Chronic diarrhoea: the causes that are missed most often
From about four weeks on, the question changes. Which pathogen turns into which mechanism. Three causes are well documented, easy to test for, and still rarely on the first list.
All articles in the gut cluster
Chronic diarrhoea is rarely a riddle. Most of the time it is a question of order. The three causes that get missed most often are not hard to find. In the first round, nobody went looking for them.
By now you know every toilet on your way to work. Not roughly. Exactly.
You have started skipping breakfast, because afterwards it always begins. You plan train journeys by which carriage has a toilet. You keep a spare pair of trousers in your backpack and tell nobody why.
You have been to the doctor. Blood count fine, stool sample negative for pathogens, colonoscopy normal. At the end came a word that explains nothing: irritable bowel.
Many people know this pattern. And many stop right here, because they believe everything has already been investigated.
Usually it has not. Not because anyone was careless, but because the first round of the work-up follows different questions than the second. This article is about the second round.
These signs belong in medical assessment, and promptly
- Blood in the stool, visible or as a coating
- Black, shiny, tarry stool
- Unintentional weight loss
- Fever
- Diarrhoea at night that pulls you out of sleep
- Repeated vomiting
- Difficulty swallowing, or the feeling that food gets stuck
- A new, persistent change in bowel habit from about 45 to 50 years of age
- Anaemia, meaning a low haemoglobin, or a low ferritin as a sign of iron deficiency
- Palpitations, confusion, barely passing urine or severe dizziness with ongoing diarrhoea
- Bowel cancer or inflammatory bowel disease in the family
If one of these red flags applies to you, it belongs in medical assessment and not in self treatment. With palpitations or signs of dehydration that means the same day, and if in doubt through the emergency number 112. Nothing in this article is a reason to postpone a recommended colonoscopy, gastroscopy or laboratory work-up, or to replace it with something else. Everything written here comes afterwards or alongside, never instead.
This article is written for adults. In children and adolescents different limits and different dangers apply, above all faster dehydration. Diarrhoea in childhood therefore belongs in paediatric hands sooner rather than later.
What is waiting for you here
- Why the four week line changes the work-up
- Bile acids that reach the large intestine, with two meta-analyses
- The SeHCAT test and the honest counterposition to it
- What a low pancreatic elastase means and what it does not
- Why microscopic colitis stays invisible without tissue samples
- Coeliac disease, lactose, fructose, sorbitol and the order trap
- The medications that appear on no list
- What calprotectin can do and what it explicitly does not cover
- An ordered work-up in six stages
- What the body loses over months of diarrhoea
When does diarrhoea become chronic, and why does that change everything
Almost everyone who comes to me describes the severity first. How often per day. How suddenly. How watery.
For the work-up, that is not the decisive number. The decisive one is the duration.
The common working definition sits in the British technology assessment of the SeHCAT test: more than three loose or liquid bowel movements a day and or a stool weight above 200 grams per day. It becomes chronic once it has lasted longer than four weeks.
These four weeks are not bureaucracy. They are a switch point.
Under four weeks the most likely explanation is an infection. The process ends on its own, and the sensible work-up reads: look for a pathogen if the course is severe, otherwise replace fluid.
Beyond four weeks the probabilities tip over. Most pathogens are gone by then, and one more standard pathogen test often brings nothing new. A mechanism that runs this long is more common. And mechanisms are questioned differently than pathogens.
Two exceptions belong here explicitly, though. After long distance travel or with an immune deficiency, parasites such as Giardia, amoebae or Cryptosporidium can stay for months. And after a course of antibiotics, Clostridioides difficile comes into question. Travel history and targeted pathogen testing therefore belong alongside the stages that follow, not behind them.
From here on the question is no longer: what did I catch?
The question is: why does water stay in the gut, or why is it being drawn in there? Every sensible investigation from week five onwards is an answer to exactly that question. Once you have understood this, you also understand why yet another standard pathogen test often brings nothing new, once travel history and immune status have been clarified.
The mechanisms, and what sets them apart in daily life
There are essentially four routes by which stool becomes liquid, plus a fifth that strictly speaking belongs to them but is distinct enough in daily life. They overlap, but they feel different, and that is diagnostically useful.
How water ends up in the gut
- Osmotic. Something dissolved stays in the gut and pulls water after it. Classic with lactose, fructose, sorbitol, magnesium and some laxatives. The recognisable sign: if you fast or leave the substance out, it stops.
- Secretory. The bowel wall actively releases fluid. Bile acids that reach the large intestine belong here, as do bacterial toxins and, rare but important, hormone producing tumours such as a VIPoma, a gastrinoma or a carcinoid. The recognisable sign: it does not stop when you fast and it can carry on at night. Exactly this combination belongs in further medical assessment.
- Inflammatory. The mucosa is irritated or damaged, fluid and protein are lost. Inflammatory bowel disease and microscopic colitis belong here. The recognisable sign: often blood or mucus, often symptoms at night, often a raised calprotectin.
- Motility. The contents move through faster than water can be taken back up. Typical with an overactive thyroid, after surgery and in a proportion of functional complaints.
- Fat malabsorption. Strictly speaking a subgroup, but distinct enough in everyday life: fat that is not broken down makes bulky, shiny, foul smelling stool that is hard to flush away. The pancreas belongs here.
This division is a thinking grid, not a diagnosis. But it explains why one single sentence from your description, for example whether the diarrhoea wakes you at night, shifts the probabilities in the examining person's mind more than any further pathogen test.
One cause is rarely thought of here. A previously unknown HIV infection or another immune deficiency can show itself first as months of diarrhoea. An HIV test therefore belongs in the medical conversation when chronic diarrhoea remains unexplained.
That is why I always ask the same thing first in the consulting room. Does it stop if you eat nothing for a day? Does it wake you at night? Is the stool watery or greasy and shiny? And since exactly when. These are not small talk questions. Every single one moves one group of causes forward and another one back.
Functional diarrhoea is a diagnosis of its own, not the end of the search
One more word on the label that many people are given. The Rome IV criteria from 2016 list five categories of functional bowel disorders, and functional diarrhoea is one of them. It sits next to irritable bowel syndrome, not inside it. The difference is pain: with irritable bowel it belongs to the picture, with functional diarrhoea it is absent. That is more than hair splitting, because the two categories lead to different next steps.
The Rome IV committee around Fermín Mearin revised the criteria for functional bowel disorders in 2016.
It set out five categories, among them irritable bowel syndrome and functional diarrhoea as separate entities, each with its own definition, epidemiology and clinical evaluation.
For you that means: irritable bowel and functional diarrhoea are not the same thing. And both are diagnoses that presuppose a work-up. They are no substitute for one.
Mearin F, Lacy BE, Chang L et al. Gastroenterology. 2016. PMID: 27144627 · DOI: 10.1053/j.gastro.2016.02.031 [Guideline]The German S3 guideline on irritable bowel syndrome from DGVS and DGNM, published in 2021, also understands irritable bowel as a diagnosis that presupposes an ordered work-up, and not as a label of last resort.
One special case is often overlooked here: the alternation between constipation and diarrhoea. When hard stool stays put and liquid stool runs past it, that looks like diarrhoea and is the opposite. More on this in the article on constipation and its causes. What matters here: a new, persistent change in bowel habit from about 45 to 50 years of age belongs in a work-up regardless of the direction.
That also makes the ranking clear: for the work-up, duration weighs more than severity. It decides which question gets asked in the first place.
The first missed cause: bile acids that reach the large intestine
Imagine a dishwasher that retrieves its detergent after every cycle and reuses it. About ninety five percent comes back, only a small remainder is lost and has to be made again.
That is exactly how your body works with bile acids. The liver makes them, the gallbladder releases them when you eat, they dissolve fat in the small intestine, and at the end of the small intestine they are almost completely taken back up.
When this retrieval jams, detergent ends up in the large intestine. And there it can do what detergent does: draw water, irritate the mucosa and speed up transit. That can show itself as watery diarrhoea, often urgent, often shortly after eating, often yellowish and sometimes at night.
Slattery and colleagues searched MEDLINE and EMBASE in 2015 for studies in which adults with diarrhoea predominant irritable bowel syndrome had been examined for bile acid loss using the SeHCAT test.
Six studies with 908 people met the criteria. The pooled proportion with bile acid malabsorption, defined as SeHCAT retention below 10 percent, was 28.1 percent, confidence interval 22.6 to 34 percent, with considerable heterogeneity.
