Gut and skin: what the axis between belly and face actually holds up

You stand in the bathroom in the morning and look at your face. Red again. Those spots again that were not there last night.

The creams are on the shelf, three opened tubes of them. You changed the cleansing gel, then the water, then the pillow. At some point you type into your phone what might be going on in your belly.

And you land in a world that tells you: your skin is the mirror of your gut. With cleanse packages running over several months, with testimonials and with before and after pictures.

Many people know this pattern. And I understand why it pulls. It explains something that feels unexplained, and it gives you something to do.

I am writing this article because I take both seriously. The connection exists, in one place even with numbers that impress me. And at the same time the research base is far narrower than every second consumer page claims. I will show you both, and with the sample sizes.

Why belly and face have anything to do with each other at all

Start with a simple question. What do your gut lining and your facial skin have in common?

Both are boundary surfaces. Both separate your inside from a world full of bacteria. Both have to be tight and permeable at the same time, ready to defend and tolerant. And both arise in the embryo from tissue that takes on exactly this double task.

The skin is the smaller surface of the two. Your gut lining, spread out, measures a multiple of it, and the largest part of your immune system sits along it. When the alarm runs constantly down there, messenger substances travel through the whole body. These messengers know no boundaries between medical specialties.

That is the basic idea of the gut-skin axis. It is not new and it is not esoteric. It is old.

Two mechanisms that get quoted everywhere

Whenever you read about the gut-skin axis somewhere, almost always the same two routes turn up. Both are well studied. Both come from mouse experiments. That matters, and it is rarely mentioned.

Then there is the barrier itself. A more permeable gut wall could let more fragments of bacteria into the blood, and the immune system reacts to that. How robust this model is and what zonulin actually says is covered in detail in Leaky gut, gut permeability and zonulin. Here one sentence is enough: the barrier is a plausible candidate, not a proven link.

And now you know why I sort the rest of this article by sample sizes and not by stories.

Rosacea, the most striking finding in the whole field

If the gut-skin axis has a hard number anywhere, it is here.

Rosacea is the skin condition with the persistent facial redness, the widened small vessels and, in one form, the inflammatory papules and pustules. It comes in flares. It reacts to heat, alcohol, stress and sun. A suddenly shooting sensation of heat with redness that comes together with diarrhoea or a racing heart does, however, belong in a work-up and should not be filed away as rosacea, because rarer causes can sit behind it. And in many people it goes along with abdominal complaints that nobody connects with the face.

In 2008 a group in Genoa looked into whether that is more than coincidence.

Now comes the part that almost nobody quotes along with it, although it is the most important one.

The group also treated 16 people with rosacea whose breath test was negative with the same antibiotic. In 13 of 16 nothing changed on the skin.

Why does that matter so much? Because antibiotics have been used in rosacea for decades, including for their anti inflammatory action on the skin itself. If the rifaximin had simply been a general anti inflammatory, it would have had to change something in the test negative group too. It did not. That speaks for the effect hanging on the finding in the gut.

The same working group looked again years later, this time over a longer period.

What the German guideline writes about it

Here it gets interesting, because on this question the German S2k guideline on rosacea is not the opponent of the gut hypothesis. It is its best ally, and a very precise one.

It counts small intestinal bacterial overgrowth explicitly among the trigger factors, alongside Demodex mites, heat, cold, water vapour, spicy food, UV radiation, exercise, alcohol and cosmetics. It gives a number: in rosacea an overgrowth is present considerably more often, 46 to 51 percent compared with 5 to 23 percent. And it writes that the evidence is thickening for a dysregulation of the microbiota of the digestive tract as a trigger factor in a subgroup.

And then comes the sentence that carries this whole article.

Under research needs the same guideline lists, word for word, the clarification of a possible association between the microbiome of the gut and rosacea. That is not a rejection. That is an open question which a professional society considers important enough to write into its research agenda.

I find this stance exemplary, and I also understand it. An unblinded study from a single centre is not enough to recommend antibiotic therapy outside its licence. That is not stubbornness, that is the usual threshold of proof.

