0
0
For ten years, the tests kept coming back normal.
That is the part of this story that is hard to shake. Researchers went back to old blood samples — frozen, filed away, taken years before anyone was sick — and found signs of IBD 10 years before diagnosis. The pattern was already sitting in the blood. Every test anyone knew how to run at the time said nothing was wrong.
Both of those things are true at once. That is what makes it worth sitting with.
Why a Clean Test Can Still Leave You Uneasy
If you have ever been handed a normal result and not quite believed it, you are not being difficult. You are being accurate.
A test does not tell you that you are well. It tells you that the one thing it was built to look for was not found, on the day it looked. Those are very different sentences. Most of us hear the first one and go home with the second one.
And if you have spent years with a gut that is "fine" — fine, but not quite right, fine in a way you have stopped mentioning to anyone — that gap between the two sentences is where you live.
What the Blood Showed: IBD 10 Years Before Diagnosis
Here is what the researchers actually did.
They took 2,000 blood serum samples from 500 people. Of those 500, two hundred went on to develop Crohn’s disease. Two hundred went on to develop ulcerative colitis. One hundred stayed healthy, and were matched to the others by age and sex so the comparison would be fair.
The samples had been collected at four points: at diagnosis, and roughly two, four and ten years before it. So for each person, there was a thin trail of blood going backwards in time, from the day they were told, to a decade before they had any reason to think about their gut at all.
Then they screened all of it against around 357,000 different antigens — bits of bacteria, viruses, food and the body’s own tissue — and counted which antibodies each sample had made. That turned up 123,980 unique antibodies across the whole set.
The patterns in the people who would later get sick were already different. Not at diagnosis. Up to ten years out.
The work was presented in October 2026 at UEG Week 2026, a European gastroenterology conference, by Dr Arno Bourgonje of the University Medical Center Groningen in the Netherlands, who carried out the research during a fellowship at the Icahn School of Medicine at Mount Sinai in New York. You can read the announcement of the findings here.
The Strangest Part: The Signal Was Quieter, Not Louder
This is the detail that stopped me, and it is not the one you would guess.
When we imagine a disease brewing for ten years, we picture something building. Rising. An alarm getting louder until somebody finally hears it.
That is not what the blood showed.
Antibody diversity was significantly lower in the people who would later develop Crohn’s or colitis, compared with the people who stayed well. Their immune systems were not shouting. Their range was narrowing. And in the group heading for Crohn’s disease, responses to encapsulated bacteria — a whole family of microbes the body normally keeps tabs on — went down steadily as diagnosis got closer, not up.
Alongside that narrowing, some specific responses did climb: elevated reactions to herpesviruses, and antibodies against flagellin, the protein that makes up the little tails bacteria swim with. In the colitis group, herpesvirus responses rose too, along with responses to the body’s own tissue.
So the ten-year signal was not an alarm. It was a shape — a few things going up while the overall repertoire quietly thinned out.
That may be the honest answer to "why did nobody catch this." You cannot hear a signal that is getting quieter if you are listening for one that gets loud. And you cannot see a pattern spread across 123,980 antibodies by looking at any one of them. It only exists when you can hold all of it at the same time, which is a thing that has only recently become possible to do.
It Also Seemed to Know the Shape of What Was Coming
The researchers fed the antibody patterns to a machine-learning model and asked it to sort people it had never seen.
Using blood drawn up to ten years early, it separated future Crohn’s cases from healthy people with an accuracy score of 0.90, and future colitis cases with 0.84. In plain terms: if you handed the model one person who would later get Crohn’s and one who never would, it ranked them in the right order about nine times out of ten.
Then it did something stranger. It separated people who would go on to have complicated Crohn’s disease — the kind that narrows or tunnels through the bowel wall and often needs surgery — from those whose disease would stay simpler. Score: 0.85. From blood taken about a decade before anyone had symptoms.
Bourgonje’s point about that is the careful one. The profile of complicated disease, he said, is "not simply the consequence of disease progression after diagnosis." It was not the illness leaving footprints as it got worse. It was there first. His summary: "complicated CD at diagnosis is associated with a distinct antibody profile years before the first clinical symptoms."
Not just whether. Something about which kind. Already written down, a decade early, in a language nobody could read yet.
What This Does Not Mean Yet
It would be easy to oversell this, so here is the fence around it, drawn by the researchers themselves.
They call it a proof of principle. Their own words are that "further research is needed before they could be translated into a clinical screening test." There is no test you can ask for. Your doctor cannot order this.
What is next, in Bourgonje’s words, is to "validate these signatures in independent, more diverse and prospectively followed populations" — a larger, more varied group of people followed forwards in time rather than looked at backwards. After that: working out whether these patterns add anything beyond the genetic and clinical markers doctors already use, building an assay that is reproducible and cheap enough to be worth running, and then showing that running it actually makes people’s lives better. That last step is the one that gets skipped in the retelling and is the only one that really counts.
Also worth saying plainly: an association is not a cause. Finding more herpesvirus antibodies in people who later got Crohn’s does not mean a virus caused it. It might be a trigger. It might be a bystander. It might be a sign of an immune system already behaving differently for some other reason entirely. Nobody knows yet, and the honest version of this story says so.
What You Can Actually Do With This
Not much medically, today. But a few things are still true and free.
- Keep your own record. You are the only one holding the whole timeline. Dates, what changed, what you ate, how long it lasted — a few lines in a cheap notebook or your phone’s notes. A pattern over two years is something no single appointment can see.