For you that means: if your diarrhoea has been filed under irritable bowel, there is a real probability that a bile acid problem sits behind it. This is not an exotic exception.
Slattery SA, Niaz O, Aziz Q, Ford AC, Farmer AD. Aliment Pharmacol Ther. 2015;42(1):3-11. PMID: 25913530 · DOI: 10.1111/apt.13227 [Meta-analysis, k=6, n=908]A working group around Tianxu Liu evaluated ten case control studies in 2023, covering 1,034 affected people and 232 healthy comparison subjects.
After SeHCAT the pooled rate was 32 percent, confidence interval 24 to 40 percent. The blood value C4 was raised, the gut hormone FGF19 lowered. The authors explicitly noted that the reference ranges for these blood markers differ between studies.
For you that means: two independent evaluations land in the same range. There are blood markers pointing the same way, but they are not standardised yet.
Liu T, Ma M, Li K, Tan W, Yu H, Wang L. J Clin Gastroenterol. 2023;57(5):451-458. PMID: 36867517 · DOI: 10.1097/MCG.0000000000001841 [Meta-analysis, k=10, n=1,034]Why this happens even without gallbladder surgery
The most common objection in the consulting room goes: but my gallbladder is still there. That is not a counterargument. Bile acid diarrhoea after gallbladder removal is a known situation of its own and is described in the article on bile, the gallbladder and TUDCA. Here we are dealing with the other form, the one that arises without any surgery at all.
The mechanism behind it was only understood over the past two decades.
The missing stop signal from the small intestine
- Bile acids reach the end of the small intestine and are taken up there by the mucosal cells.
- Those cells answer by releasing the hormone FGF19 into the blood. It is the message back to the liver: there is enough.
- The liver then throttles its new production of bile acids. A classic feedback loop.
- If less FGF19 arrives, the throttling does not happen. The liver keeps producing, although enough is already in circulation.
- The small intestine can no longer retrieve this surplus completely. The excess reaches the large intestine and draws water there.
This is why the simple idea of a broken retrieval mechanism falls short. In a proportion of those affected it is not the uptake that is disturbed, it is the brake. Among others, Julian Walters described this, and he explicitly compared the recognition gap around this diagnosis with the historical recognition gap around coeliac disease.
Julian Walters at Imperial College London summarised in 2010 what was known about primary bile acid diarrhoea, meaning the form not caused by surgery.
Around 30 percent of those who would otherwise receive the diagnosis of irritable bowel with diarrhoea or functional diarrhoea show an abnormal SeHCAT retention. As the central mechanism he named the disturbed FGF19 feedback with increased new synthesis in the liver.
For you that means: this diagnosis looks like irritable bowel and feels like irritable bowel. It still has a name of its own.
Walters JRF. Expert Rev Gastroenterol Hepatol. 2010;4(5):561-567. PMID: 20932141 · DOI: 10.1586/egh.10.54 [Mechanism Review]How this is found: SeHCAT, blood markers, treatment trial
The reference test is called SeHCAT. You swallow a capsule containing a radioactively labelled substance that resembles a bile acid. Seven days later it is measured how much of it is still in your body. If little is left, you are losing a lot.
Because this is a nuclear medicine investigation involving radiation, two points belong with it. In pregnancy it counts as a contraindication. During breastfeeding, feeding has to be interrupted as instructed by the nuclear medicine department carrying out the test. Whether the radiation exposure is justified in your case is decided by that department under radiation protection law, not by this article.
How much is too little depends on the chosen threshold. Common cut-offs are 15, 10 and 5 percent remaining activity. It is at these thresholds that the debate flares up.
Wedlake and colleagues at the Royal Marsden Hospital found eighteen studies in 2009, together covering 1,223 people, in which people with irritable bowel diarrhoea had been examined by SeHCAT. The individual analyses rest on different subsets of those studies.
A severe loss, meaning under 5 percent remaining activity, was found in 10 percent, under 10 percent in 32 percent, under 15 percent in 26 percent. For the response to a bile acid binder a dose response relationship emerged: 96 percent at a remaining activity below 5 percent, 80 percent below 10 percent, 70 percent below 15 percent. These response rates come from largely open observations without a placebo group and are more likely to overstate the effect.
For you that means: the test says not only whether, but also how strongly. And the strength says something about how reliably a treatment might take hold.
Wedlake L, A'Hern R, Russell D, Thomas K, Walters JRF, Andreyev HJN. Aliment Pharmacol Ther. 2009;30(7):707-717. PMID: 19570102 · DOI: 10.1111/j.1365-2036.2009.04081.x [Systematic Review]One could stop here and demand the test for everyone. That is exactly what the professional societies do not do, and they have reasons. Those belong here.
Why the professional societies hold back on SeHCAT
The British technology assessment by Rob Riemsma and colleagues from 2013 was commissioned work for the NICE diagnostics programme, with a systematic review and a health economics model.
The result is uncomfortable for both sides. Only three studies examined the connection between SeHCAT result and response to a bile acid binder at all, with small numbers and differing thresholds. There is no true reference standard for the diagnosis. Over a short time horizon and across various scenarios, the controlled treatment trial was the economically cheaper strategy. The authors' conclusion: considerable decision uncertainty, no strategy clearly superior.
What this means in practice: if your doctor would rather run a controlled treatment trial than the test, that is not convenience, it is a reasoned position with a paper behind it. And the other way round: if you want to raise the test, the prevalence figures give you an equally reasoned argument. Whether it is justified in the individual case is decided by the examining department, because this involves radiation.
Riemsma R, Al M, Corro Ramos I et al. Health Technol Assess. 2013;17(61):1-236. PMID: 24351663 · DOI: 10.3310/hta17610 [Systematic Review]In Germany the SeHCAT test is available at specialised nuclear medicine centres, not everywhere, and reimbursement is not settled across the board. Both belong on the table beforehand.
That leaves the question of what follows from a finding.
Wilcox, Turner and Green evaluated thirty publications in 2014, covering 1,241 adults in whom bile acid malabsorption had been diagnosed and treated.
Cholestyramine, a prescription only bile acid binder, was the best studied. In 801 evaluated cases an improvement was reported in 70 percent, range 63 to 100 percent. These figures come from open observations without a comparison group and without placebo, so they are more likely to overstate the effect.
The weak point was tolerability. Bile acid binders can be poorly tolerated, with constipation, bloating and nausea among the reported effects. They can reduce the uptake of fat soluble vitamins and the bioavailability of medication taken at the same time, and they cannot be used where the bile ducts are completely obstructed.
For you that means: with confirmed bile acid loss a binder may come into question. Whether, which one and with what timing gaps is set by your doctor alone. On dosing I deliberately say nothing here.
Wilcox C, Turner J, Green J. Aliment Pharmacol Ther. 2014;39(9):923-939. PMID: 24602022 · DOI: 10.1111/apt.12684 [Systematic Review]Bile acid binders are not a self treatment. They bind not only bile acids, but also thyroid hormone, blood thinners, fat soluble vitamins and many other active substances. Anyone taking them without medical supervision can render an existing treatment ineffective without noticing. Their use belongs in medical hands, with a clear diagnosis beforehand and timing gaps alongside.
One point often gets lost here: a normal colonoscopy does not touch this question at all. Bile acids cannot be seen through an endoscope.
The second missed cause: a pancreas that delivers too little
There is one description that makes me prick up my ears. The stool is not simply liquid but bulky, pale, shiny, hard to flush away. Alongside it a weight loss nobody can explain, and a feeling of fullness after fatty meals.
That is the picture of a fat digestion that cannot keep up. The pancreas delivers the enzymes that break down fat, protein and starch. When too little of that arrives, fat stays undigested in the gut. This is called exocrine pancreatic insufficiency.
The European guideline from 2024, carried by seven professional societies, defines it as a decline in pancreatic secretion below the level that allows normal nutrient digestion. In practice that means: less tooling arrives than the work requires.
John Leeds and colleagues in Sheffield examined three groups: 314 people with diarrhoea predominant irritable bowel syndrome, 105 with chronic diarrhoea of another kind and 95 comparison subjects without diarrhoea.
A faecal elastase below 100 micrograms per gram was found in 19 of the 314 people with irritable bowel, meaning 6.1 percent, but in nobody from the other two groups. After enzyme supplementation, stool frequency, consistency and abdominal pain improved measurably in the group with low elastase, and not in the comparison group.
For you that means: about one in sixteen people carrying the label of irritable bowel diarrhoea had a relevant pancreatic weakness here. Not many. But enough to make it worth asking.