The counter evidence belongs in the same section

If I stopped here, you would have read a very convincing article. And an incomplete one.

And now you know why I look closely in papulopustular rosacea with concurrent abdominal complaints, and why with pure redness and no abdominal symptoms at all I do not, at least not at first.

What else plays a part in rosacea, and why the gut does not explain everything

There is one finding in rosacea that is numerically far stronger than anything from the abdomen. It is rarely mentioned on gut health pages, and when it is, it is often wrong.

Helicobacter pylori, the chapter with four answers

Hardly any germ is named more often in this context. And hardly any has such an inconsistent body of data.

Added to this is the Danish register cohort with a hazard ratio of 1.04, so nothing, and the American cross sectional study in which people with rosacea even had Helicobacter less often than the general population. The German guideline sums this up in two words: discussed controversially.

The triggers that actually appear in the guideline

Before you think of the gut in rosacea, it is worth a look at the list that affected people report themselves. The German guideline gives it with percentages.

Does anything strike you? The strongest single trigger is sunlight, the second strongest is mental tension. Neither appears in any gut cleanse offer.

On diet the guideline is very precise, and I like quoting it because it says two things at once. With the recommendation strength should and 100 percent agreement in the consensus procedure it advises avoiding foods that lead to vessel widening, for example alcohol, spicy food as well as hot food and drinks. In the same breath it states: even though no special rosacea diet can be recommended, these factors should be avoided.

In addition it names foods with plenty of biogenic amines, such as red wine or cheese, as a possible aggravating factor. If you wonder why exactly these foods keep turning up, the explanation is in Histamine intolerance and DAO deficiency.

Coeliac disease, inflammatory bowel disease and one important mistake

The Danish register cohort found a raised risk of coeliac disease in rosacea, hazard ratio 1.46. That is an association from which something actually follows, namely a question in the history taking.

What that means in practice is unspectacular and still important. Anyone with rosacea and persistent bowel complaints should have the bowel complaints taken seriously and not brushed aside as a side issue. The route leads to diagnostics, not to a cleanse. More on this in Crohn disease and ulcerative colitis viewed integratively.

And now you know why in rosacea I first ask about sun protection, sleep and triggers, before a breath test even comes up.

What a stool test shows in rosacea, and why little follows from it

At some point the question about the stool test comes up in almost every conversation. You send in a sample, get a chart back, and on it are bacterial names with bars.

For rosacea there are exactly two serious investigations of this kind. I will put them side by side, because together they say more than each does alone.

That leaves the study type that has been marketed for a few years as proof of causality: Mendelian randomisation. It uses genetic variants as a natural random allocation.

What follows from this for the stool test in skin problems? By current standing, no course of action. A microbiome profile can be interesting, it can raise questions, and it can generate insight in studies. A skin treatment cannot be derived from it. What a stool test can do and what it cannot is described in Stool testing, PCR and dysbiosis diagnostics.

And now you know why I do not recommend a microbiome stool test for a purely skin related concern.

Acne, where the gut ends and metabolism begins

In acne the situation is different from rosacea. Here there is a very well documented route from inside to outside. It just does not run through the gut flora.

Blemished skin, causes in the gut or in metabolism

The sebaceous gland is an organ that listens to hormones. Two signals can turn it up: androgens and the insulin like growth factor IGF-1. When a lot of insulin is on the move, IGF-1 can rise, and at the same time the binding proteins that otherwise bind it can fall. More free IGF-1 can mean more sebum, stronger keratinisation and more food for the bacteria in the follicle.

That is not a theory out of nowhere. It has been measured in three places.

And then there is a natural experiment that has stayed in my mind for years.

Milk, and why yoghurt drops out

A meta-analysis of observational studies by Mohadeseh Aghasi and colleagues compared the highest with the lowest intake category: dairy overall odds ratio 2.61, milk in total 1.48, low fat milk 1.25, skimmed milk 1.82. For yoghurt and cheese no significant association was found.

