Is fiber really healthy? And when is it too much?

You meant it seriously. Whole grain bread instead of white. Oats with flaxseed. Two spoons of bran in the muesli, because that is supposed to be good for you.

And then you sit there in the evening, your belly is tight as a ball, and your bowel movements have got worse rather than better.

Many people know this pattern. In my consultations I hear the same sentence almost every week: but I am doing everything right. You probably are. The advice you are following just may not fit your gut. That is what this text is about: holding a good reputation honestly against the data, assumption by assumption.

The best reputation in the whole supermarket, and where it comes from

Thirty grams a day. That is the reference value of the German Nutrition Society. Since then the number sits on muesli boxes and in apps. Hardly anyone asks where it came from.

It came from East Africa, in the 1960s.

The British surgeon Denis Burkitt observed there that colorectal cancer, diverticula and constipation were rare. People ate a very fiber rich diet. Out of that grew an influential hypothesis: the West eats too little fiber, and that is where its bowel diseases come from.

The contrast is still there today. In 2015 the colorectal cancer rate among African Americans was around 65 per 100,000, among people in rural South Africa below 5 per 100,000 [Case Series, n=40]. Except that it is not only the fiber that differs there, but also fat, meat, movement and life expectancy.

And now you know why a number on a muesli box has a backstory that rarely gets told along with it.

What happens when you put a good reputation into a randomized trial

A hypothesis only becomes knowledge once somebody has tried to refute it. The test: if fiber protects against colorectal cancer, then in people whose adenoma has just been removed it should lower the return of these precursors.

In the same issue stood the blinded twin study. David Alberts and colleagues gave 1,429 people 13.5 or 2 grams of wheat bran daily. After a good three years: 47.0 versus 51.2 percent with a new adenoma, adjusted odds ratio 0.88, interval 0.70 to 1.11 [RCT, n=1,429]. The point estimate leans slightly in the protective direction, the interval includes the null.

In 2017 Cochrane pooled five randomized trials with 4,798 participants: risk ratio 1.04, interval 0.95 to 1.13 [Meta-analysis, k=5, n=4,798]. The authors rate their own evidence as low.

And now you know why a recommendation and the testing of that recommendation are two different things.

The other side, presented honestly and not as a straw man

At this point you might think: case closed. That would not be honest. There is a second, very large body of data, and it points in the opposite direction.

In 2019 Andrew Reynolds and his team analyzed 185 prospective studies and 58 controlled trials in the Lancet. In the observational data, all cause mortality, coronary heart disease, stroke, type 2 diabetes and colorectal cancer were 15 to 30 percent lower in the highest fiber consumers, with the largest risk reduction at 25 to 29 grams per day [Meta-analysis, k=243]. In the controlled parts, however, the endpoints were weight, blood pressure and cholesterol.

Then there is EPIC: 519,978 people from ten countries, relative risk in the highest fifth 0.75, calibrated 0.58 [Cohort, n=519,978]. That is where the statement comes from that doubling fiber intake could lower colorectal cancer risk by 40 percent.

People who eat a lot of whole grains, vegetables and legumes smoke less often on average, move more and go to screening more regularly. In that calculation, fiber is a marker for a whole pattern of living.

On other levels the data are better. In people who already have diabetes, an analysis of 42 controlled trials found HbA1c 2.00 mmol/mol lower on more fiber, and in the accompanying cohorts 14 fewer deaths per 1,000 participants [Meta-analysis, k=42, n=1,789]. More on that in the piece on blood sugar spikes.

And now you know why two types of study can sound so different without either of them getting the maths wrong.

Fiber is not one substance, it is a family that disagrees

Imagine somebody asks you whether metal is good for the body. You would ask back: which metal. Iron is something quite different from lead.

With fiber we almost never ask that question back. The nutrition table shows one number, and underneath it gather substances with opposing properties. The split into soluble and insoluble is not wrong, it just carries less than you might think.

How big this difference is in practice is shown by a Dutch trial. 275 people with irritable bowel syndrome received 10 grams of psyllium, 10 grams of bran or placebo for twelve weeks. In the first month, 57 percent responded to psyllium compared with 35 percent on placebo. Bran showed no robust advantage and had the most early dropouts, main reason: symptoms got worse [RCT, n=275].

Two fibers, the same amount, the same diagnosis, opposite experience.

The cholesterol side supports the viscosity axis as well. A meta-analysis across 28 randomized trials found LDL 0.33 mmol/l lower at a median of 10.2 grams of psyllium per day, around 13 mg/dl [Meta-analysis, k=28, n=1,924]. More on that in the cholesterol myth. And a 2026 network meta-analysis across 17 trials ranked viscous soluble fibers at the top for straining, frequency and consistency [Meta-analysis, k=17, n=1,423], with the advice to choose fiber by symptom rather than by a blanket amount.

And now you know why a spoon of bran and a spoon of psyllium do two different things in your gut.

The finding nobody likes: sometimes less can be more

There is one situation in which I regularly work against a reflex. Somebody has constipation, already eats a lot of fiber, and the next piece of advice is: even more fiber. Sometimes that goes well. Sometimes the belly gets harder.

The other side shows the same signal, just from the front. A meta-analysis across seven randomized trials with 287 people found a response rate of 77 percent on fiber versus 44 percent on placebo, but in the same data set significantly more bloating [Meta-analysis, k=7, n=287].

Both papers point from two directions towards the same thing: fiber can generate gas in constipation. For some people that is the price of a benefit. For others it seems to be the whole effect.

Tested under blinding, the picture gets even more sober. In a four week double blind trial in 250 adults with functional constipation, there was no time by group effect for stool frequency, Bristol score and straining [RCT, n=250]. Stool became softer, the number of bowel movements did not rise above placebo.