Leeds JS, Hopper AD, Sidhu R et al. Clin Gastroenterol Hepatol. 2010;8(5):433-438. PMID: 19835990 · DOI: 10.1016/j.cgh.2009.09.032 [Cohort, n=514]The test for it is unspectacular: a single stool sample in which pancreatic elastase-1 is measured. No fasting, no preparation, no equipment.
But it is exactly this simplicity that creates a misunderstanding. A low value is not a verdict.
Daniel de la Iglesia and colleagues at Puerta de Hierro in Majadahonda evaluated thirteen studies in 2025, covering 888 people, in which faecal elastase had been tested against the reference standard of fat excretion.
At the 200 microgram per gram cut-off the pooled sensitivity was 0.94, but the specificity only 0.69. At the stricter cut-off of 100 the specificity rose to 0.82 while sensitivity fell to 0.88. The authors pointed explicitly to the limited predictive values.
Important for the interpretation: these figures come mostly from populations in which a weak pancreas is common. In the subgroup analysis sensitivity was 0.98 in cystic fibrosis and specificity 0.81 in chronic pancreatitis. The authors accordingly recommend the test for initial screening in people at increased risk. In unselected chronic diarrhoea it can therefore perform more weakly than the pooled figures suggest.
For you that means: the test picks up almost every genuine weakness, yet it also flags people who do not have one. A low value is a door opener, not the end of the investigation.
de la Iglesia D, Agudo-Castillo B, Galego-Fernández M, Rama-Fernández A, Domínguez-Muñoz JE. United European Gastroenterol J. 2025;13(8):1571-1582. PMID: 40569793 · DOI: 10.1002/ueg2.70061 [Meta-analysis, k=13, n=888]There is a further catch: elastase is measured in a stool sample. If the stool is watery, everything in it is diluted. A low value can then be a dilution effect and say nothing about the pancreas.
This is why the European guideline contains a sentence that is easily skimmed over: the diagnosis should rest on an overall assessment, from symptoms, nutritional status and a secretion test. Not on a laboratory value alone.
A laboratory value answers no question. It shifts a probability.
Whether the elastase value means anything depends on how likely the problem was before the test. With greasy shiny stool, weight loss and fullness after fatty food, a low value means a lot. With watery diarrhoea and none of these signs, the same value means considerably less. That is not a contradiction, that is statistics.
And one more thing belongs at this point. A weak pancreas is never the whole answer, it is always a question of why. Behind it can sit chronic pancreatitis, a state after pancreatic surgery, cystic fibrosis, long standing diabetes and, especially with new symptoms and weight loss, a mass in the pancreas.
A first time low elastase value together with weight loss therefore does not belong answered with enzymes. It belongs answered with a medical search for the cause, including imaging of the pancreas.
And why enzymes are still not the first step
At this point the wish often comes up to simply try digestive enzymes. Understandable, they are freely available and sound harmless.
I still take a step back. In my practice I ask the acid question before the enzyme question: is there enough stomach acid up top in the first place? The thought behind it: stomach acid could be the start signal for the digestion that follows, and if too little arrives up top, more than just enzymes could be missing further down. This is a clinical working rule without a strong study base, and I mark it explicitly as such. More on this in the article on low stomach acid. When enzymes are up for discussion at all is the topic of the article on digestive enzymes.
What remains is simple: enzymes taken on suspicion, without diagnostics beforehand, blur exactly the signal you need. If things get better afterwards, you do not know why. If they do not get better, you do not know either.
One thing above all can be held on to here: a single stool value in this field is neither an acquittal nor a verdict.
The third missed cause: an inflammation you cannot see
This is the section this article was written for.
Many people arrive with a report that says: macroscopically normal. They read that as the all clear. But it only says: there was nothing to see.
A normal colonoscopy rules out what can be seen. Microscopic colitis cannot be seen. It is found exclusively in tissue samples. If none were taken, the question has not been answered, it was never asked in the first place.
The European guideline from 2021 defines the condition exactly this way: chronic inflammatory bowel disease with a normal or near normal endoscopic appearance, chronic watery, non bloody diarrhoea and characteristic changes in the tissue.
There are three forms. In collagenous colitis a thickened band of collagen lies beneath the mucosal surface. In lymphocytic colitis immune cells migrate in between the mucosal cells. The incomplete form shows the changes in a milder degree. All three look normal through the endoscope, and all three can run for years under the label of irritable bowel.
Stephan Miehlke and a European author team from gastroenterology, pathology and basic research developed the European guideline on microscopic colitis in 2021 following AGREE II methodology.
It states: the endoscopic appearance is normal or near normal, the diarrhoea chronic, watery and non bloody, and the diagnosis rests on characteristic histological changes. Among the treatment options named are oral budesonide and bile acid binders.
For you that means: if your diarrhoea is watery, non bloody and chronic and the colonoscopy showed nothing, the question about stepwise biopsies is justified. It is a treatable diagnosis, but only if somebody looks for it. The treatment options named there are prescription only and belong in medical hands. Budesonide is a prescription only corticosteroid preparation that acts mainly in the gut and reaches the rest of the body less than classic cortisone does. That does not make it free of side effects. Reported effects include increased susceptibility to infection, effects on bone density and the eyes with repeated or long courses, and an influence on adrenal function. After stopping, symptoms also frequently return. Dose, duration and follow-up checks are set by your doctor. Neither starting nor stopping it is something you decide alone.
Miehlke S, Guagnozzi D, Zabana Y et al. United European Gastroenterol J. 2021;9(1):13-37. PMID: 33619914 · DOI: 10.1177/2050640620951905 [Guideline]Who is affected in particular
The condition has a distinct profile. That makes it well predictable in the consulting room, once you know it.
Collagenous colitis
3.05 : 1Ratio of women to men. Incidence 4.14 per 100,000 person years. Median age at diagnosis 64.9 years.
Lymphocytic colitis
1.92 : 1Ratio of women to men. Incidence 4.85 per 100,000 person years. Median age at diagnosis 62.2 years.
Jinlu Tong and colleagues evaluated twenty five epidemiological studies on microscopic colitis in 2015.
Incidence was 4.14 per 100,000 person years for the collagenous and 4.85 for the lymphocytic form. Women were affected three times as often in the collagenous form, the median age at diagnosis was 64.9 and 62.2 years respectively, and incidence rose with age. For proton pump inhibitors they calculated an odds ratio of 2.68, for SSRI antidepressants 2.41.
For you that means: if you are a woman past sixty, have had watery diarrhoea for weeks and the colonoscopy was normal, the question about tissue samples belongs on the table.
Tong J, Zheng Q, Zhang C, Lo R, Shen J, Ran Z. Am J Gastroenterol. 2015;110(2):265-276. PMID: 25623658 · DOI: 10.1038/ajg.2014.431 [Meta-analysis, k=25]That brings us to medication. And here it gets methodologically interesting, because the obvious answer is probably too simple.
Why the odds ratios for acid blockers and others can disappear
Zahid Tarar and colleagues evaluated thirteen studies in 2022 with 304,482 people, among them 34,194 with microscopic colitis. The decisive part was their subgroup analysis: once against randomly selected controls from the population, once against people who also had diarrhoea.
Against random controls the numbers looked dramatic: proton pump inhibitors with an odds ratio of 4.55, SSRI 3.23, anti-inflammatory painkillers 3.27, statins 2.23. Against diarrhoea controls the raised risk was gone, for proton pump inhibitors even reversed at 0.68. The explanation is close at hand: people with abdominal complaints take exactly these medications more often.
What this means in practice: suspicion of a medication is justified and belongs to be reviewed. But it is no proof, and it does not justify stopping anything on your own.
Tarar ZI, Farooq U, Gandhi M et al. Diseases. 2022;11(1):6. PMID: 36648871 · DOI: 10.3390/diseases11010006 [Meta-analysis, k=13, n=304,482]One signal, however, survives this scrutiny somewhat better than the others.
Anna and Olli Nevalainen systematically searched MEDLINE and Scopus in 2023 for observational studies on acid suppressing medication and immune mediated diseases. Microscopic colitis was the most frequently studied condition.
For proton pump inhibitors overall the pooled odds ratio was 3.28, but the confidence interval ran from 0.98 to 11.0 and the certainty of evidence by GRADE was low. For lansoprazole and collagenous colitis the odds ratio was 14.5, confidence interval 9.37 to 22.3, with low heterogeneity and moderate certainty of evidence by GRADE. That figure rests on only three observational studies covering 1,725 exposed people, however, not on all 25 articles in the review.
