That is fibre.

The soluble/insoluble split, taught everywhere, is a laboratory classification based on whether the fibre dissolves in water. It correlates loosely with effects and misleads in important cases. The two properties that actually predict what a fibre does are:

• Viscosity: does it form a gel? Viscous fibres (beta-glucan, pectin, psyllium, guar) slow gastric emptying, slow glucose absorption, and trap bile acids, which is how they lower LDL cholesterol.
• Fermentability: do the bacteria eat it? Highly fermentable fibres (inulin, pectin, resistant starch, GOS) produce short-chain fatty acids and also produce gas.

Psyllium is the illustrative exception: it is soluble and very viscous but poorly fermentable, which is exactly why it is the fibre supplement of choice for irritable bowel syndrome. It gives the gel benefits without the gas.

What fibre does in the gut

In short: It adds bulk and holds water, slows absorption where it is viscous, traps bile acids, and, most importantly, feeds bacteria.

1. Bulk and transit. Insoluble fibre holds water and increases stool mass, which stretches the colon wall and stimulates the movement that propels it. Wheat bran is the most effective bulking agent by mass, and each gram of it can add several grams to stool weight. This is why insoluble fibre helps constipation and why it can worsen symptoms in an already irritated bowel.

2. Viscosity effects. Gel-forming fibres slow gastric emptying and create a physical barrier in the small intestine that slows the diffusion of glucose and fat to the absorptive surface. This flattens the post-meal glucose curve. It also traps bile acids and carries them out of the body, and since bile acids are made from cholesterol, the liver must draw on circulating cholesterol to replace them. That is the mechanism behind the LDL-lowering effect of oat beta-glucan, which is one of the few health claims allowed on food packaging in both the EU and US: 3 g a day of oat beta-glucan lowers LDL by roughly 5 to 10 percent.

3. Feeding the microbiome, which is where most of the interesting biology is.

The microbiome, without the hype

In short: Roughly 38 trillion bacteria in the colon, dominated by a few phyla, whose main service is fermenting what you cannot digest.

Current estimates put the number of bacterial cells in the body at around 38 trillion, against roughly 30 trillion human cells, so the "ten times more bacteria than cells" statistic that circulated for decades was an overestimate that has been corrected. It is roughly one to one, which is still remarkable.

Almost all of them live in the colon, in near-total absence of oxygen, at about 10¹¹ cells per gram of contents. A typical adult carries several hundred species, dominated by two phyla: Bacillota (formerly Firmicutes) and Bacteroidota (formerly Bacteroidetes), with smaller contributions from Actinomycetota (including Bifidobacterium), Pseudomonadota, and Verrucomicrobiota (including Akkermansia muciniphila).

Two facts to hold onto. Individual microbiomes differ enormously: two healthy people may share only a modest fraction of their species. And function is more conserved than taxonomy: different species doing the same job, so what the community does matters more than exactly who is in it.

What the microbiome actually does for you

Fermentation into short-chain fatty acids. The central service. Bacteria ferment fibre and resistant starch anaerobically, producing mainly:

Butyrate is the one to know. Colonocytes derive most of their energy from butyrate rather than from blood glucose, which is unusual for a human cell. So a diet with almost no fermentable fibre literally underfeeds the lining of your own colon, thinning the mucus layer and weakening the barrier. There is good animal evidence that fibre-deprived microbes begin consuming the intestinal mucus layer instead, bringing bacteria closer to the epithelium and promoting inflammation.

Other functions:

• Vitamin synthesis: vitamin K2 and several B vitamins (biotin, folate, B12, riboflavin), though how much is absorbed in the colon is debated and it should not be relied upon.
• Immune education: gut bacteria induce regulatory T cells, and germ-free animals have visibly underdeveloped immune systems. Much of the "hygiene hypothesis" literature about allergy and autoimmunity runs through this.
• Colonisation resistance: a dense resident community occupies niches and outcompetes incoming pathogens. Broad-spectrum antibiotics remove this, which is how Clostridioides difficile takes over.
• Bile acid transformation, which affects fat digestion and signalling.
• Drug metabolism, which is real and clinically significant: gut bacteria inactivate the Parkinson's drug levodopa, activate the prodrug sulfasalazine, and reactivate metabolites of the chemotherapy drug irinotecan, which causes its characteristic diarrhoea.
• Gut-brain signalling via the vagus nerve, immune mediators, and microbial metabolites. The animal work here is striking. The human work is early.

