Celiac disease test: the blood tests, the biopsy, and why you keep eating gluten until then

Short answer

A celiac disease test usually starts with one blood test — tissue transglutaminase IgA (tTG-IgA), drawn together with total IgA — and, if that is positive, an upper endoscopy with several duodenal biopsies to confirm it (1, 5). Both only work while you are still eating gluten, so you keep eating it until your doctor says the workup is finished (2).

If you have spent years being told it is stress, or if you already feel better off gluten and now resent being asked to eat it again, you're not imagining it and it isn't your fault. The average gap between the first gut symptoms and a celiac diagnosis has been measured at around eleven years, and most Americans who have the disease have not been told so (9, 18). The gluten rule is not doctors doubting you. It is the only way the tests can see anything at all.

Note: this article explains what guidelines and research say about testing and is not medical advice; your doctor or gastroenterologist decides which tests fit you. Red flags, the same wherever they appear below: see a clinician if diarrhea or digestive discomfort lasts more than two weeks, and sooner for unintentional weight loss, blood in your stool or black, tarry stools, vomiting blood, repeated vomiting or trouble swallowing (16). After an upper endoscopy, contact your doctor immediately for fever, trouble swallowing, throat, chest or abdominal pain, or bleeding including black stools, and call 911 for severe chest pain, heavy bleeding or trouble breathing (20).

The celiac disease test path at a glance

Almost everything below hangs on one condition: you are eating gluten while it happens.

A blood test is a few minutes and a vial. The endoscopy is a morning, sedation and someone to drive you home. The slow part, for most people, is the waiting in between — and the gluten you have to keep eating through it.

Why can't I just go gluten-free and see if I feel better?

Because removing gluten starts repairing exactly what the tests are built to detect, and within months both the blood test and the biopsy can read normal in someone who genuinely has celiac disease. The AGA's clinical practice update puts it plainly: reducing or avoiding gluten before diagnostic testing "is discouraged as it may reduce the sensitivity of both serology and biopsy testing" (2). NIDDK gives patients the same instruction, and so does Mayo Clinic (3, 16).

The numbers behind that are not subtle. tTG antibodies have a half-life of about six months; on a gluten-free diet roughly 8 in 10 people with diagnosed celiac disease test negative within a year, and about 9 in 10 by five years (9). The lining of the intestine heals on its own timetable: symptoms often ease within weeks, while full regrowth of the villi can take up to three years or more (9). A test taken partway through that repair can easily land in the normal range and tell you nothing true.

The second problem is that feeling better off gluten does not identify what you have. Plenty of people improve on a gluten-free diet without celiac disease — because of non-celiac gluten or wheat sensitivity, because wheat is a major source of fermentable carbohydrates that irritate an irritable bowel, or because cutting out bread quietly changed half of what else they ate. "I feel better" is a reason to test, not a result.

And the label does matter, even for someone who would happily eat gluten-free forever. A confirmed diagnosis is what gets your first-degree relatives screened, what puts nutrient and bone checks on your chart, and what tells a dietitian how strict the diet has to be. Without it, you carry a lifelong restriction and none of the follow-up — and you cannot get the diagnosis back later without eating gluten again for weeks or months (7).

If bloating is the symptom pushing you toward the gluten-free aisle, it is worth reading what else produces it before you cut anything: we cover that in bloated all the time.

The celiac blood test: what tTG-IgA actually measures

tTG-IgA, and why total IgA is drawn with it

tTG-IgA measures IgA antibodies your immune system makes against tissue transglutaminase, an enzyme in the gut wall. In celiac disease, gluten sets off that reaction; in almost nothing else does it happen the same way. It is the preferred first test for children and adults, and guidelines pair it with a total IgA level in the same draw (1, 5).

The Celiac Disease Foundation puts tTG-IgA at about 93% sensitivity and 96% specificity in people eating gluten — good, not perfect (4). Mildly raised results also turn up in other autoimmune conditions such as type 1 diabetes and Hashimoto's thyroiditis, which is one reason a positive result is a step and not a verdict (4).

Total IgA is there for a specific trap. If your body makes very little IgA, an IgA-based test comes back negative no matter how much celiac disease you have.