For you that means: for the acid blocker group as a whole the signal is uncertain, for one particular substance with one particular form it is clearer. Both show an association, not a cause. A good reason to have the indication reviewed medically. No reason to stop the acid blocker on your own.
Nevalainen A, Nevalainen OPO. Int J Risk Saf Med. 2023;34(3):207-225. PMID: 36442213 · DOI: 10.3233/JRS-220012 [Meta-analysis, k=25]Why biopsies are still not taken in everyone
If tissue samples are the sole route to this diagnosis, why are they not always taken? Here too there is an honest voice on the other side.
The yield of untargeted random biopsies is low
A two centre audit from Western Australia examined all colonoscopies between 2009 and 2018 in which random biopsies had been taken from macroscopically normal mucosa: 872 examinations, 48.7 percent of them because of diarrhoea.
Only 1.5 percent of all biopsies were positive for microscopic colitis, 3.1 percent in the diarrhoea group. The cost per diagnosis made was 10,862 Australian dollars. The authors' conclusion: biopsies where suspicion of an organic cause is high, not as a reflex in everyone.
What this means in practice: a good history beats the reflex biopsy. If the profile fits, meaning chronic watery diarrhoea, older age, female sex, matching medication, the hit rate rises considerably. This is why it is worth raising these points actively in the conversation.
Seenarain V, Idrees M, Mogridge J, Sinha A, Thompson A. ANZ J Surg. 2020;90(12):E163-E167. PMID: 32856361 · DOI: 10.1111/ans.16248 [Cohort, n=872]A normal colonoscopy does not mean nothing is there. It means nothing was visible.
That difference decides whether you keep looking or stop. And it is the reason why one question in the follow-up conversation can be worth more than another investigation: were tissue samples taken from several segments, including where the mucosa looked normal?
That explains part of why this diagnosis so often takes years. It is not hiding. It is simply being looked for in a place where it is not visible.
The other causes you must not forget
I have covered three causes in depth, because they are barely explained anywhere online in German. The following ones appear almost everywhere. They still belong here, because their order and their pitfalls are rarely explained alongside.
Coeliac disease: test first, then leave things out
Coeliac disease is a permanent immune mediated reaction to gluten from wheat, barley and rye. It is not an intolerance in the colloquial sense, but a disease with effects reaching far beyond the gut.
Prashant Singh and an international team evaluated 96 studies that had examined the frequency of coeliac disease in the population between 1991 and 2016.
The pooled seroprevalence was 1.4 percent among 275,818 people, the biopsy confirmed prevalence 0.7 percent among 138,792 people. For Europe and Oceania they calculated 0.8 percent, for women 0.6 against 0.4 percent in men.
For you that means: common enough that it is looked for as standard in chronic diarrhoea. And rare enough that a single abnormal antibody is not yet a diagnosis.
Singh P, Arora A, Strand TA et al. Clin Gastroenterol Hepatol. 2018;16(6):823-836.e2. PMID: 29551598 · DOI: 10.1016/j.cgh.2017.06.037 [Meta-analysis, k=96, n=275,818]Do not eat gluten free before the work-up has been done
A gluten free diet before coeliac diagnostics can make the diagnosis impossible. Both the antibodies in the blood and the changes in the small bowel lining recede on a gluten free diet. Both investigations require that you are still eating gluten containing food regularly at the time of the test.
The American guideline from 2023 therefore puts serology on a gluten containing diet at the start and places the small bowel biopsy for confirmation alongside it for most adults. A biopsy free strategy is discussed only for selected children.
Anyone who leaves gluten out first and tests afterwards ends up without a diagnosis. Not healthy, just without an answer. And without a diagnosis there is no follow-up, no family screening and no framing of the long term consequences. The order reads: test first, then leave things out. How the diagnostics run in detail is described in the article on recognising coeliac disease.
To be distinguished from this is gluten sensitivity without coeliac disease, where the antibodies stay negative and the tissue sample normal, while the symptoms are real. That is a topic of its own and is covered in the article on gluten and gliadin without coeliac disease.
Lactose, fructose, sorbitol: and the order trap
Sugars that are not taken up in the small intestine pull water after them and are fermented by bacteria in the large intestine. That makes gas and liquid stool.
This is tested with breath tests. The North American consensus of 2017 set doses and thresholds: lactulose 10 grams, glucose 75 grams, fructose 25 grams, lactose 25 grams, positive from a hydrogen rise of at least 20 ppm above baseline.
And then there is a sentence in it that is easily lost in everyday practice: bacterial overgrowth of the small intestine should be excluded before a test for carbohydrate malabsorption, in order to avoid false positive results.
A positive fructose breath test does not prove that you cannot tolerate fructose.
It only shows that somewhere in your gut hydrogen was produced from fructose. If the bacteria sit too far up, meaning in the small intestine rather than the large one, the same rise appears without fructose uptake being disturbed at all. This is why the question about bacterial overgrowth of the small intestine comes before the sugar test, not after it. And an abnormal test without accompanying symptoms during the measurement is no diagnosis anyway.
For sorbitol the situation is thinner still. The North American consensus names agreed test doses for lactulose, glucose, fructose and lactose, but not for sorbitol. That is the statement in itself: for sorbitol a standardised test dose is missing. It remains an observation, not a diagnosis with a threshold. And sorbitol sits in more products than most people suspect, from sugar free chewing gum to dried fruit.
Anyone moving into an elimination protocol after an abnormal test should know that leaving things out is a time limited diagnostic measure and not a permanent diet. How that runs in a structured way is described in the article on the FODMAP diet.
After the infection: when it simply does not stop
A sentence I hear often: ever since that one stomach bug on holiday it has never fully gone away.
Fabiane Klem and colleagues at the Mayo Clinic evaluated 45 cohort studies with 21,421 people after infectious enteritis.
Twelve months after the infection 10.1 percent had irritable bowel syndrome, and the risk was 4.2 fold higher in the first year. After bacterial infections the proportion was 13.8 percent, after protozoal infections 41.9 percent, the latter based on few studies with considerable heterogeneity. A raised risk also showed up in women and where anxiety or depression accompanied the course.
For you that means: the link to the infection is not imagination. It has a name and a data base, but it does not replace looking into the other causes.
Klem F, Wadhwa A, Prokop LJ et al. Gastroenterology. 2017;152(5):1042-1054.e1. PMID: 28069350 · DOI: 10.1053/j.gastro.2016.12.039 [Meta-analysis, k=45, n=21,421]A Danish analysis from 2019 arrived at similar orders of magnitude: 12 percent after Campylobacter, 12 percent after Salmonella, 11 percent after Shigella. For viral and parasitic pathogens only few studies exist, and the authors say so explicitly.
What this means in practice: the pathogen itself matters less than the fact that the onset coincided in time with an infection. That belongs in your history. More on the post infectious course is in the article on irritable bowel and its causes, and if parasites were involved, in the article on Blastocystis and Giardia.
Thyroid: a blood test that often gets forgotten
An overactive thyroid can speed up transit from mouth to large intestine. What moves through faster can be thickened less well. That can already explain the diarrhoea, even when nothing is found in the gut itself.
Ellen Ebert described the gastrointestinal effects of thyroid disease in 2010. Graves disease accounts for 60 to 80 percent of cases of overactivity, mouth to caecum transit is accelerated, and fatty stool can arise through hyperphagia and adrenergic stimulation. In underactivity, by contrast, slowed motility with constipation predominates.
Keep this in mind if palpitations, weight loss with a good appetite, restlessness, sweating or tremor have joined the picture. The TSH value in the blood is a simple, cheap test and belongs early in the work-up. More on this in the article on an overactive thyroid. Important if you already take a thyroid preparation: existing thyroid medication is never changed, reduced or stopped on your own. Every adjustment belongs discussed and supervised medically.
Inflammatory bowel disease
Crohn's disease and ulcerative colitis are the best known organic causes of chronic diarrhoea. They come more often with blood, mucus, fever, night time symptoms and weight loss, and as a rule they show up in the colonoscopy and in calprotectin. Their treatment is a large topic of its own and is covered in the article on Crohn's disease and colitis. Only one thing counts here: they belong to the diagnoses that are actively looked for in the first round, and red flags are no reason to wait.
Across all these causes the same pattern shows up: the order of the tests weighs just as heavily as their selection.
The medications that appear on no list
If I had to guess which section of this article changes something most often, it would be this one.
