Prebiotics: what actually feeds the microbiome

You are standing in the kitchen reading the back of a tub. Inulin from chicory root, it says. Below that a sentence about a healthy microbiome, next to it a leaf in green.

You thought this through. You do not want to suppress anything, you want to build something up. So you stir a spoonful into water.

Four days later you are sitting on the sofa in the evening, undoing the top button of your trousers, thinking: this cannot be the right direction.

Many people know this pattern. The belly is fuller, not lighter. The flatulence starts earlier in the day. And the obvious conclusion is then: I simply do not tolerate fibre.

That conclusion is understandable, but it stops too early. The evidence says something more precise, and that is the core message of this article: it can make a measurable difference which substrate you take, not only how much.

So this is not about prebiotics for everyone. It is about the right substrate, at the right time, for a particular gut.

Four terms that constantly get mixed up

Picture the shelf in the pharmacy. Four tubs side by side. One says probiotic, the next prebiotic, then synbiotic, then postbiotic.

The words sound like variants of the same product. They describe four completely different things, though, and the difference is not word play. It decides what you are actually holding.

The terms do not come from marketing. They come from ISAPP, an international expert body for probiotics and prebiotics that sets its definitions in consensus papers and revises them regularly.

Probiotics are the residents

Under the ISAPP consensus, a probiotic is defined as a live microorganism that confers a health benefit when administered in adequate amounts. Three conditions: alive, adequately dosed, with a documented benefit.

More on that is already in the cluster. Which capsule form survives the stomach passage better and what sets spore formers apart from classic strains is covered in Probiotics: spore form or capsule. Here we stay with the boundary.

Prebiotics are the food

The definition currently in force comes from the ISAPP consensus paper of 2017. A prebiotic is a substrate that is selectively utilised by host microorganisms and that confers a health benefit.

Synbiotics are the combination, and there are two designs

Under the ISAPP paper of 2020, a synbiotic is a mixture of live microorganisms and substrates that are selectively utilised. The interesting part is the distinction inside it.

Complementary means: both parts meet their respective criteria on their own. A documented probiotic plus a documented prebiotic, each with its own data base, both together in one tub.

Synergistic means: the substrate was chosen specifically for the microorganisms it comes with. In that case it does not have to meet the individual criteria of a prebiotic, because it was never meant as a standalone prebiotic.

In practice that means: the word synbiotic on a pack does not yet tell you whether the components were matched to each other or merely placed side by side.

Postbiotics are what is left over

The ISAPP definition of 2021 reads: a preparation of inanimate microorganisms and their components that confers a health benefit. With or without metabolites.

There is a thinking error hidden here that turns up very often online.

And now you know why four similar sounding tubs on the same shelf answer four different questions.

What separates a prebiotic from a dietary fibre

The next question follows almost automatically: is every dietary fibre a prebiotic, then?

The answer is no. And the reason is more interesting than it first sounds.

A dietary fibre is a definition from nutrition science: indigestible for our own enzymes. A prebiotic is a definition from microbiology: selectively utilisable by beneficial microorganisms, with a documented benefit.

The first says something about what we cannot do. The second says something about what they can do.

Wheat bran as the textbook case

Wheat bran is a classic dietary fibre. It binds water, increases stool bulk and speeds up transit. That is a physical effect, not a microbial one.

This is exactly where things get methodologically difficult, and the German S3 guideline on irritable bowel syndrome names it with remarkable openness.

Translated: when a fibre is helpful, nobody knows for certain whether that comes from the fermentation or from the bulk. As long as that stays open, a prebiotic is hard to evaluate as a tool in its own right.

That is not a sign of ignorance. That is clean methodology, and it is the reason for the cautious guideline position we will look at more closely later.

Soluble beats insoluble

The question of amount, meaning the famous 30 grams a day, is already covered at length in Fibre: which myths hold up. Here we stay with the question of quality.

And now you know why the question about the working principle matters more in this debate than any number of grams.

The substrates one by one: what each can do and how well it is studied

Now it gets concrete. Because prebiotic is an umbrella term for very different molecules, and their evidence base is spread very unevenly.

Some have been tested in dozens of trials. Some rest on a single dose finding study from industry. Both sit on the same shelf, often in the same price range.

One image first, to make the differences tangible: picture the colon as a long road. Some substrates are matchsticks. They burn immediately, right at the start of the road, with a lot of gas. Others are logs. They glow slowly and arrive further down.

Inulin and oligofructose: best studied, most often misunderstood

Inulin is the substance that opens almost every discussion about prebiotics. It sits in chicory root, in Jerusalem artichoke, in globe artichokes, in onions and leek. Oligofructose is the short chain version of it.

With constipation the picture looks a little friendlier, but here too a closer look pays off.

Galactooligosaccharides: less was better

Galactooligosaccharides, GOS for short, are made from lactose and occur naturally in pulses. Breast milk contains a related group of compounds, which explains their popularity.