And now you know why the sentence just eat more fiber is not the answer for some people.

Why gas is not the same thing as pain

Two people eat the same portion of onions. One notices nothing. The other doubles over two hours later. The obvious explanation goes: she produces more gas. That is exactly what was measured, and the measurement does not support the explanation.

This is where the KPNI lens, clinical psychoneuroimmunology, adds something. The gut has its own nervous system with more nerve cells than the spinal cord. Between a stretch stimulus and a conscious sensation there are amplifier stages, and in visceral hypersensitivity they are turned up. Stress, lack of sleep and silent inflammation can turn that dial.

A picture for it: two apartments, the same doorbell. In one the volume knob sits at three, in the other at ten. It rings equally often and feels completely different.

That is exactly where the FODMAP logic comes in. In the reference trial, the symptom score in people with IBS fell to 22.8 millimeters on a low FODMAP diet, compared with 44.9 millimeters on a typical Australian diet [RCT, n=38]. In the healthy volunteers nothing changed. It is meant as a limited diagnostic phase: details in the low FODMAP guide and in the piece on IBS causes. Here too the type of fiber has a say: a meta-analysis across 14 trials with 906 people found a benefit with a number needed to treat of 10, which came entirely from soluble fiber [Meta-analysis, k=14, n=906].

And now you know why two people with the same amount of gas can have two completely different evenings.

When less can be an option, and when it clearly is not

Now comes the section where I choose my words most carefully. Here a body of evidence quickly turns into a set of instructions, and that would be wrong. In the specialist literature it is discussed for a few situations whether a time limited reduction of fiber intake might ease things. These are possibilities, not recommendations.

And now the counterweight. Carol Brotherton and colleagues surveyed 1,619 people with inflammatory bowel disease, all in remission at the start. In Crohn disease, those who did not avoid high fiber foods had around 40 percent fewer flares, adjusted odds ratio 0.59 [Cohort, n=1,619]. In ulcerative colitis no such association was found. It is self report, and the reverse direction is plausible, because people who feel worse avoid more.

And now you know why the answer to the question about the right amount of fiber almost always starts with a question back.

Diverticula: a story that was tested and did not hold

Hardly any story in digestive medicine is as catchy as this one. Too little fiber makes hard stool. Hard stool needs more pressure. More pressure pushes the bowel wall outwards at weak points. And there are your diverticula.

The story has one advantage: you can test it. That is exactly what a research group from North Carolina did.

The same group repeated the question in 2013 in a different sample, analysing only people who did not know about their diagnosis. Between fiber intake and diverticulosis, comparing 25 against 8 grams per day, no association was found: odds ratio 0.96 [Cohort, n=2,108]. That too remains a cross sectional view.

And now you know why a particularly catchy explanation sometimes survives so long for exactly that reason.

Butyrate, and why a capsule is not a plate

At this point you might get the impression that I consider fiber superfluous. The opposite is the case. The stronger argument simply lies elsewhere: not in the fiber, but in what your bacteria make out of it.

When fermentable fibers reach the colon, bacteria break them down into short chain fatty acids. One of them is butyrate, the main fuel of the colonic mucosa. A picture: you throw something into the compost that you cannot digest yourself, and out of the compost comes the heating for your house.

A 2011 mechanism review describes for butyrate a role in fluid transport in the intestinal epithelium, an easing of mucosal inflammation, a strengthening of the barrier, and a modulation of visceral sensitivity and motility [Mechanism Review].

The immunological line is cleanly studied, but it comes from animal models. In experiments in mice, butyrate triggered the differentiation of regulatory T cells and eased an experimental colitis [In vivo, mouse]. A beautiful piece of work. And a mouse.

The same goes for the flip side. In a gnotobiotic mouse model, chronic fiber deprivation made the microbiome switch to the host's mucus glycoproteins, the barrier thinned out, and a pathogen came closer to the epithelium [In vivo, mouse]. In humans, long term data are missing, see gut permeability and zonulin.

In humans there is at least one strong signal. In a two week diet swap between 20 African Americans and 20 rural South Africans, mucosal biomarkers of cancer risk as well as microbiome and metabolome reversed in both directions, with clearly more butyrate production in the group eating the fiber rich diet [Case Series, n=40]. What was measured, though, were biomarkers, not cases of disease. A 2023 review closes honestly: clinical attempts to raise butyrate levels in humans have produced mixed results [Mechanism Review]. More in the piece on L-glutamine and butyrate.

Why a blanket fiber recommendation helps so little

Three things decide at the same time, and none of them appears in the recommendation of 30 grams.

The amount. Below 10 grams per day, the pooled analysis showed a higher colorectal cancer risk, above 30 grams no additional advantage.

The type. Viscous or not, fermentable or not, coarse or fine. Two equally large portions of fiber can have opposite effects.

The person. In the double blind trial in 250 adults, baseline microbiome profiles helped predict who responded to which preparation [RCT, n=250]. The same fiber simply does not do the same thing in everyone.

And then there is the difference between a supplement and a plate. Per 10 grams of total fiber per day the relative risk for colorectal cancer was 0.90, per three servings of whole grains per day 0.83, while fruit and vegetable fiber on their own showed nothing [Systematic Review]. A plate brings matrix, water, structure, polyphenols and chewing work with it. A capsule brings one substance. You will find the same figure of thought in the pieces on unprocessed food and on the calorie myth.

If you would like your own situation looked at, you will find the option to book an appointment below this article.

And now you know why an honest answer to the question about the right amount of fiber begins with three questions back.

Frequently asked questions about fiber

Fiber in the bigger picture

The fiber question rarely hangs in the air on its own. It touches fermentation, barrier, motility and perception all at once.