The European consensus on malabsorption from 2025, carried by ten professional societies, puts it unusually plainly: the medical history and the medication history are equally essential for recognising risk. Many cause lists summarise this under adverse drug effects. I find it more helpful to name the substances that actually come up in the consulting room.
Alberto Rubio-Tapia and colleagues at the Mayo Clinic in Rochester collected twenty two people between 2008 and 2011 with unexplained chronic diarrhoea and bowel damage while taking the blood pressure drug olmesartan. Coeliac disease had been excluded in every case.
Thirteen were women, the median age was 69.5 years, the median weight loss 18 kilograms, range 2.5 to 57 kilograms. Fourteen had to be admitted to hospital. Tissue samples showed villous atrophy in fifteen. Transglutaminase antibodies were negative, and a gluten free diet brought no improvement. Important for the interpretation: only people who improved after medically supervised withdrawal were included in this case series at all. The improvement was therefore an inclusion criterion and not a success rate. Within this selected group the mean weight gain was 12.2 kilograms. How often olmesartan produces this picture in the first place is not something the paper says.
For you that means: a medication can produce a picture that looks like coeliac disease and does not respond to a gluten free diet. The decisive step here was not the next diet, it was going through the medication list.
Rubio-Tapia A, Herman ML, Ludvigsson JF et al. Mayo Clin Proc. 2012;87(8):732-738. PMID: 22728033 · DOI: 10.1016/j.mayocp.2012.06.003 [Case Series, n=22]Two qualifications, so this does not turn into panic. First, these are twenty two cases, and the vast majority of people taking olmesartan experience none of it. Second, the documented evidence relates to this one substance. Whether other sartans can do the same is not covered by this case series and remains an open question.
With metformin the situation is different in kind, but at least as important in practice. Fabrice Bonnet and André Scheen summarised in 2017 what is known about gastrointestinal intolerance. Diarrhoea and nausea are the most common side effects, a considerable proportion of those treated cannot tolerate the preparation at an adequate dose, and genotype variants, comorbidities and co-medication all play into it. As strategies they name a slower titration of the immediate release form and a switch to an extended release form.
Important here: these are options for a conversation, not an instruction. A diabetes medication is not left out, switched or reduced on your own.
The medication list that belongs at the appointment
- Olmesartan and other blood pressure drugs
- For olmesartan a severe, coeliac like bowel injury is documented. If you take a sartan and have had diarrhoea for weeks, that belongs raised actively. Stopping only under medical supervision.
- Metformin
- Diarrhoea and nausea are the most common side effects, often tied in time to the start or a dose increase. Extended release form and titration are discussable routes. Never change on your own.
- Proton pump inhibitors, meaning acid blockers
- The data are split, but clear for lansoprazole and collagenous colitis. What makes sense is a medical review of whether the indication still stands. More on this in the article on heartburn and acid blockers.
- SSRI antidepressants
- Associated with microscopic colitis in case control studies, with the same methodological limitation as the acid blockers. Psychiatric medication is never stopped on your own.
- Anti-inflammatory painkillers and statins
- Both turn up in the same analyses. For painkillers and the collagenous form a moderate signal persisted even at low heterogeneity.
- Antibiotics
- Both during the course and weeks afterwards. If diarrhoea appears newly after antibiotic treatment or is particularly severe, that belongs in medical assessment. What may be useful for rebuilding afterwards is described in the article on the gut after antibiotics.
- Magnesium and magnesium containing preparations
- The osmotic effect is undisputed and dose dependent, but I did not find a reliable frequency figure for adults in this research. I therefore list it as a clinical observation, not as a documented number. It is still worth raising the dose and the salt form.
- Laxatives, including herbal ones
- Regular use, including as a tea, belongs on the list. So do sugar substitutes in sugar free products.
None of this is stopped, reduced or switched on your own. The purpose of this list is a better conversation, not a self experiment. Write everything down, including supplements, including preparations from the supermarket, including what you only take occasionally. And bring the packaging along if you are unsure.
This is why a sheet of paper with your medication list can sometimes bring more than the next test.
What the stool test can do, what it cannot, and in what order to proceed
If one value in this field gets over-interpreted, it is calprotectin.
Calprotectin is a protein from white blood cells. When these cells migrate into the bowel lining because something is inflamed there, part of it ends up in the stool. So the test does not measure the gut, it measures the presence of inflammatory cells in it. Its strength lies in the all clear.
Yoon An and colleagues evaluated eighteen studies from primary care and outpatient settings in 2019, in which faecal calprotectin was used to distinguish organic from functional complaints.
For organic against functional, sensitivity and specificity were each 81 percent. The decisive part is the arithmetic behind it: at an assumed prevalence of one percent the positive predictive value was only 4.2 percent, while the negative predictive value was 100 percent.
For you that means: a low calprotectin is a reliable all clear. A mildly raised one, where the disease is rare, is usually not a sign of disease. It can also come from painkillers, an infection that has passed, or a small trace of blood.
An YK, Prince D, Gardiner F et al. Med J Aust. 2019;211(10):461-467. PMID: 31680263 · DOI: 10.5694/mja2.50384 [Meta-analysis, k=18]An American analysis from 2015 reaches the same core statement from the other side: with a calprotectin up to 40 micrograms per gram or a CRP up to 0.5, the probability of inflammatory bowel disease is at most one percent. Erythrocyte sedimentation rate and lactoferrin added little on their own.
What calprotectin explicitly does not cover
- Microscopic colitis. It can be present with a normal calprotectin. The test is no substitute for a tissue sample.
- Coeliac disease. Looked for through antibodies in the blood and the small bowel biopsy, not through the stool.
- Bile acid loss. Runs without inflammation. Calprotectin is typically normal alongside it.
- Exocrine pancreatic insufficiency. A different stool test, a different question, a different statement.
- Carbohydrate malabsorption. Investigated through breath tests.
And what about microbiome analysis?
The question comes up almost every time, so here it is briefly and clearly. None of the nine guidelines and consensus documents I evaluated for this article lists microbiome sequencing as a diagnostic tool in chronic diarrhoea. Calprotectin, by contrast, is backed by clean test performance data.
That does not mean microbiome data are uninteresting. It means that in this particular diagnostic question they so far decide nothing that would not also be decided without them. Where their value lies and where their limits sit is covered at length in the article on stool testing, PCR and dysbiosis diagnostics.
The order that wastes no round
The British guideline on the investigation of chronic diarrhoea from 2018 was developed by a group of thirteen, which included two patient representatives. Its basic principle is simple: the work-up follows pre-test probability and red flags, instead of running every test in everybody. The American guideline from 2019 on laboratory evaluation in functional diarrhoea and irritable bowel with diarrhoea follows the same idea.
This is what it looks like in ordered form.
Six stages, in this order
Check the red flags
Blood, tarry stool, weight loss, fever, night time symptoms, vomiting, difficulty swallowing, a new persistent change in bowel habit from about 45 to 50 years of age, anaemia, family history.
If one applies: medical assessment straight away, all further stages run alongside, not insteadGo through the medication list
Completely, including supplements, laxatives, sugar substitutes and everything taken only occasionally. Costs nothing and occasionally answers the whole question.
Free of chargeOften skippedBasic laboratory panel
Blood count, inflammatory markers, ferritin, electrolytes and TSH. The thyroid belongs in here, because an overactive thyroid can trigger diarrhoea without any bowel finding at all.
Blood countCRPFerritinTSHCoeliac serology on a gluten containing diet
This stage deliberately comes before any dietary trial. Anyone eating gluten free beforehand makes the investigation unusable. The European malabsorption consensus of 2025 puts it so that in malabsorption without another obvious explanation, coeliac disease should always be looked for.
The order decidesCalprotectin in the stool
To gauge whether inflammatory bowel disease is likely. Low means the all clear for this one question. Mildly raised means little at first and belongs read in context.
Strong negative findingWeak positive findingTargeted next steps, depending on the picture
Only now does it get specific: colonoscopy with stepwise biopsies in chronic watery diarrhoea, pancreatic elastase with greasy shiny stool and weight loss, bile acid diagnostics or a controlled treatment trial where the pattern fits, breath tests after bacterial overgrowth has been excluded.
Stepwise biopsiesElastaseBile acidsRunning across all stages is the travel and infection history. After long distance travel, with an immune deficiency or after a course of antibiotics, targeted pathogen testing belongs alongside, for parasites such as Giardia, amoebae or Cryptosporidium and for Clostridioides difficile. This order is orientation for your conversation, not a protocol to work through. It can change at any point if a finding comes in between. And it replaces no medical decision about what makes sense in your case. If a colonoscopy has been recommended, it stays recommended, even if stage 2 turns up something suspicious.