What a "celiac panel" usually includes

Labs sell the tests in bundles, and the bundles differ. A typical celiac panel contains some combination of:

• tTG-IgA — the workhorse first test.

• Total serum IgA — the validity check on the test above.

• EMA-IgA (endomysial antibody) — near-100% specific but less sensitive, with 5–10% of people with celiac disease testing negative; labor-intensive, so it is kept for confirming unclear cases, usually on a second sample (4, 5).

• DGP-IgA and DGP-IgG (deamidated gliadin peptide) — useful in IgA deficiency, in very young children, and sometimes alongside a weak tTG-IgA (4, 5).

• tTG-IgG — for people with IgA deficiency. In anyone with normal IgA, an IgG-based tTG result is not specific and should not be read as evidence of celiac disease on its own (2).

Older tests for anti-gliadin antibodies (AGA-IgA, AGA-IgG) still appear on some panels and in a lot of direct-to-consumer kits; they were replaced for good reason and are not what current algorithms use (5).

If you are IgA deficient

Selective IgA deficiency affects roughly 1 in 500 people and about 2% of people with celiac disease, and the two travel together more often than chance (14). It is the single most common reason a person with real celiac disease gets a clean tTG-IgA result. The AGA calls it "an infrequent but important explanation for why patients with CD may be negative on IgA isotype testing despite strong suspicion," and recommends measuring total IgA and switching to IgG-based tests — tTG-IgG and DGP-IgG — when it turns up (2, 5).

This is worth checking on your own lab report. If you only see a tTG-IgA line and no total IgA, ask whether IgA deficiency has ever been ruled out.

Children, and children under 2

For children over 2, the path looks like the adult one: tTG-IgA with total IgA first (1). Under 2, antibody tests are less reliable, and the ACG's 2023 update looked specifically at whether adding DGP helps in that age group; tTG-IgA remains the preferred single test, with IgG-based tests used in children with IgA deficiency (1). The Celiac Disease Foundation adds that children under 3 are generally screened at 2 or 3 unless symptoms appear earlier, and that a pediatric gastroenterologist should be the one evaluating poor growth or persistent diarrhea in a toddler (4).

What the number means: weak positive versus ten times the limit

The size of a positive tTG-IgA changes what happens next, which is why the lab report shows your value and the lab's upper limit of normal (ULN) side by side.

• Negative, IgA normal, low suspicion — Celiac disease is unlikely; look elsewhere (5)

• Negative, IgA normal, high suspicion — ACG still recommends a biopsy, to catch seronegative celiac disease (5)

• Negative, IgA undetectable — Repeat with tTG-IgG and/or DGP-IgG (5)

• Weak to moderate positive — Referral for endoscopy; EMA-IgA or DGP-IgA may be added first for confidence (5)

• More than 10× the upper limit — The high-confidence band — in children, a possible route to diagnosis without biopsy; in adults, still normally confirmed by biopsy (1, 6)

One practical point: different labs use different assays and different upper limits, so "32" at one lab is not "32" at another. If you will be repeating the test over time, having it run at the same facility makes the numbers comparable (9).

At-home kits and finger-prick tests

Point-of-care celiac tests exist and some have performed comparably to lab tests in studies, but the AGA notes they can be difficult to quality-control in inexperienced hands, and that getting a fast result is rarely the deciding factor in a celiac workup (2). Treat a home result — in either direction — as a reason to go to a clinician, not as an answer. And the gluten rule applies to home kits exactly as it does to lab ones.

Do I need an endoscopy biopsy if my celiac blood test is positive?

In adults, usually yes: the ACG's 2023 guideline recommends upper endoscopy with multiple duodenal biopsies to confirm the diagnosis in both children and adults with suspected celiac disease, calling it a strong recommendation (1). The reason is that a lifelong diet is a serious thing to assign on an antibody number alone, and that the biopsy also shows whether something other than celiac disease is damaging the gut.

What the pathologist is looking for

Not "gluten." The biopsy is read for the pattern celiac disease leaves: an increase in intraepithelial lymphocytes (above 25 per high-power field counts as lymphocytosis), crypt hyperplasia, and flattening of the villi — graded on the Marsh scale. The AGA recommends the report describe where each sample came from, the Marsh grade, the villus-to-crypt ratio and the lymphocyte count (2).