What shapes your microbiome

The honest state of microbiome science

This is a field where the gap between what is demonstrated and what is claimed is wide, so it is worth stating plainly.

Well established: the microbiome ferments fibre into SCFAs; it educates the immune system; antibiotics disrupt it; faecal microbiota transplantation cures recurrent C. difficile infection at high rates and is an approved therapy; diet changes composition quickly.

Plausible, being actively researched: links to obesity, type 2 diabetes, inflammatory bowel disease, colorectal cancer, allergy, and possibly mood.

Overstated: almost every consumer claim. Microbiome testing kits sold direct to consumers cannot currently tell you anything actionable, because there is no agreed definition of a healthy microbiome, results vary by sampling and sequencing method, and the recommendations generated from them are generic. Most probiotic supplements do not establish residence in the gut and their effects, where measurable, stop when you stop taking them.

The obesity story specifically is a cautionary example. Early work showing that transplanting microbiota from obese mice made lean mice fatter was genuinely exciting. Attempts to translate it to humans have been far less impressive, the widely cited Firmicutes-to-Bacteroidetes ratio has not held up as a marker, and human FMT trials for obesity have produced modest and inconsistent results.

Probiotics, prebiotics, and the rest of the -biotics

In short: Prebiotics are food for your existing bacteria and work reliably; probiotics are live bacteria and work for a short list of specific indications.

Where probiotics have real evidence, and it is strain-specific rather than species-specific, let alone generic:

• Antibiotic-associated diarrhoea: reasonably good evidence for Saccharomyces boulardii and certain Lactobacillus strains, with meaningful reductions in incidence.
• Acute infectious diarrhoea in children: modest reduction in duration, roughly a day, though two large 2018 trials found no benefit, which tempered enthusiasm considerably.
• Necrotising enterocolitis in preterm infants: this is the strongest evidence base in the whole field, with substantial reductions in a devastating condition.
• Irritable bowel syndrome: some strains help some people. Response is unpredictable.
• Pouchitis after colectomy: good evidence for a specific multi-strain formulation.

Where the evidence does not support the claims: general "gut health" in healthy people, immunity, mood, weight loss, and skin. A 2016 systematic review found no consistent effect of probiotics on the faecal microbiota composition of healthy adults, which is awkward for the entire premise of the consumer category.

Practical guidance if you want to try one: choose a product naming the specific strain (e.g. Lacticaseibacillus rhamnosus GG, not "Lactobacillus blend"), matched to a condition with evidence for that strain, at the dose used in the trials, and give it four weeks.

Prebiotics work more predictably because they do not need to survive, colonise, or compete. They just feed what is already there. Inulin, FOS, and GOS reliably increase Bifidobacterium. The dose-limiting problem is gas: 5 g is usually fine, 10 to 20 g causes noticeable bloating in many people. Which is the correct place to mention the exception.

When fibre is the problem: FODMAPs and IBS

In short: In irritable bowel syndrome, the fermentation that benefits most people causes pain and bloating, and a structured elimination can identify which fibres are responsible.

FODMAP stands for Fermentable Oligosaccharides, Disaccharides, Monosaccharides, And Polyols. They are short-chain carbohydrates that are poorly absorbed in the small intestine, draw water in osmotically, and are rapidly fermented in the colon:

In most people this is exactly what you want. In irritable bowel syndrome, where the gut is hypersensitive to distension, the same process produces pain, bloating, and altered bowel habit. The low-FODMAP diet, developed at Monash University, produces symptom improvement in roughly 50 to 75 percent of people with IBS, which is a strong result for a dietary intervention.

Two cautions that are frequently ignored. It is a diagnostic protocol, not a permanent diet: strict elimination for two to six weeks, then systematic reintroduction to identify personal triggers, then the least restrictive long-term diet possible. And staying on it permanently reduces fibre intake and measurably reduces beneficial bacteria, which is the opposite of what you want. It should be done with a dietitian.

