IBS pain is not in your head: enzymes poke the gut's nerves

If you have irritable bowel syndrome, you know the appointment. The camera finds nothing. The blood work is clean. And somewhere near the end comes the sentence that does the damage: there is nothing wrong with your bowel. This is not a rare conversation. In the Rome Foundation Global Study, which questioned 73 076 adults across 33 countries, 40.3 % of the internet respondents met the criteria for at least one disorder of gut-brain interaction ⁴. That is a great many people being told their pain has no address.

A study published in the journal Gut in 2026, led by Lejla Ridžal and Anita Annaházi with colleagues at the Technical University of Munich, went looking for the address ¹.

What they actually did

They collected stool from patients recruited through centres in three countries: 21 with the diarrhoea-predominant subtype, 9 with the constipation-predominant subtype, and 18 healthy volunteers as controls ¹. The stool was spun down and only the liquid kept — the supernatant, the soup the gut wall is actually bathed in.

Then they dripped that soup onto living nerve networks. Not brain tissue: the submucous plexus of the distal colon, part of the enteric nervous system — the mesh of neurons woven into the gut wall that decides, largely without asking the brain, when the gut squeezes, secretes and reports trouble. Using an imaging method that shows which neurons fire, they watched what each sample of fluid did.

Fluid from both IBS groups activated those neurons significantly more strongly than fluid from the healthy controls ¹. Something in the stool of IBS patients talks directly to the gut's own nervous system.

Two illnesses wearing one name

Here is the part that should change how the condition is discussed. When the researchers blocked serine and cysteine proteases — enzymes whose job is cutting other proteins — the effect of the diarrhoea-subtype fluid went away, and it depended on PAR-1, a receptor that sits on the cell surface and is switched on by being snipped. In the constipation-subtype samples, none of that applied; those mediators worked through some other route ¹.

The protein census matched the split. 47 proteins differed between diarrhoea-subtype and healthy stool, several of them immunoglobulin components — antibodies, the residue of a quiet immune reaction rather than a raging one ¹. A combination of amylases, trypsin-2 and one immunoglobulin protein told diarrhoea-subtype samples from healthy ones well enough that the authors describe it as high diagnostic performance ¹.

This did not come from nowhere. In 2007, in The Journal of Clinical Investigation, colonic biopsies from IBS patients were shown to release more proteolytic activity than controls, to sensitise mouse sensory neurons in culture, and to do it through the related receptor PAR-2 ². In 2017, again in Gut, one culprit enzyme was pinned down as trypsin-3, released by the intestinal lining itself and able to signal to human submucosal neurons as well as mouse sensory ones — again through PAR-2 ³. What the 2026 work adds is the whole stool fluid rather than a single purified enzyme, a different receptor (PAR-1), and the split between the subtypes.

Where the enzymes come from is not settled. Stool fluid holds human enzymes and the working output of an enormous resident population — a colon community dominated by two great bacterial groups, one of them Bacteroidota ⁵. This study measured the enzymes. It did not trace their makers.

Now the caveats, out loud

The neurons were guinea pig, not human. That is the single biggest limit, and no amount of enthusiasm removes it. Human gut nerves may answer differently.

The neurons were also in a dish, exposed directly to stool fluid. A neuron lighting up under a microscope is not a person doubled over at work. The step from one to the other is exactly the step this study did not take.

The constipation group was 9 people ¹. Nine. Any statement about that subtype is a first sighting, not a finding to build on.

No treatment was tested. No patient received a protease inhibitor or a PAR-1 blocker. And the marker panel — amylases, trypsin-2, an immunoglobulin protein — is reported as high diagnostic performance across 21 diarrhoea-subtype samples against 18 controls. A combination that separates thirty-nine stool samples is a promising candidate, not a test you can ask your doctor for.

What honestly changes today is the conversation, not the prescription. There is now a measurable, physical thing in IBS stool that makes gut nerves fire, it differs between the two subtypes, and in a dish it can be blocked. "Nothing wrong with your bowel" was never a finding. It was the limit of what we had been able to measure.