How many samples, and from where

More than one, and not all from the same spot. Damage in celiac disease is patchy, so guidelines call for multiple biopsies from different parts of the duodenum, including the duodenal bulb — with the caveat that the bulb contains Brunner glands and other normal variants that can muddy interpretation (2). A 2026 Cleveland Clinic Journal of Medicine review describes the same practice: multiple samples from the distal duodenum and the bulb, precisely because the disease can be patchy (10).

Sampling quality is a real-world weak point, not a theoretical one. The AGA notes that several comparative studies have found biopsy sampling falling short, that interpretation quality varies between pathology departments, and that around 5% of biopsies end up diagnostically uncertain (2). If your biopsy was inconclusive, that is a known failure mode with a known answer — ask, rather than assume the question is settled.

The no-biopsy route, and who it is for

For children, there is a validated shortcut. ESPGHAN's 2020 criteria allow a diagnosis without biopsy when tTG-IgA is above 10× the upper limit of normal and an endomysial antibody test on a second, separate blood sample is positive (6). The ACG endorses that combination for children (1). In practice this has cut the need for endoscopy in children by roughly 30–50% (2).

For adults the ACG is more cautious, and says so: the same criteria "may be considered after the fact, as a diagnosis of likely CD" in symptomatic adults unwilling or unable to have an endoscopy — a conditional recommendation, which in guideline language means weak (1). The evidence is thinner in adults than in children. A multicentre study across eight countries found that of 42 adults with tTG-IgA above 10× ULN, 40 — 95% — had villous atrophy on biopsy (12). Good, but not the near-100% predictive value seen in children. Most US adults will still be offered the endoscopy, and in the US it is often also the way other causes get excluded.

Two things the biopsy can't fix

It cannot see damage that has already healed, which is the whole reason for the gluten rule. And a normal biopsy with a clearly positive blood test does not automatically end the discussion: that combination can mean early-phase disease, a false-positive antibody result, or a sampling miss, and the standard next move is to stay on gluten and retest in 1–3 months (5).

What if I already stopped eating gluten? The gluten challenge

Then the honest answer is that your current tests cannot be trusted, and getting a real answer means eating gluten again on purpose, for a defined period, under a clinician's supervision — that is what a gluten challenge is (7). It is the least popular paragraph in any celiac article, and the most important one.

There is no single agreed protocol, which is itself worth knowing before someone quotes you a number as if it were settled:

Where does the short version come from? A 2013 trial in Gut put 20 adults with biopsy-proven celiac disease on either 3 g or 7.5 g of gluten a day for 14 days. By day 14 the villus-height-to-crypt-depth ratio had fallen from 2.2 to 1.1 and intraepithelial lymphocytes had risen from 32.6 to 51.8 per 100 enterocytes; antibody titres rose only slightly by day 14 and markedly by day 28. The authors concluded a 14-day challenge at 3 g a day or more produces histological and serological change in most adults (8).

That is the key asymmetry, and it is the thing most people get backwards: a short challenge is often enough to rule celiac disease in, and rarely enough to rule it out. If two weeks of gluten produces villous atrophy, you have your answer. If it doesn't, you have learned very little, and the NASPGHAN group's advice is to keep going rather than call it negative (7).

A few practical things worth knowing before you start:

• Serology moves slower than you'd like. Mild rises in antibodies — under 10× the upper limit — are supportive but not diagnostic, and in a very short challenge celiac blood tests add little; the endpoint that counts is the biopsy (7).

• Longer gluten restriction means a longer challenge. If you have only been off gluten for under two weeks, the gut architecture has probably not had time to restore itself, which shortens what is needed (7).

• Symptoms are expected, and are information, not failure. Studies report abdominal pain, diarrhea and bloating, plus fatigue, headaches and brain fog; higher doses bring them on faster. Report them rather than silently cutting the dose, so the plan can be adjusted (7). If brain fog is the symptom that worries you most, we've written about what causes brain fog more broadly.

• Not during pregnancy. The Celiac Disease Foundation is explicit that gluten challenges should never be done during pregnancy, and that a physician experienced in celiac disease should supervise one and move to biopsy quickly if symptoms get severe (4).

• It is not a solo project. The NASPGHAN group's framing is that clinicians should not remove gluten before testing is complete in the first place, and that when a challenge is needed it belongs in a plan made with a gastroenterologist, with scheduled check-ins (7).