Resistant starch: four types, a chapter of its own

Resistant starch is starch that reaches the small intestine undigested. The literature distinguishes four types, and that classification is established enough to describe here without an extra citation.

RS1 is physically enclosed, for instance in whole grains and seeds. RS2 sits in the natural granule structure, for example in green banana and raw potato. RS3 only forms when cooked starch cools down, which is retrogradation. RS4 is chemically modified and comes from food technology.

RS3 is the form that kitchens and forums argue about. It gets its own section shortly, because that is where most of the numbers circulate that nobody can document.

Partially hydrolysed guar gum: the quiet heavyweight

Hardly anyone knows it. In the irritable bowel literature it is nevertheless the best studied prebiotic substrate. It comes from the seed of the guar bean and is enzymatically shortened so that it no longer gels.

An important limitation: this trial was open label, so not blinded. The switching endpoint is vulnerable to expectation effects. It stands here anyway, because it asks an honest question, namely which substrate people voluntarily keep taking.

In children there is a signal pointing the same way, although from a pilot study: 60 children between 8 and 16 years with functional abdominal pain, four weeks, efficacy 43 versus 5 percent compared with placebo (Romano 2013, PMID 23345946). Thirty children per arm are too few for a strong statement. As a hint it still counts.

Acacia fibre: gentle in reputation, thin in evidence

Acacia fibre, that is gum arabic, is regarded in practice as particularly well tolerated. It earned that reputation over years. The study base behind it is modest.

Beta glucans: the hardest clinical endpoint, only not in the gut

Beta glucans from oats and barley are soluble fibres that form a viscous layer in the small intestine and are fermented in the colon.

Pectin and polyphenols: the two outliers

Pectin sits in apples, citrus peel and carrots and is undisputedly a soluble fibre. For this article I did not find a robust randomised trial with a clinical gut endpoint. So it stands here as a substrate, without a statement about effect. That is the honest version.

Polyphenols from tea, berries, grapes, cocoa and pomegranate are a special case. They are largely not absorbed in the small intestine and reach the colon, where bacteria transform them.

A review from 2026 summarises the evidence on polyphenols and the microbiome and speaks explicitly of prebiotic like effects, while calling for further mechanistic and longitudinal studies (Rakhra 2026, PMID 41622499). That is exactly how it should be phrased.

And now you know why two people with the same powder can have completely different experiences.

Resistant starch from the kitchen: what really forms on cooling

Hardly any nutrition tip has spread as fast as this one: boil potatoes, let them cool, eat them the next day. Resistant starch is said to form, and it is said to feed the gut.

The tip is right in principle. The numbers circulating online mostly are not.

What happens physically as it cools

During cooking the starch granules swell and lose their order. The chains lie loosely side by side and our amylases reach them easily.

As it cools, these chains rearrange and pack into dense crystalline regions. That process is called retrogradation. Picture woollen threads drifting chaotically in warm water and pulling together into a firm braid as they cool.

Our enzymes reach that braid less easily. The starch travels further back and lands with the bacteria. That is RS3.

Two numbers that are everywhere and documented nowhere

Two figures keep turning up on health insurer pages, advice portals and magazine sites. For neither of them did I find a traceable primary study.

The first is the twelve hour rule, according to which resistant starch forms after about twelve hours. What is documented are the values above: 10 hours at room temperature and 24 hours in the fridge, with different results. A clean threshold at twelve hours is not what the data give.

The second is the claim that cooled potatoes have 3.5 kilocalories less per 100 grams. No primary source could be found for that either. I list it here only in order to explicitly not adopt it.

And what does that do to the gut?

This is where it gets interesting, because the most honest study on the whole topic carries a small disappointment.

And now you know why yesterday's potato salad remains a good idea but is not a therapy.

What comes out of it: butyrate, propionate, acetate

Up to here it was about the food. Now it is about what the bacteria make from it.

When microorganisms ferment fibres, short chain fatty acids are produced. Above all three: acetate, propionate and butyrate. Acetate mostly goes into the bloodstream, propionate is processed further in the liver, and butyrate largely stays on site.

Butyrate is therefore the substance that gets the most attention. The reason for that comes from a paper carried out in a mouse model.

Everything else about butyrate, the mucosa and the supplement questions is covered at length in L-glutamine and butyrate for the gut lining. Here it is only about the path from substrate to metabolite.

Not everyone makes the same thing from the same food

An observation from a Dutch paper that tested two fibre mixtures against single substrates fits with this.

And now you know why the same fibre can feed two different metabolic routes in two people.

Why prebiotics make things worse for some people at first

Now comes the section this article hangs on.

Because the most common feedback on prebiotic powders is not I feel better. It is: at the start it was worse, and at some point I dropped it.

Two explanations for that circulate online. One says: that is an initial worsening, you have to push through. The other says: you are feeding your overgrowth.

The first explanation is not documented. The second is plausible but likewise not cleanly documented. What has actually been measured is a third thing, and it is more interesting than either.

What the largest analysis shows

The second large analysis did not even arrive at a number for prebiotics.