- Say the quiet part out loud at the appointment. "This has been going on for three years and I have stopped mentioning it" is more useful information than today’s symptom score.
- Treat "normal" as one data point, not a verdict. It is reasonable to accept a result and still ask what would make you come back.
- Stop the 2am searching. Checking your body over and over feels like vigilance and works like fuel. If that loop is familiar, this is worth reading on why checking makes it worse.
And if you already have a diagnosis, the flip side of this story matters too: feeling better and being healed are not the same measurement, which is something researchers have found inside remission as well.
Being Read Before You Have Words for It
There is one more thing in this that has nothing to do with medicine.
For ten years, those people were already being described in full. Accurately. In detail. Including the part about how severe it would get. The description existed the whole time; what was missing was anyone able to read it.
There is a very old idea — older than any of this science, carried around in poetry for thousands of years — that a person is already known at exactly that depth by God. Not watched. Read. Understood from a long way off, before the thought finishes forming, before there are words for it, and certainly before there are symptoms. The people who wrote that down did not find it frightening. They found it to be the most comforting thing they knew.
What is on offer there is not a cure, and it would be dishonest to pretend otherwise. It is something smaller and more durable: that none of those ten silent years were blank. Nothing in you was ever unread, including the stretch when every test said nothing at all.
If the not-knowing is the part that actually keeps you up, you might find our free How Anxious Am I? assessment useful — it takes about two minutes and names what the worry is actually made of.
The Decade Nobody Got to Watch
Five hundred people gave a little blood on an ordinary day. Most of them will never know that the sample sat in a freezer for a decade and then said something true about them that nobody could hear at the time.
That is the quietly hopeful part of this, and it is not a promise about your gut. It is just this: "we did not find anything" and "there was nothing there" were never the same sentence. The body has been telling the truth the whole time. We are only now learning to read the handwriting.
Questions & Answers
Can a blood test detect IBD before symptoms start?
Not yet, and not one you can ask for. Research presented at UEG Week 2026 showed that antibody patterns in stored blood samples differed in people who later developed Crohn’s disease or ulcerative colitis, up to ten years before they were diagnosed. The researchers describe it as a proof of principle and state that further research is needed before it could become a clinical screening test. No such test is currently available to patients or doctors.
What did the IBD antibody study actually find?
Researchers analysed 2,000 blood serum samples from 500 people: 200 who later developed Crohn’s disease, 200 who later developed ulcerative colitis, and 100 healthy people matched by age and sex. Samples had been taken at diagnosis and roughly two, four and ten years earlier. Screening against around 357,000 antigens turned up 123,980 unique antibodies. Antibody diversity was significantly lower in the people who went on to develop either disease, and distinct antibody patterns were visible up to ten years before diagnosis.
Does Epstein-Barr or another virus cause inflammatory bowel disease?
No study has shown that it does, and this one does not claim it. The research found elevated antibody responses against herpesviruses in people who later developed Crohn’s disease, and rising herpesvirus responses in those who later developed ulcerative colitis. That is an association, not a cause. A raised antibody response could reflect a trigger, a bystander infection, or an immune system already behaving differently for some other reason. The question is unresolved.
How accurate was the model at predicting IBD ten years early?
A machine-learning model using the antibody patterns scored 0.90 for separating future Crohn’s disease cases from healthy controls and 0.84 for future ulcerative colitis cases, using blood taken up to ten years before diagnosis. It also scored 0.85 at separating people who would later develop complicated Crohn’s disease from those whose disease stayed uncomplicated. A score of 0.90 means that given one person who later developed the disease and one who did not, the model ranked them in the correct order roughly 90% of the time. These are research results on a single stored sample set, not validated clinical performance.
Why did nobody notice IBD in these blood samples at the time?
Two reasons. First, the signal pointed the wrong way for anyone looking: overall antibody diversity went down rather than up, and in the group heading towards Crohn’s disease, responses to encapsulated bacteria declined steadily as diagnosis approached. Tests are usually built to catch something rising. Second, the pattern only exists across tens of thousands of antibodies at once. Screening 357,000 antigens in a single sample is recent technology; no individual test from a decade ago could have seen a pattern of that shape.
A Question Worth Sitting With
If a test existed that could tell you, with reasonable accuracy, that you were likely to develop a serious illness in about ten years — with no treatment available to change the outcome — would you want to be told? And does your answer change if it could also tell you how severe it was going to be? There is no right answer here and smart people land in both camps. Tell us where you come down in the comments.
If You Want to Share This
Short: They went back to blood samples taken ten years before anyone was diagnosed with IBD — and the pattern was already there. The weird part is the signal was getting quieter, not louder. That is why nobody caught it. https://bgodinspired.com/index.php/health-and-wellness/ibd-10-years-before-diagnosis/
Longer: This reframed something for me. "We did not find anything" and "there was nothing there" are not the same sentence. Researchers found antibody signatures of Crohn’s and colitis in stored blood up to a decade before diagnosis — and in some cases could tell how severe the disease would become. No test exists yet. But for anyone who has been told for years that nothing is wrong and never quite believed it: you were not imagining it. https://bgodinspired.com/index.php/health-and-wellness/ibd-10-years-before-diagnosis/
For anyone living with it: If you love someone who spent years being told nothing was wrong before they got an IBD diagnosis, this one is worth ten minutes. The body was describing it the whole time. Nobody could read the handwriting yet. https://bgodinspired.com/index.php/health-and-wellness/ibd-10-years-before-diagnosis/