Does a celiac genetic test (HLA-DQ2/DQ8) confirm celiac disease?

No. It can help rule celiac disease out, and it can never rule it in. Almost everyone with celiac disease carries HLA-DQ2 or HLA-DQ8 — but so do 25–30% of people in the general population, the overwhelming majority of whom will never develop the disease (4). NIDDK says the same in plainer words: not having the variants makes celiac disease very unlikely, while having them does not mean you have it (3).

A negative result is the useful one. The Celiac Disease Foundation puts a negative genetic test at roughly a 99% probability of not developing celiac disease (4). ARUP's algorithm states it in one line: "A negative HLA result rules out CD. A positive HLA result is permissive of CD but cannot be used to confirm a diagnosis" (5).

So where does it earn its place? The AGA reserves it for second-line use: when serology and histology disagree, when someone is seronegative but has suspicious biopsy findings, when a historic diagnosis was made without modern antibody tests — including in people diagnosed as very young children — and when a gluten-free diet has already been started and nobody wants to put the person through a challenge for nothing (2). That last case is the common one in real life. If you quit gluten years ago, a negative HLA test can close the question without a single slice of bread; a positive one tells you only that a challenge is still the way forward.

It also helps with family planning of a sort: first-degree relatives who share the same genotype have been estimated to carry up to a 40% risk, against 7–20% when the genotype is unknown (4).

One caveat: the test is not cheap, is often paid out of pocket, and can be run on blood, saliva or a cheek swab (4). It is also the one celiac test whose result will never change, which is both its strength and the reason not to order it reflexively.

Celiac disease, gluten sensitivity, wheat allergy: which test answers which question

These three get collapsed into "gluten problem" constantly, and they are tested in completely different ways.

Two honest limits on the middle column. There are no validated biomarkers for non-celiac gluten or wheat sensitivity, and the Salerno criteria may not cleanly separate it from functional gut disorders (10). It also isn't settled that gluten is the trigger rather than another component of wheat (15). None of that makes the symptoms less real — it means the diagnosis is built by elimination, and the elimination has to start with the condition that has a real test.

My celiac blood test was negative, but I still feel ill

A negative antibody test does not automatically close the question, and there are three different reasons why.

Seronegative celiac disease. Up to 5–16% of people with celiac disease do not produce detectable antibodies (9). This is exactly why the ACG says a biopsy should follow negative serology when pretest suspicion was high (5). The AGA adds the mirror case: if celiac disease is strongly suspected and biopsies are negative, tTG-IgA should still be run, and repeat biopsies considered then or later (2).

You weren't eating enough gluten. Half a slice of toast on Sundays is not a gluten-containing diet. This is the most common fixable reason for a negative test, and it is worth saying out loud to your doctor if it applies.

It genuinely isn't celiac disease. If both blood test and biopsy are negative while you are eating gluten, celiac disease can reasonably be concluded against (9) — and then the symptoms still need an explanation. The usual next candidates are the ones that look identical from the outside: irritable bowel syndrome, which has its own diagnostic route described in how IBS is diagnosed; small intestinal bacterial overgrowth, tested by breath test as covered in how to test for SIBO; lactose or fructose intolerance; and histamine intolerance. Stress and the gut–brain axis can amplify all of them without being the cause of any.

One more scenario trips people up: a positive blood test with a clean biopsy. That is sometimes called "potential celiac disease," and experts genuinely disagree about what to do next — some watch and recheck, some treat. It is a conversation to have with a clinician who sees celiac disease often, not a question with a single published answer (9).

Who gets tested for celiac disease, and when

The US does not screen everyone. The US Preventive Services Task Force concluded in March 2017 that the evidence is insufficient to weigh the benefits and harms of screening asymptomatic people — a grade I statement, not a recommendation against testing symptomatic ones (11). A literature surveillance report in August 2024 found nothing new that would justify revisiting it. The ACG's 2023 position is the same in practice: case finding rather than mass community screening (1).