What stands out in this list is that the cheapest stage can often be the most productive one.
What the body loses along the way, and why that should not wait
There is an attitude I understand and still consider risky: I will just live with it. Understandable, because many people went a long time without an answer. Risky, because months of diarrhoea are not merely unpleasant, they shift a balance sheet.
What gets lost, in this order.
Fluid and electrolytes. Sodium, potassium and magnesium go with the water. You may notice this as weakness, muscle cramps, palpitations or dizziness on standing up. This is the part that becomes dangerous first when the course is severe.
Important: palpitations, confusion, barely passing urine, severe dizziness or a dry tongue with ongoing diarrhoea are not a point for waiting. They are a reason to seek medical help the same day, and if in doubt through the emergency number 112. Taking high doses of electrolytes on your own is not the answer here. With impaired kidney or liver function in particular it can do harm, and the same applies to magnesium and zinc preparations at higher doses.
Bile acids and with them fat digestion. If bile acids are lost through the stool over time, they are missing for the breakdown of fat. What hangs on fat then hangs along with it: the fat soluble vitamins A, D, E and K. This chain is physiologically compelling, and the European malabsorption consensus names nutrient deficiencies as a consequence of malabsorption in general. Solid frequency figures specifically for bile acid diarrhoea I did not find in this research. I therefore write this as mechanistically compelling, not as quantified.
Zinc. Here is the most concrete number in the whole chapter.
Khursheed Jeejeebhoy summarised in 2009 what balance studies show about zinc requirements during artificial nutrition.
Without losses through the gastrointestinal tract the estimated requirement was 3 milligrams per day, and on average 12 milligrams per day in people with diarrhoea and fistula losses. Zinc is a component of numerous enzyme systems.
For you that means: these figures apply to intake through a vein and cannot be transferred to capsules or tablets, because only part of what is swallowed is taken up. What they show is the direction. Anyone losing a lot through the gut can need considerably more, simply to balance the books. That could help explain why skin, hair, sense of taste, wound closure and immune defence can suffer after months of diarrhoea. How much is needed in the individual case belongs in medical hands.
Jeejeebhoy K. Gastroenterology. 2009;137(5 Suppl):S7-S12. PMID: 19874952 · DOI: 10.1053/j.gastro.2009.08.014 [Mechanism Review]The uncomfortable addition to this comes from Nancy Krebs. The gastrointestinal tract is the main site of zinc losses, and the amount excreted depends among other things on diarrhoea and fatty stool. At the same time sensitive biomarkers are missing for picking up especially mild zinc deficiency.
A normal zinc level in the blood is no proof that everything is in order.
With long lasting diarrhoea the situation counts more than the laboratory value. That is not a licence for self supply with supplements, quite the opposite. It is an argument for having the losses assessed medically, instead of relying on a single measurement. I deliberately name no doses here. They depend on too many factors that are only known in the individual case.
Iron. With chronic diarrhoea it can run short by two routes: through reduced uptake and through small, often unnoticed blood losses. A low ferritin therefore belongs in the basic panel and is at the same time a red flag that needs further assessment. Why uptake in the gut can be the real bottleneck is described in the article on iron deficiency and absorption problems.
How you might notice deficiency states in everyday life, without a laboratory: exhaustion that sleep no longer clears. Skin that turns dry. Small wounds that take longer to close. Hair that thins. A sense of taste that flattens. Infections that come more often and stay longer. Nails that break. None of this proves anything on its own. Together with months of diarrhoea they are a reason to raise the topic.
And this is why enduring it is the wrong strategy. Not because diarrhoea is dangerous in itself, but because a deficit running for months across several balance sheets at once eventually shows an effect. Which micronutrients play which role in energy metabolism is the topic of the article on micronutrients as cofactors.
Chronic diarrhoea is rarely clarified by running more tests. Most of the time it is clarified because somebody asks the right questions in the right order. And the first of them costs nothing: which medications do you take, and since exactly when has it been like this?
If you are right at the beginning and all of this feels like too much, start with the whole picture. How diagnostics, nutrition and rebuilding fit together sensibly in the gut is described in the overview article on the gut reset.
Time in this field is therefore a factor of its own, not merely a side effect.
Frequently asked questions about chronic diarrhoea
When does diarrhoea count as chronic?
The British HTA report on the SeHCAT test gives the common working definition: more than three loose or liquid bowel movements a day and or a stool weight above 200 grams per day, chronic once it has lasted longer than four weeks. What matters is the duration, not the severity. Beyond that line the diagnostic question shifts from the search for a pathogen to the search for a mechanism.
I have had diarrhoea for weeks, but the colonoscopy was normal. What does that mean?
A normal colonoscopy rules out what can be seen. Microscopic colitis cannot be seen. The European guideline defines it explicitly through a normal or near normal endoscopic appearance alongside characteristic changes in the tissue. Without stepwise biopsies from several segments of the large intestine, that question has not been answered. Coeliac disease, bile acid loss and a weak pancreas are also still open, because none of them becomes visible in the large intestine.
What is bile acid loss, and can I have it even though my gallbladder is still there?
Yes. About 95 percent of bile acids are taken back up at the end of the small intestine. What slips through draws water into the large intestine and speeds up transit. This can happen without any surgery at all. One central mechanism is a disturbed feedback loop through the gut hormone FGF19, which leads the liver to produce more bile acids than the small intestine can retrieve. Two meta-analyses found bile acid malabsorption on the SeHCAT test in 28.1 percent and 32 percent of people with diarrhoea predominant irritable bowel syndrome.
What is the SeHCAT test, and can I get it in Germany?
With the SeHCAT test you swallow a capsule containing a radioactively labelled substance that resembles a bile acid. Seven days later it is measured how much of it is still in your body. A low retention value points to bile acid loss. Because this is a nuclear medicine investigation involving radiation, it counts as a contraindication in pregnancy, and during breastfeeding, feeding has to be interrupted as instructed by the nuclear medicine department. Whether it is justified is decided by that department. In Germany the test is available at specialised nuclear medicine centres, not everywhere. The cost question belongs on the table beforehand. The British HTA report also notes that there is no true reference standard and that cut-off values differ between centres. This is why a medically supervised treatment trial is a well founded alternative as well.
My faecal pancreatic elastase is low. Do I now have pancreatic insufficiency?
Not yet. A meta-analysis of 13 studies with 888 people found a sensitivity of 0.94 at the 200 microgram per gram cut-off, but a specificity of only 0.69. So the test picks up almost every genuine weakness, yet it also flags people who do not have one. These figures come mostly from populations at increased risk, such as chronic pancreatitis or cystic fibrosis, and cannot simply be transferred to unselected chronic diarrhoea. With watery stool the value can come out too low through dilution alone. The European guideline therefore asks for an overall assessment from symptoms, nutritional status and a secretion test, not a verdict from a single laboratory value.
What is microscopic colitis, and why is it invisible during a colonoscopy?
Microscopic colitis is a chronic inflammation of the lining of the large intestine that only shows up under the microscope. Through the endoscope the mucosa looks normal or near normal. Typical is chronic watery, non bloody diarrhoea. There are three forms: collagenous colitis, lymphocytic colitis and the incomplete form. Without tissue samples from several segments of the large intestine the diagnosis cannot be made. In the collagenous form women are affected about three times as often, and the median age at diagnosis is around 65 years.
My calprotectin is mildly raised, but the colonoscopy was normal. What does that mean?
Calprotectin is an inflammation marker from white blood cells in the stool. Its strength lies in the low value: below about 40 micrograms per gram the probability of inflammatory bowel disease is at most one percent. Its positive predictive value, by contrast, is weak. In a meta-analysis of 18 studies it was only 4.2 percent at an assumed prevalence of one percent. A mildly raised value alongside a normal colonoscopy is therefore often not a sign of disease. It can also come from painkillers, an infection that has passed, or a trace of blood.
Could my blood pressure medication or my diabetes medication be behind the diarrhoea?
Both are described. For the blood pressure drug olmesartan, a Mayo Clinic case series documented 22 people with severe, coeliac like bowel disease, a median weight loss of 18 kilograms and villous atrophy. A gluten free diet brought no improvement. Only people who improved after medically supervised withdrawal were included in the series, however. That was an inclusion criterion and is therefore not a success rate. For metformin, diarrhoea and nausea are the most common side effects, dose dependent and often influenced by extended release forms or slower titration. Important: blood pressure and diabetes medication is never stopped or reduced on your own. That belongs in the consulting room.