What stands out is that skimmed milk sits higher at 1.82 than milk overall at 1.48 and that fermented products drop out. A separate value for whole milk does not appear in this work, so the comparison has to be read cautiously. And the top value for dairy overall has a very wide confidence interval of 1.20 to 5.67, so it could be considerably smaller than the figure 2.61 makes it look. The pattern still fits the insulin explanation. What that means in practice and how robust observational data on milk are overall is described in Dairy products, healthy or not.

On blood sugar itself I keep it short here, because there is a separate article for it. What matters is not the individual sugar content but how fast and how high blood sugar rises after a meal and how often that happens in a day. Everything practical about it is in Avoiding blood sugar spikes. And if your skin changes with your cycle or is concentrated on the chin and jaw, the hormonal side is worth a look in Hormonal acne from within.

And how much of that is gut?

Honest answer: so far little, and the little there is is new.

And now you know why with blemished skin I first ask about breakfast and not about your bowel movement.

Atopic dermatitis, where prevention and treatment part ways

With atopic eczema there is one sentence that I consider the most useful in this whole article. It goes: preventing and treating are two different questions, and they have different answers.

It all started in Finland in 2001.

And that is exactly what happened. The halving turned into a considerably smaller effect in the pooled analyses, but one that stayed stable.

A second pooled analysis across 16 studies and 2,797 children arrived at a relative risk of 0.74 and named as its most important limitation that the studies defined eczema very differently. And an updated analysis across 21 studies found the effect above all with combination preparations, with intake by the mother and only up to about the age of two.

The most important sentence of this third analysis is a negative one: for asthma, hay fever, wheezing, allergic diseases overall and sensitisation, no significant effect was found.

The same group also looked at prebiotics, meaning the fibres that feed bacteria. It suggests them for infants who are not exclusively breastfed and advises against them in exclusively breastfed infants, both conditional and with very low certainty of evidence. On prebiotics in pregnancy or breastfeeding there was not a single study, which is why the panel deliberately gave no recommendation on that.

This one sentence says a lot about the ground on which prebiotics for the complexion are marketed in this country. What these substances actually do in the gut is described in Prebiotics and resistant starch.

And now the treatment

Up to here it has been about children who do not yet have eczema. The more common question is a different one: you already have atopic dermatitis. Do probiotics do anything?

Why the German guideline says no so clearly

The German S3 guideline on atopic dermatitis formulates a strong negative recommendation, consensus based and with 100 percent agreement: for treatment, general dietary measures such as supplements or a blanket avoidance of certain foods such as cow milk or gluten should not be used. The same applies to vitamins.

That sounds harsh. The reason behind it is not dogma, though, but safety, and I find it convincing.

One note belongs here, because gluten comes up. If a suspicion of coeliac disease is ever in the room for you or your child, then the diagnostic work-up belongs before the avoidance. A gluten free diet that begins beforehand can make the diagnosis impossible, because antibodies and mucosal findings recede under avoidance.

Hand eczema and dyshidrotic eczema, the little blisters on the palms and the sides of the fingers, are often assumed to come particularly clearly from the gut. Robust intervention studies on this do not exist. What applies to atopic eczema applies here all the more: first the dermatological classification, then everything else.

The hygiene hypothesis, cleanly classified

What makes this finding interesting for the skin nonetheless is the barrier parallel. In atopic eczema the skin barrier is disturbed, often through a predisposition in the filaggrin gene. The skin loses water, irritants and allergens get in more easily, and the immune system gets to know them through the skin instead of through the gut. This order is considered one explanation for why food allergies are more common in children with eczema. The parallel to the gut barrier is obvious, and it is exactly that: a parallel. More on this in Leaky gut and zonulin.

And now you know why with an existing atopic dermatitis I do not start with probiotics but with the barrier and the dermatological treatment.

Psoriasis, the finding with the clearest practical consequence

In psoriasis the data look different. Here there is no cleanse story, but there is a number from which something follows.

That is the only sentence in this whole article from which a clear action follows. And it leads not to a preparation but to an examination.

Why this fits together so well is understandable immunologically. Psoriasis and inflammatory bowel diseases share risk genes, they share signalling pathways in the immune system, and they are treated in part with the same antibody medications. Two diseases at two boundary surfaces, with a shared immunological ground. That is something other than a chain from gut to skin.