Case finding means testing people whose odds are already raised. The logic, as the AGA describes it, is to lift the pretest probability from the roughly 1% population rate to 5–10% or above, where a positive result actually means something (2). The large US multicentre screening study behind that estimate also found celiac disease far more often in relatives of people who have it than in people at no particular risk, which is where the whole case-finding logic comes from (18). The groups that qualify:

• Symptoms that fit. Chronic unexplained digestive symptoms, poor growth in a child, weight loss, iron-deficiency anemia, delayed puberty (5). Celiac disease also shows up without gut symptoms at all — as low iron, weak bones, an itchy blistering rash, mouth sores, headaches, numbness in the hands and feet, joint pain or raised liver enzymes (16).

• First-degree relatives. A meta-analysis put the risk in asymptomatic first-degree relatives at 7.5% (2); the Celiac Disease Foundation uses the rule of thumb of about 1 in 10 versus 1 in 100 (4). Children of people with celiac disease are recommended to be screened at least once even without symptoms (9).

• Associated conditions. Type 1 diabetes above all — celiac disease is present in 5–10% of children with it (2) — plus autoimmune thyroid and liver disease, Down, Turner and Williams syndromes, and selective IgA deficiency (4).

If you have a rash that blisters and itches, say so before the blood draw. Dermatitis herpetiformis is celiac disease showing up in the skin, and it is diagnosed from small skin samples taken on and next to the rash rather than from the gut (3).

Living with a celiac workup: the questions people ask next

How long does the whole thing take?

If you are already eating gluten: a blood draw, then results in days, then — if positive — a gastroenterology referral and an endoscopy, with the queue usually being the slow part. Weeks, mostly. If you stopped gluten first, add the challenge: the NASPGHAN guide's default is a minimum of 12 weeks with a check-in at 4–6 weeks, and a shortened 6–12 weeks when symptoms are hard to tolerate (7). Planning around it beats improvising through it.

I'm pregnant, or trying. Should I wait?

If you are already eating gluten, the blood test itself is just a blood draw, and the decision about endoscopy is one for your obstetric clinician and gastroenterologist together. What is clear is the other case: the Celiac Disease Foundation says gluten challenges should never be done during pregnancy (4). If you are gluten-free and planning a pregnancy, that argues for sorting the testing out before, not during.

Will I need the endoscopy again later?

Usually not routinely. After diagnosis, follow-up is individualized; blood tests are repeated regularly while repeat biopsies are much less often needed, and your clinician decides which (9). The AGA suggests follow-up serology at 6 and 12 months after diagnosis and yearly after that, and notes that negative follow-up antibodies do not guarantee the gut has healed — which is why persistent symptoms can still earn a repeat biopsy (2).

My test was negative. Can I be tested again later?

Yes. Celiac disease can develop at any age, so a negative test is a snapshot, not a lifetime clearance — people with ongoing or new symptoms can be rechecked (9). For gene-positive first-degree relatives, the Celiac Disease Foundation suggests periodic rescreening, though its own page gives two different intervals — every 3–5 years in one place and every 2–3 years in another, which is a good illustration of why the interval is a question for your clinician rather than a web page (4).

My kid needs testing. Is it the same?

Mostly, with two differences. Children can qualify for the no-biopsy route when tTG-IgA is above 10× ULN and EMA is positive on a second sample (6, 1). And a gluten challenge in a child is weighed against school, growth and daily life, which is why the NASPGHAN group builds in a formal check-in at 4–6 weeks to adjust the plan rather than grinding through it (7).

Everyone says "just try gluten-free for a month." What do I say back?

That a month gluten-free is the one thing that can take the answer away. Feeling better proves nothing about which condition you have, and the tests lose sensitivity as the gut repairs (2). The sequence that works is test first, then trial — in that order, and treatment decisions after a diagnosis belong to your clinician and a dietitian, not to this page.

How to bring this up with your doctor

What to say. Lead with the facts that change the plan: "I have had [symptom] for [how long]. I am still eating gluten. Celiac disease runs in my family / I have type 1 diabetes / my iron has been low." The combination of symptom, gluten status and risk factor is what moves celiac testing up the list.

What to ask for. "Can we run tTG-IgA with a total IgA?" Total IgA is the part most often left off, and without it a negative result is not fully interpretable (2, 5). Then: "If it is positive, who do I see for the endoscopy?" and "If it is negative and I still feel like this, what is the next step?"