Can an acid blocker cause diarrhoea?
It is being studied, and the data are split. Against randomly selected controls, a meta-analysis found an odds ratio of 4.55 for proton pump inhibitors and microscopic colitis. Compare instead with other people who also have diarrhoea, and that association disappears. Only one signal stays more stable: lansoprazole and collagenous colitis, with an odds ratio of 14.5 at low heterogeneity. That figure rests on only three observational studies, however, and shows an association, not a cause. What makes sense is to have the indication for the acid blocker reviewed medically. It is not stopped on your own.
I want to eat gluten free to see whether things get better. Is that a good idea?
Not before the coeliac work-up has been done. Both the antibodies in the blood and the tissue sample from the small intestine require that you are still eating gluten regularly at the time of the test. The American guideline from 2023 therefore puts serology on a gluten containing diet at the start and places the small bowel biopsy for confirmation alongside it for most adults. Anyone who goes gluten free first can make both investigations unusable and is left without a diagnosis. Test first, then leave things out.
My diarrhoea started with a stomach bug and never fully went away. Is that normal?
It is common enough to have a name. A meta-analysis of 45 studies with 21,421 people found irritable bowel syndrome in 10.1 percent twelve months after infectious enteritis, with a 4.2 fold higher risk in the first year. After protozoal infections the proportion was considerably higher, although based on few studies. The link to the infection absolutely belongs in your history. It does not replace looking into the other causes.
Diarrhoea at night that wakes me up: is that a red flag?
Yes, symptoms that pull you out of sleep are among the signs that belong in prompt medical assessment. Functional complaints typically pause during sleep. Diarrhoea that does not shifts the probabilities towards an organic cause, for example microscopic colitis, inflammatory bowel disease or bile acid loss. Together with blood in the stool, unintentional weight loss, fever or anaemia, this is no reason to wait.
What does my body lose when the diarrhoea goes on for months?
First water and electrolytes, above all sodium, potassium and magnesium. Then, once bile acids are lost, fat digestion and with it the fat soluble vitamins A, D, E and K. For zinc there is the most concrete number of all, but it comes from a very particular setting: in balance studies during artificial, that is intravenous, nutrition the estimated requirement without losses through the gut was 3 milligrams per day, and on average 12 milligrams per day with diarrhoea and fistula losses. These figures cannot be transferred to intake by mouth, because only part of swallowed zinc is taken up at all. They show the direction, not the amount. A sensitive laboratory value for mild zinc deficiency is missing, so the situation counts more than the measurement. Iron can be affected too. That belongs in medical hands, not supplemented on suspicion.
Related topics from other areas
Fibre
What the number 30 is worth and when more is not better
Sleep and the microbiome
Why the rhythm runs along in the gut
Gut brain axis
Vagus, stress and the motor function of the gut
Histamine intolerance
DAO, symptoms and the limits of diagnostics
Paleo and AIP
Elimination protocols in autoimmune conditions
Inflammatory foods
What processing can set off in the body
Scientific sources
- Arasaradnam RP, Brown S, Forbes A et al. Guidelines for the investigation of chronic diarrhoea in adults: British Society of Gastroenterology, 3rd edition. Gut. 2018;67(8):1380-1399. PMID: 29653941 · DOI: 10.1136/gutjnl-2017-315909 [Guideline]
- Smalley W, Falck-Ytter C, Carrasco-Labra A, Wani S, Lytvyn L, Falck-Ytter Y. AGA Clinical Practice Guidelines on the Laboratory Evaluation of Functional Diarrhea and Diarrhea-Predominant Irritable Bowel Syndrome in Adults (IBS-D). Gastroenterology. 2019;157(3):851-854. PMID: 31302098 · DOI: 10.1053/j.gastro.2019.07.004 [Guideline]
- Mearin F, Lacy BE, Chang L, Chey WD, Lembo AJ, Simren M, Spiller R. Bowel Disorders (Rome IV). Gastroenterology. 2016. PMID: 27144627 · DOI: 10.1053/j.gastro.2016.02.031 [Guideline]
- Layer P, Andresen V, Allescher H et al. Update S3-Leitlinie Reizdarmsyndrom (DGVS und DGNM). Z Gastroenterol. 2021;59(12):1323-1415. PMID: 34891206 · DOI: 10.1055/a-1591-4794 [Guideline]
- Lenti MV, Hammer HF, Tacheci I et al. European Consensus on Malabsorption. Part 1: Definitions, Clinical Phenotypes, and Diagnostic Testing for Malabsorption. United European Gastroenterol J. 2025;13(4):599-613. PMID: 40129317 · DOI: 10.1002/ueg2.70012 [Guideline]
- Dominguez-Muñoz JE, Vujasinovic M, de la Iglesia D et al. European guidelines for the diagnosis and treatment of pancreatic exocrine insufficiency. United European Gastroenterol J. 2024;13(1):125-172. PMID: 39639485 · DOI: 10.1002/ueg2.12674 [Guideline]
- Miehlke S, Guagnozzi D, Zabana Y et al. European guidelines on microscopic colitis: United European Gastroenterology and European Microscopic Colitis Group statements and recommendations. United European Gastroenterol J. 2021;9(1):13-37. PMID: 33619914 · DOI: 10.1177/2050640620951905 [Guideline]
- Rubio-Tapia A, Hill ID, Semrad C, Kelly CP, Greer KB, Limketkai BN, Lebwohl B. American College of Gastroenterology Guidelines Update: Diagnosis and Management of Celiac Disease. Am J Gastroenterol. 2023;118(1):59-76. PMID: 36602836 · DOI: 10.14309/ajg.0000000000002075 [Guideline]
- Rezaie A, Buresi M, Lembo A et al. Hydrogen and Methane-Based Breath Testing in Gastrointestinal Disorders: The North American Consensus. Am J Gastroenterol. 2017;112(5):775-784. PMID: 28323273 · DOI: 10.1038/ajg.2017.46 [Guideline]
- Slattery SA, Niaz O, Aziz Q, Ford AC, Farmer AD. Systematic review with meta-analysis: the prevalence of bile acid malabsorption in the irritable bowel syndrome with diarrhoea. Aliment Pharmacol Ther. 2015;42(1):3-11. PMID: 25913530 · DOI: 10.1111/apt.13227 [Meta-analysis, k=6, n=908]
- Liu T, Ma M, Li K, Tan W, Yu H, Wang L. Biomarkers for Bile Acid Malabsorption in Diarrhea-predominant Irritable Bowel Syndrome: A Systematic Review and Meta-analysis. J Clin Gastroenterol. 2023;57(5):451-458. PMID: 36867517 · DOI: 10.1097/MCG.0000000000001841 [Meta-analysis, k=10, n=1,034]
- Wedlake L, A'Hern R, Russell D, Thomas K, Walters JRF, Andreyev HJN. Systematic review: the prevalence of idiopathic bile acid malabsorption as diagnosed by SeHCAT scanning in patients with diarrhoea-predominant irritable bowel syndrome. Aliment Pharmacol Ther. 2009;30(7):707-717. PMID: 19570102 · DOI: 10.1111/j.1365-2036.2009.04081.x [Systematic Review]
- Riemsma R, Al M, Corro Ramos I et al. SeHCAT for the investigation of bile acid malabsorption and measurement of bile acid pool loss: a systematic review and cost-effectiveness analysis. Health Technol Assess. 2013;17(61):1-236. PMID: 24351663 · DOI: 10.3310/hta17610 [Systematic Review]
- Wilcox C, Turner J, Green J. Systematic review: the management of chronic diarrhoea due to bile acid malabsorption. Aliment Pharmacol Ther. 2014;39(9):923-939. PMID: 24602022 · DOI: 10.1111/apt.12684 [Systematic Review]
- Walters JRF. Defining primary bile acid diarrhea: making the diagnosis and recognizing the disorder. Expert Rev Gastroenterol Hepatol. 2010;4(5):561-567. PMID: 20932141 · DOI: 10.1586/egh.10.54 [Mechanism Review]
- de la Iglesia D, Agudo-Castillo B, Galego-Fernández M, Rama-Fernández A, Domínguez-Muñoz JE. Diagnostic Accuracy of Fecal Elastase-1 Test for Pancreatic Exocrine Insufficiency: A Systematic Review and Meta-Analysis. United European Gastroenterol J. 2025;13(8):1571-1582. PMID: 40569793 · DOI: 10.1002/ueg2.70061 [Meta-analysis, k=13, n=888]
- Leeds JS, Hopper AD, Sidhu R et al. Some patients with irritable bowel syndrome may have exocrine pancreatic insufficiency. Clin Gastroenterol Hepatol. 2010;8(5):433-438. PMID: 19835990 · DOI: 10.1016/j.cgh.2009.09.032 [Cohort, n=514]
