Celiac Testing: The Blood Tests, the Biopsy, and Why You Must Not Go Gluten-Free First

It usually starts with a well-meaning experiment. Someone feels bloated and exhausted, reads a few posts about gluten, empties the bread bin, and, feeling somewhat better, finally books a doctor’s appointment. By then, the single most useful piece of evidence may already be gone.

Celiac disease affects roughly 1 in 100 people worldwide, yet most who have it don’t know. Part of the problem is that its signature tests measure your immune system’s live reaction to gluten. Take gluten away, and the antibodies fade, the intestinal lining begins to heal, and the tests can come back reassuringly, and misleadingly, normal.

So before you change a single meal, it’s worth understanding how the testing actually works: what the blood panel measures, why the biopsy still matters for adults, what the Marsh stages mean, and how to get an honest answer the first time around.

Why you must not go gluten-free before celiac testing

Every celiac test in routine use depends on one thing: an immune system actively responding to gluten. The blood tests detect antibodies your body makes while it is attacking gluten fragments and its own tissue. The biopsy looks for the physical damage that attack leaves behind, flattened villi, inflamed lining, crowded immune cells.

Remove gluten, and both signals begin to disappear. Antibody levels typically fall within weeks to months, and the intestinal lining starts repairing itself. A person with genuine celiac disease who has been strictly gluten-free for a few months can test entirely normal. That’s not a cure; it’s a blind spot.

The consequences of a missed diagnosis are practical, not theoretical. A confirmed diagnosis unlocks follow-up monitoring for nutrient deficiencies and bone health, screening for first-degree relatives, and, in some countries, prescription or insurance support for the diet. Without it, you’re left guessing for life, and going back on gluten later to re-test (a formal “gluten challenge”) is unpleasant for many people.

The NHS and other major health bodies are unambiguous on this point: keep eating gluten-containing foods in more than one meal every day until all testing, including any biopsy, is finished. If you’ve already cut gluten, don’t panic, there’s a section on gluten challenges below, but do tell your doctor exactly when you stopped and how strictly.

What does a celiac blood test actually measure?

The first-line celiac blood test is a simple blood draw checked for tissue transglutaminase IgA antibodies, usually written as tTG-IgA. Tissue transglutaminase is an enzyme in your intestinal wall. In celiac disease, gluten fragments trigger the immune system to manufacture antibodies against this enzyme, effectively friendly fire against your own gut. Detecting those antibodies is strong evidence the disease process is underway.

Alongside it, most labs run a total IgA level. This matters more than it sounds. Roughly 2 to 3 percent of people with celiac disease have selective IgA deficiency, they make little or no IgA at all, which is several times more common than in the general population. In those people, a tTG-IgA test will read negative no matter how severe the disease is, simply because the measuring stick is missing. If total IgA is low, the lab switches to IgG-based versions of the tests instead.

Results usually come back within days as a number compared against a reference range. A strongly positive result, many times the upper limit of normal, carries different weight than a borderline one, which is one reason doctors want the actual figure, not just “positive.” No blood level, however high, is treated as a final diagnosis in adults on its own; it’s the ticket to the confirmatory step, the biopsy.

The full antibody panel, decoded

Depending on your results, age, and IgA status, a doctor may order more than the single tTG-IgA test. Each addition answers a specific question rather than simply piling on data.

Test	What it detects	When it’s used
tTG-IgA	IgA antibodies against tissue transglutaminase	First-line test for most people over age 2
Total serum IgA	Whether you produce normal IgA at all	Run alongside tTG-IgA to catch IgA deficiency
EMA (endomysial antibody, IgA)	Antibodies against connective tissue around muscle fibers	Confirms borderline tTG results; highly specific but labor-intensive
DGP-IgG / DGP-IgA	Antibodies against deamidated gliadin peptides	IgA-deficient patients and children under about 2
tTG-IgG	IgG version of the main test	Backup when total IgA is low

The EMA test deserves a note: it is read by a technician looking at a fluorescence pattern under a microscope, which makes it more expensive and somewhat observer-dependent, but its specificity approaches 99 percent. Labs typically reserve it as a tiebreaker.

One test you may see marketed, older “anti-gliadin” antibody tests, has largely been retired from adult diagnosis because too many healthy people test positive. If a direct-to-consumer kit leans on it, treat the result with caution and confirm through a clinical lab.

How accurate is the celiac blood test?

For a screening tool, tTG-IgA performs remarkably well, with the crucial caveat that the person is eating gluten. Studies summarized by the NIH’s digestive disease institute put its sensitivity around 90 to 98 percent (it catches most true cases) and its specificity around 95 to 99 percent (it rarely flags people without the disease).

Still, no test is perfect, and the gaps are worth knowing:

• False negatives happen with IgA deficiency, a reduced-gluten diet, very young children whose antibody responses are immature, and occasionally in people with milder intestinal damage.
• False positives occur more often at low, borderline antibody levels and can be associated with other autoimmune conditions, such as type 1 diabetes or autoimmune liver disease.

This is why the number matters. A tTG-IgA result ten times the upper limit of normal is very likely to reflect real celiac disease; a result nudging just past the cutoff warrants a confirmatory EMA test and careful interpretation rather than an immediate verdict.

