Does prebiotic fibre actually raise Akkermansia? The human answer is no, or not yet

Search for inulin and Akkermansia and you will find a great many pages telling you the first feeds the second. We sell inulin. We are not going to tell you that.

Here is what the trials actually found.

The mouse evidence, which is real

In 2011 Amandine Everard and colleagues fed prebiotic fibre to two different mouse models of obesity, genetically obese mice and diet-induced leptin-resistant mice, and tracked what happened to both the gut community and the animals' glucose and lipid handling (Everard et al., 2011). The microbial composition changed, and so did the metabolism. Prebiotic fibre is not inert; it reorganises a gut.

That work sits alongside the fibre-deprivation experiments described in the previous post, where removing fibre pushed a defined community into eating the host's mucus layer instead (Desai et al., 2016). Both point the same way: what you feed the community determines what the community does.

Both are mice.

The human evidence, which is thinner and more awkward

The cleanest human test of the prebiotic idea came in 2013, when Evelyne Dewulf and colleagues ran a double-blind trial of inulin-type fructans, a 50/50 inulin and oligofructose mix, in 30 women with obesity, published in Gut (Dewulf et al., 2013).

It worked, in the sense that it did something specific rather than nothing. The microbiota shifted selectively. Two organisms rose clearly: Bifidobacterium, which is the classic and expected prebiotic response, and Faecalibacterium prausnitzii, a well-regarded butyrate producer. Both correlated inversely with markers the researchers were watching. The metabolic changes in the women themselves were, in the authors' own word, modest.

Akkermansia muciniphila is not among the organisms reported as responding.

That absence is the part worth sitting with. This was a well-run trial of exactly the fibre type most often described online as Akkermansia food, in humans, with the microbiota measured, and the headline result was about other bacteria.

The awkward bit gets worse when you look at the biology

There is a reason these trials keep coming up quiet, and it is not bad luck.

In 2018 Ose and colleagues did the obvious experiment nobody quotes: they took six oligosaccharides and tested them against 31 of the main human gut anaerobes, one organism at a time, to see which bacteria could actually metabolise which fibres (Ose et al., 2018). Bifidobacteria ate fructo-oligosaccharides readily, which is exactly why Bifidobacterium is the reliable prebiotic responder.

Akkermansia muciniphila did not grow significantly on any of the six.

That is a fairly direct answer to the question in the title. Inulin is not Akkermansia food, because Akkermansia largely cannot eat it. It eats mucin. Any effect fibre has on it has to be indirect, routed through the rest of the community or through the mucus layer itself, which is a longer and looser causal chain than a label implies when it says "feeds your Akkermansia".

And then the trial that inverts the whole idea

In 2020 Rodriguez and colleagues, working in the same Louvain group that originated much of this field, ran native inulin against a placebo in people with obesity and asked a different question: not what does inulin change, but who does inulin work for (Rodriguez et al., 2020).

The answer, in their words: the pre-intervention levels of Anaerostipes, Akkermansia and Butyricicoccus drove the decrease in body mass index in response to inulin. Their conclusion is that you should characterise someone's microbiota before giving them a prebiotic.

Read that carefully, because it points the opposite way to the marketing. Akkermansia here is not the thing inulin produces. It is one of the things that predicts whether inulin will do anything for you at all. It is the starting condition, not the result.

What that means, carefully

None of this proves inulin cannot affect Akkermansia in a person. Dewulf's was a small trial in a specific population; absence from a reported result is not proof of no effect; growth in a test tube is not the same as growth in a colon; and abundance measured in stool is an imperfect proxy for what happens in the mucus layer, which is where this organism actually lives.

But "we cannot rule it out" is a very different sentence from "inulin feeds Akkermansia", and only one of them is honest about the state of the evidence.

The reasonable position, as far as we can tell, is this. Fibre availability shapes whether gut bacteria eat your food or your mucus — demonstrated, in mice. Inulin-type fructans reliably shift the human gut community toward Bifidobacterium and F. prausnitzii — demonstrated, in a small trial. Akkermansia does not appear to consume these fibres directly, and in the trial that looked hardest it behaved as a predictor of response rather than a product of it. That more Akkermansia follows from taking inulin is not established.

So why do we sell it

Because chicory inulin is a good ingredient on its own terms, and it has an authorised claim that rests on its own evidence rather than on ours: chicory inulin contributes to normal bowel function by increasing stool frequency, at 12 g a day. That is what the European Union assessed and permitted, and it is what our Chicory Inulin label says.

We think the environment argument is genuinely interesting, and we would rather set it out with its gaps visible than quietly imply the human data says something it doesn't. If a trial lands showing inulin raises Akkermansia in people, we will write that post gladly, and we will link the paper.

Until then the fibre is fibre, and the bacterium comes in a different bottle.