The measured mechanism: not more gas, more sensation

How this heightened sensitivity develops and what may influence it is covered in Irritable bowel: the search for causes. Here the mechanism is enough.

The order: why building up before calming down often goes wrong

In my view this is the practically most important point of the whole topic.

Many people start with the building up part, because it sounds like the right idea. Do not suppress, nourish. Do not take away, give.

That is a good attitude. It is just the second question here, not the first.

If there are too many bacteria in the small intestine, if methane producers are slowing transit, or if the bowel wall overreacts to normal stretch, then a rapidly fermentable substrate meets the least favourable possible starting point. It gets processed at the front, where it does not belong, or it arrives in a belly that already reports every small change in volume.

So the order here may decide more than the amount.

There is a separate text on the stomach acid question: Low stomach acid and betaine HCl. On the situations from step 2 you will find SIBO in the small intestine, IMO and methanogens and Bile and bile acids. And if it is about leaving something out for a limited period, the approach is described in Using FODMAP properly.

And what about SIBO?

The honest answer is unsatisfying. There are no robust studies on prebiotics in bacterial overgrowth. And even the diagnostic work up is methodologically contested.

A second review arrives independently at the same caution: cumulative hydrogen production after lactulose was no higher in people with irritable bowel syndrome than in controls (Yao 2018, PMID 29035975). More gas is simply not automatically more complaint.

An overview from 2025 covering 58 meta-analyses on eleven dietary interventions in irritable bowel syndrome classifies this similarly: most effects were small and the certainty of evidence under GRADE was low to very low. Moderate certainty was found only for abdominal pain under probiotics and for symptoms under a low FODMAP approach (Zeraattalab-Motlagh 2025, PMID 39110917).

And now you know why the same spoonful of powder can lead to three very different evenings in three people.

Fermented foods: the second route

There is a second route to the microbiome, and it does not run through a powder but through a jar in the fridge.

Sauerkraut, kimchi, kefir, yoghurt, miso, tempeh, kombucha. These foods are created through intentional microbial growth and enzymatic conversion. That is how the ISAPP consensus paper of 2021 defines them (Marco 2021, PMID 33398112).

What matters is the boundary that the same paper sharpens explicitly: a fermented food is not a probiotic. Sauerkraut contains microorganisms, but without a defined strain, without a defined dose, without a documented benefit for that particular strain.

That is not a put down. It is a category of its own, and it has one of the most interesting studies of recent years on its side.

Honesty requires this too: 18 people per group is a small sample, all were healthy, and the endpoints were surrogate markers. No transfer to people with bowel disease can be drawn from that.

And now you know why the jar in the fridge deserves a category of its own and is not a replacement for the plate.

Akkermansia muciniphila and the mucus layer

One name has been turning up everywhere in microbiome discussions for some years now: Akkermansia muciniphila. The second part of the name means mucus loving, and that describes it well.

This bacterium does not live in the middle of the bowel contents but in the mucus layer that covers the intestinal epithelium. It feeds on mucin, meaning on that very mucus.

At first that sounds the wrong way round. A bacterium that eats our protective layer is supposed to be useful? The explanation may lie in the balance: moderate mucin breakdown can stimulate new production, so the layer stays in motion and stays dense. Only when too many mucus eaters find too little other food can the ratio tip.

How closely the mucus layer and barrier function are linked is covered at length in Leaky gut, intestinal permeability and zonulin.

The contradiction I am not smoothing over

Now comes the point where the evidence gets uncomfortable.

And now you know why the most discussed gut bacterium of recent years is not an argument for a capsule but one for the plate.

Diversity on the plate: what that means in practice

That leaves the question everyone asks in the end: so what do I eat?

I am deliberately not giving you a number of weeks or a target amount. Not out of reticence, but because the data do not support it.

The argument for that is already above. Fibres do not promote the microbiome in general, they promote particular strains. Those who have them might benefit more. Those who do not, less. And nobody knows in advance which line up sits in your colon.

From that follows no numerical target but a strategy: breadth instead of amount. Different substrates can reach different strains and different sections of the colon.

That is not a plan. That is a map. How much of it fits in which phase is an individual question.

Two things matter to me here.

First: the framework stays the same. Safety and assessment first, then the question of the situation, then selection and breadth. Turning that order around can make things unnecessarily hard, and the belly often joins in.

Second: a permanently low fibre diet is not a goal, at most a way station. The animal data on the mucus layer are an argument for keeping the road back to breadth open. The ACG guideline also stresses explicitly that a restrictive approach should stay time limited.

If you want to read on: the question of amount is covered in Fibre: which myths hold up, the question of inflammation promoting patterns in Inflammatory foods, and how much sleep and rhythm play into this in Sleep and the microbiome.

And now you know why diversity is the most honest recommendation this field allows to be drawn from the evidence.

Common questions about prebiotics and resistant starch

Where this topic connects to the rest of the body

What gets fermented in the colon does not stay in the colon. It is tied to the mucosa, to the immune system, to the nervous system and to the question of how well nutrients arrive at all.