What to bring. Dates, not adjectives. A few weeks of what you ate and how you felt, any previous lab results including old iron and liver panels, your family history, and — if you have already cut gluten — exactly when you stopped and how strictly. That last detail decides whether a gluten challenge is needed and how long it has to be (7).

If you are told to just try the diet. It is reasonable to answer: "I would rather test first, because going gluten-free now makes both the blood test and the biopsy less sensitive, and I would have to do a gluten challenge later to get an answer." That is the AGA's position, not a patient opinion (2, 7).

If the first test is negative and nothing changes. Ask whether your pretest suspicion was high enough to warrant a biopsy anyway, and whether IgA deficiency was excluded (5). A second opinion from a gastroenterologist who sees celiac disease often is a normal request, not an escalation.

Red flags, the same wherever they appear in this article: see a clinician if diarrhea or digestive discomfort lasts more than two weeks, and sooner for unintentional weight loss, blood in your stool or black, tarry stools, vomiting blood, repeated vomiting or trouble swallowing (16). After an upper endoscopy, contact your doctor immediately for fever, trouble swallowing, throat, chest or abdominal pain, or bleeding including black stools, and call 911 for severe chest pain, heavy bleeding or trouble breathing (20).

Start understanding your body

Learn what affects your energy, stress, sleep, and daily state. Get the app.

Explore your data

How Welltory helps — and what it cannot do

The limits first, and they are absolute here. Welltory is a general wellness product, not a medical device. It does not diagnose, detect, predict, monitor, prevent, treat or mitigate any condition, and it cannot find celiac disease, confirm it, rule it out, or tell you whether a particular meal did anything to your gut. There is no heart-rate, HRV or sleep signal that stands in for tTG-IgA or a duodenal biopsy. Welltory's own help center also notes that measurements may be inaccurate for people under 18 and for pregnant people (19). We did not publish a chart of our own user data for this article, because no Welltory metric bears on celiac testing and a chart implying otherwise would be misleading.

What it can do is hold the record you are going to be asked for anyway. A celiac workup runs on dates: when symptoms started, when you last ate gluten, what happened in the weeks between. During a gluten challenge, clinicians specifically ask patients to report symptoms as they develop so the dose or duration can be adjusted rather than abandoned (7). A dated log is easier to hand over than a memory of a bad month.

You note the event; the app records the rest. Add a tag from the plus sign at the top of the feed — "started gluten challenge", "two slices of bread", "cramps all evening", "endoscopy". When the Today screen marks a stress or rest stretch, it asks "What happened?"; tap a suggested tag, type or talk, and correct the tags before saving. Resting heart rate, heart rate variability, stress minutes, Battery and sleep are recorded the same way every day, which is what makes two weeks comparable to the two before them (Battery and stress minutes need iOS with an Apple Watch or Oura; detailed sleep analysis needs an Apple Watch).

Over a few months, My Patterns shows what your tagged days have in common. My Patterns collects tags from stress and rest stretches on iOS with an Apple Watch or Oura, and patterns start appearing at around seven tagged events in a month — so a single test week will not produce one. Whatever it surfaces is an association, not a cause, and it is not evidence about gluten.

A record you made, for an appointment. The Journal (Premium) keeps your measurements, tags and notes, and from the web app you can export a CSV (Dashboard → choose a chart → Export). Bringing a dated page to a gastroenterologist is more useful than reconstructing three months on the spot.

And the one thing the diary must never do: it is not a test, and a bad stretch in it is not a reason to take yourself off gluten. Doing that before the workup is finished is the single action that can cost you the diagnosis (2).

Red flags, the same wherever they appear in this article: see a clinician if diarrhea or digestive discomfort lasts more than two weeks, and sooner for unintentional weight loss, blood in your stool or black, tarry stools, vomiting blood, repeated vomiting or trouble swallowing (16). After an upper endoscopy, contact your doctor immediately for fever, trouble swallowing, throat, chest or abdominal pain, or bleeding including black stools, and call 911 for severe chest pain, heavy bleeding or trouble breathing (20).

How we made it

Made with AI tools, then edited and fact-checked by the Welltory team. See our Editorial & AI policy.

Jane Smorodnikova, co-founder of Welltory, built the methodology behind how we read physiological data.

Written by Tatsiana Yashyna.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data