How much fibre, and where to get it

In short: Around 30 g a day, from a variety of plant sources, increased gradually with water.

Recommendations: the UK sets 30 g a day for adults, the US sets 25 g for women and 38 g for men (14 g per 1,000 kcal), and the EFSA suggests at least 25 g. Actual intakes in both the UK and US average around 15 to 20 g. Estimates of Hadza hunter-gatherer intake run to 100 g a day, and of Palaeolithic diets similarly high.

The dose-response evidence is unusually good. A large 2019 Lancet meta-analysis commissioned by the WHO, covering 185 observational studies and 58 trials, found that people eating the most fibre had roughly 15 to 30 percent lower all-cause and cardiovascular mortality, coronary heart disease incidence, stroke, type 2 diabetes, and colorectal cancer, with benefit continuing to rise up to at least 25 to 29 g a day and evidence of further benefit beyond. Whole grains showed the same pattern. That is about as consistent as nutritional epidemiology gets.

The pattern is clear: pulses are the densest practical source by a wide margin, and a single cup of beans or lentils delivers half a day's fibre. Adding pulses to your week is the highest-leverage fibre change available, ahead of bran, supplements, and everything else.

The "30 plants a week" idea comes from the American Gut Project, which found that people eating more than 30 different plant species per week had more diverse gut microbiomes than those eating fewer than 10. It is observational and it makes a good target because it pushes toward variety rather than quantity. Herbs, spices, nuts, seeds, grains, and different varieties all count, which makes 30 much easier than it sounds.

Increase gradually. Going from 15 g to 35 g overnight produces gas, bloating, and cramping, because the bacterial community needs weeks to adapt. Add 5 g a week. And increase water alongside, because bulking fibres without adequate fluid can worsen constipation rather than relieve it, which is the most common way fibre advice backfires.

The bottom line

• Fibre is plant carbohydrate your enzymes cannot cut. Its two properties that matter are viscosity (which slows glucose and traps bile acids) and fermentability (which feeds bacteria).
• Colon cells run primarily on butyrate produced by bacteria fermenting fibre, so a low-fibre diet underfeeds your own gut lining.
• The evidence linking higher fibre intake to lower mortality, heart disease, diabetes, and colorectal cancer is among the strongest and most consistent in nutrition, and most people eat about half the recommended amount.
• Pulses are the densest practical source. Variety of plants matters as much as total grams.
• Prebiotics reliably feed your existing bacteria. Probiotics work for a short list of specific strains and indications and are otherwise oversold. Consumer microbiome tests are not yet actionable.

Sources and notes

Fibre classification, viscosity, and fermentability follow standard nutrition texts and the Institute of Medicine definitions. The dose-response mortality and disease evidence is Reynolds et al., The Lancet, 2019, commissioned by WHO. Oat beta-glucan and LDL is the basis of authorised health claims by EFSA and the FDA. Short-chain fatty acid production and butyrate as colonocyte fuel follow Roediger's original work and subsequent reviews. Microbial cell counts revised to roughly 1:1 with human cells follow Sender, Fuchs, and Milo, PLoS Biology, 2016. Fibre deprivation and mucus layer degradation follow Desai et al., Cell, 2016. Human milk oligosaccharides and Bifidobacterium follow Bode's reviews. Probiotic evidence by indication follows Cochrane reviews for antibiotic-associated diarrhoea and the preterm NEC literature; the two large negative paediatric gastroenteritis trials are Freedman et al. and Schnadower et al., New England Journal of Medicine, 2018. The finding that probiotics do not consistently alter healthy adult microbiota is Kristensen et al., Genome Medicine, 2016. Low-FODMAP evidence follows the Monash University programme. The 30-plants finding is from the American Gut Project, McDonald et al., mSystems, 2018.

Open questions. Almost nothing about the microbiome is yet actionable at the individual level: there is no agreed definition of a healthy microbiome, and consumer testing cannot currently generate useful recommendations. Whether the obesity associations seen in mice translate to humans has repeatedly failed to replicate.

👉 Next: the vitamins, all thirteen of them, with what each does and which ones people actually run short of.