And now you know why this number is the practically most useful one in the whole field.

What follows from this in practice, and where the evidence ends

Now everything is on the table. Time to sort it.

When a gut work-up in skin problems may be worth it

The order I stand for across the whole cluster

Before anything else happens, I ask whether there is enough stomach acid up top, because it stands at the beginning of the digestive chain and is overlooked most often. Why that is so is described in Low stomach acid and betaine HCl. That is a question, not a recommendation for a self experiment. Acidifying preparations can do harm with a gastric ulcer, with gastritis and alongside anti inflammatory painkillers, so this question belongs in an examination and not in a trial run. And it is explicitly not an invitation to tinker with ongoing medication. Anyone taking an acid blocker, a laxative, an antibiotic or an immunosuppressive medication does not stop it on their own, does not reduce it and does not switch it. Every change to it belongs in a medical conversation and with medical supervision, because an abrupt stop can cause complaints of its own.

After that comes what actually feeds the microbiome, meaning plant variety and fibre. Why the often quoted number 30 is more marketing than physiology is described in Fibre myths. The overall concept with phases and diagnostics is in Gut reset, the whole gut approach.

And now you know why my answer to the question about a gut cleanse for the complexion almost always begins with a question back.

Why gut cleanse programmes for the skin often promise more than the data hold

I write this section reluctantly, because it turns against something that is actually close to me. I work a lot with the gut. I consider it underrated. And still this belongs here, because otherwise nobody says it.

When a gut cleanse promises you a better complexion, four things stand between that promise and the data.

First: the studies are small

Look at the sample sizes that appear in this article. Eleven people in Chen. Twelve in Nam. Seventeen against fourteen in Lesiak. Twenty seven in Nelson. Two hundred and thirty one spread across three probiotic studies in acne.

These are numbers from which research arises. They are not numbers from which treatment recommendations arise. For comparison: the Danish register cohort covered 49,475 people with rosacea, the psoriasis meta-analysis almost eight million participants. When the same question is answered once with eleven and once with fifty thousand people, that is not a tie.

Second: the findings contradict each other

The clearest case stands further up. Lactobacillus, Megasphaera and Acidaminococcus are reduced in Taiwan and increased in Korea, in the same condition.

When two investigations point in opposite directions for the same bacteria, then a statement such as your gut flora looks like this and that in this skin condition simply cannot hold. At least one of the two investigations must be wrong, and probably both are too small to hit anything.

Third: association is not cause

Rosacea and inflammatory bowel disease are associated in both directions. People with bowel disease more often have rosacea, and people with rosacea more often develop a bowel disease.

A chain from gut to skin cannot explain that. A shared immune predisposition that shows itself at both boundary surfaces explains it very well. The difference is practical: in the first case you treat the gut and the skin follows. In the second case you treat both organs where they are ill.

Fourth: skin conditions fluctuate on their own

This is the most underrated point, and it decides how much testimonials are worth.

Rosacea runs in flares. Acne changes with season, cycle, sleep and tension. Atopic dermatitis gets better in summer for many people and worse in winter. Anyone who starts a cleanse in a bad phase experiences an improvement with high probability. In many cases that improvement would have come without the cleanse as well.

Exactly for this reason the unblinded skin assessment in the Parodi study is a real problem, even though I consider this study the strongest in the field. And exactly for this reason the blinded assessment in Smith is so valuable.

What remains nonetheless

Now comes the part that matters to me, because otherwise this section sounds like a refusal. It is not one.

That is less than what is advertised. And clearly more than nothing.

On gut cleansing, enemas and colon hydrotherapy I deliberately say nothing here, neither good nor bad. That belongs in a text of its own, and that text is Gut cleansing, enema and colon hydrotherapy.

Common questions about gut and skin

These are the questions I am asked most often on this topic. The answers deliberately carry numbers, so that you can use them further.

Where gut and skin dock onto the rest of the body

Both organs hang on the same foundations. On what you eat, on your sleep, on your nervous system and on what your immune system currently considers dangerous.