- Tong J, Zheng Q, Zhang C, Lo R, Shen J, Ran Z. Incidence, prevalence, and temporal trends of microscopic colitis: a systematic review and meta-analysis. Am J Gastroenterol. 2015;110(2):265-276. PMID: 25623658 · DOI: 10.1038/ajg.2014.431 [Meta-analysis, k=25]
- Tarar ZI, Farooq U, Gandhi M et al. Are Drugs Associated with Microscopic Colitis? A Systematic Review and Meta-Analysis. Diseases. 2022;11(1):6. PMID: 36648871 · DOI: 10.3390/diseases11010006 [Meta-analysis, k=13, n=304,482]
- Nevalainen A, Nevalainen OPO. Autoimmune and immune-mediated inflammatory diseases after exposure to acid-suppressive medication: A systematic review and meta-analysis. Int J Risk Saf Med. 2023;34(3):207-225. PMID: 36442213 · DOI: 10.3233/JRS-220012 [Meta-analysis, k=25]
- Seenarain V, Idrees M, Mogridge J, Sinha A, Thompson A. Random colonic biopsies in macroscopically normal colonoscopies: is there any benefit? A two-centre audit of current practice. ANZ J Surg. 2020;90(12):E163-E167. PMID: 32856361 · DOI: 10.1111/ans.16248 [Cohort, n=872]
- Rubio-Tapia A, Herman ML, Ludvigsson JF et al. Severe spruelike enteropathy associated with olmesartan. Mayo Clin Proc. 2012;87(8):732-738. PMID: 22728033 · DOI: 10.1016/j.mayocp.2012.06.003 [Case Series, n=22]
- Bonnet F, Scheen A. Understanding and overcoming metformin gastrointestinal intolerance. Diabetes Obes Metab. 2017;19(4):473-481. PMID: 27987248 · DOI: 10.1111/dom.12854 [Mechanism Review]
- Singh P, Arora A, Strand TA et al. Global Prevalence of Celiac Disease: Systematic Review and Meta-analysis. Clin Gastroenterol Hepatol. 2018;16(6):823-836.e2. PMID: 29551598 · DOI: 10.1016/j.cgh.2017.06.037 [Meta-analysis, k=96, n=275,818]
- Klem F, Wadhwa A, Prokop LJ et al. Prevalence, Risk Factors, and Outcomes of Irritable Bowel Syndrome After Infectious Enteritis: A Systematic Review and Meta-analysis. Gastroenterology. 2017;152(5):1042-1054.e1. PMID: 28069350 · DOI: 10.1053/j.gastro.2016.12.039 [Meta-analysis, k=45, n=21,421]
- Svendsen AT, Bytzer P, Engsbro AL. Systematic review with meta-analyses: does the pathogen matter in post-infectious irritable bowel syndrome? Scand J Gastroenterol. 2019;54(5):546-562. PMID: 31112663 · DOI: 10.1080/00365521.2019.1607897 [Meta-analysis]
- Menees SB, Powell C, Kurlander J, Goel A, Chey WD. A meta-analysis of the utility of C-reactive protein, erythrocyte sedimentation rate, fecal calprotectin, and fecal lactoferrin to exclude inflammatory bowel disease in adults with IBS. Am J Gastroenterol. 2015;110(3):444-454. PMID: 25732419 · DOI: 10.1038/ajg.2015.6 [Meta-analysis]
- An YK, Prince D, Gardiner F et al. Faecal calprotectin testing for identifying patients with organic gastrointestinal disease: systematic review and meta-analysis. Med J Aust. 2019;211(10):461-467. PMID: 31680263 · DOI: 10.5694/mja2.50384 [Meta-analysis, k=18]
- Jeejeebhoy K. Zinc: an essential trace element for parenteral nutrition. Gastroenterology. 2009;137(5 Suppl):S7-S12. PMID: 19874952 · DOI: 10.1053/j.gastro.2009.08.014 [Mechanism Review]
- Krebs NF. Update on zinc deficiency and excess in clinical pediatric practice. Ann Nutr Metab. 2013;62(Suppl 1):19-29. PMID: 23689110 · DOI: 10.1159/000348261 [Mechanism Review]
- Ebert EC. The thyroid and the gut. J Clin Gastroenterol. 2010;44(6):402-406. PMID: 20351569 · DOI: 10.1097/MCG.0b013e3181d6bc3e [Mechanism Review]
- The prevalence figures for bile acid loss come from irritable bowel cohorts, not from the general population. The figure circulating online that bile acid malabsorption affects about one percent of the population could not be traced back to a solid primary source. It therefore does not appear in this text.
- The blood markers C4 and FGF19 are not standardised. The direction of the effects is consistent across studies, but the reference ranges differ between studies, and the authors of the meta-analysis themselves call for better determination of test performance. Mechanistically plausible, formally not yet sufficiently examined.
- The treatment trial with a bile acid binder is not validated as a diagnostic procedure. It is economically attractive, but a study is missing that shows it identifies the same people as the SeHCAT test. A non-specific stool thickening effect could wrongly be counted as a diagnosis.
- On the availability of the SeHCAT test in Germany there is no solid public overview. Figures on the number of centres and on reimbursement that circulate online come mostly from hospital websites. This is why the text says only that the test is available at specialised centres and that the cost question belongs settled beforehand.
- The medication associations in microscopic colitis are methodologically open to attack. High against random controls, gone against diarrhoea controls, with very high heterogeneity. Only the signal for lansoprazole and the collagenous form persists, with moderate certainty of evidence and low heterogeneity, but it rests on only three observational studies and shows an association, not a cause.
- Olmesartan is documented, the substance class is not. The case series concerns olmesartan exclusively. Extending the statement to all sartans is not covered by these data and remains an open question. In addition, only people who improved after withdrawal were included in the case series at all. The improvement is therefore an inclusion criterion and not a success rate, and the paper says nothing about how often this picture occurs.
- On magnesium as a cause of diarrhoea I have no verifiable frequency figure for adults. The osmotic effect is undisputed and dose dependent. I therefore carry it as a clinical observation without a number and without a source.
- For sorbitol an agreed test dose is missing. The North American breath test consensus names doses for lactulose, glucose, fructose and lactose, but not for sorbitol. Sorbitol malabsorption is therefore an observation and not a diagnosis with a threshold.
- The figures on the post infectious course after protozoa and after viral infections are uncertain. They rest on few studies with considerable heterogeneity. The order of magnitude after bacterial pathogens is far better secured.
- Fat soluble vitamins in bile acid diarrhoea are mechanistically compelling, but not quantified. Frequency figures specifically for this constellation I did not find in this research.
- Zinc cannot be measured reliably in mild deficiency. The routes of loss are well documented, a sensitive biomarker is missing. This is why no target value and no dose appears here. The balance figures of 3 and 12 milligrams also come from artificial, intravenous nutrition and cannot be transferred to intake by mouth.
- The diagnostic accuracy of faecal elastase comes mostly from high risk populations. The 2025 meta-analysis recommends the test for initial screening in people at increased risk, such as chronic pancreatitis or cystic fibrosis. In unselected chronic diarrhoea it can perform more weakly than the pooled figures suggest.
- The response rates to bile acid binders were not collected under controlled conditions. The figures of 96, 80 and 70 percent come from largely open observations without a placebo group. They are more likely to overstate the effect, and the individual proportions rest on different subsets of the included studies.
- Two of the cited guidelines are usable only as a framing reference. For the AGA guideline from 2019 and the German S3 guideline on irritable bowel syndrome from 2021 there is no abstract in PubMed. Only their existence and subject were cited, no individual strengths of recommendation and no figures.
- What deliberately does not appear here. No dosing information, no copyable protocol and no advice to change, reduce or stop an existing medication. That applies explicitly to blood pressure, diabetes and thyroid preparations, to acid blockers and to psychiatric medication. Nothing in any section of this article implies that a recommended colonoscopy, endoscopy or laboratory work-up should be postponed or replaced. What I describe from my consulting room is marked as observation and is not a study result.