It also explains the standing rule in adult medicine: blood tests screen, biopsies confirm. A positive antibody test tells you the odds are high. The biopsy tells you what is actually happening in the tissue, and gives you a baseline against which future healing can be measured. Skipping that step to start the diet early trades a few weeks of patience for a lifetime of uncertainty.

What are the warning signs of celiac disease?

The textbook picture, chronic diarrhea, weight loss, a bloated belly, is real but increasingly the minority presentation in adults. Celiac disease has earned its reputation as a great imitator because the damage it causes is fundamentally about malabsorption, and malabsorption can show up almost anywhere in the body.

Digestive warning signs include persistent diarrhea or, counterintuitively, constipation; bloating and gas; abdominal pain; nausea; and pale, foul-smelling stools. But watch equally for the signs that don’t point at the gut:

• Unexplained iron-deficiency anemia that doesn’t respond as expected to treatment, one of the most common adult presentations
• Persistent fatigue and “brain fog”
• Bone loss or fractures at a younger age than expected, from impaired calcium and vitamin D absorption
• Mouth ulcers and dental enamel defects
• Tingling or numbness in hands and feet
• An intensely itchy, blistering rash on elbows, knees, scalp, or buttocks, dermatitis herpetiformis, celiac’s skin manifestation
• In children: faltering growth, delayed puberty, irritability

Unexplained infertility and recurrent miscarriage have also been linked to undiagnosed celiac disease in some studies, which is why some clinicians include it in fertility workups.

The honest caveat: every symptom on this list has many other, often more common, causes. The point of the list isn’t self-diagnosis: it’s knowing when a celiac blood test is a reasonable thing to ask about, while you’re still eating gluten.

What do celiac stools look like?

People search this question constantly, and it deserves a straight answer. When celiac disease significantly damages the small intestine, fat stops being absorbed properly and passes through instead. The result is steatorrhea: stools that are pale or clay-colored, unusually bulky, greasy-looking, notably foul-smelling, and prone to floating or leaving an oily film that makes them hard to flush.

The mechanism is simple plumbing. Your small intestine’s villi, millions of finger-like projections that multiply its absorptive surface to roughly the size of a studio apartment’s floor, are where dietary fat gets taken up. Flatten those villi, as celiac disease does, and fat sails through to the colon, where bacteria ferment it into the characteristic odor and gas.

Two important honesty checks, though. First, steatorrhea is a sign of many conditions, including pancreatic insufficiency and gallbladder disease; it is not a celiac fingerprint. Second, and more importantly, many people with confirmed celiac disease have completely normal-looking stools. Some are constipated. Some have no digestive symptoms at all: the “silent” type discussed below. Studies suggest classic diarrhea is the presenting complaint in well under half of newly diagnosed adults today.

So use stool changes as one data point, not a verdict in either direction. Persistent changes lasting more than two weeks, especially with weight loss, or ever with blood, belong in a doctor’s office, not a search bar.

What are the four types of celiac disease?

Clinicians commonly describe four presentations, and knowing them explains why so many diagnoses take years.

• Classical celiac disease looks the way medical textbooks once described it: diarrhea, weight loss, malabsorption, and in children, poor growth. This is the easiest form to spot and, in adults, no longer the most common.
• Non-classical celiac disease presents with little or no diarrhea. Instead: anemia, fatigue, bone thinning, migraines, mouth ulcers, or neurological symptoms. Many adults in this group visit several specialists before anyone orders a celiac blood test.
• Silent (asymptomatic) celiac disease is found in people with positive antibodies and clear intestinal damage on biopsy who report feeling fine, often discovered when a relative is diagnosed and family screening begins. Interestingly, many “silent” patients only recognize how unwell they were after the diet improves their energy or mood.
• Potential (sometimes called latent) celiac disease describes positive blood antibodies with a normal or near-normal biopsy. These patients carry the genetic susceptibility and the immune activity but not yet the villous damage; some progress to full disease over time, so they’re usually monitored rather than dismissed.

You may also encounter the term refractory celiac diseasea rare situation where the intestine fails to heal despite a strict, verified gluten-free diet. It affects a small fraction of patients and requires specialist care, but it’s a category of disease course, not a fifth presentation type.

The genetic test: powerful for ruling out, weak for ruling in

Nearly everyone with celiac disease carries one of two gene variants: HLA-DQ2 (about 95 percent of patients) or HLA-DQ8 (most of the rest). These genes build the molecular “display cases” that present gluten fragments to immune cells; without them, the celiac reaction essentially cannot get started.

That biology gives the genetic test a lopsided usefulness. A negative result, no DQ2, no DQ8, makes celiac disease extraordinarily unlikely, now and in the future. Its negative predictive value exceeds 99 percent. For someone with ambiguous symptoms who has already been gluten-free for a year and dreads a gluten challenge, a negative gene test can close the question without one.

A positive result, though, tells you very little. Somewhere between 25 and 40 percent of the general population carries DQ2 or DQ8, yet only about 1 in 100 people develops celiac disease. Carrying the genes means the door is unlocked, not that anyone has walked through it. A positive genetic test never diagnoses celiac disease and never justifies starting a gluten-free diet on its own.

Where genetics earns its keep: people already on a gluten-free diet who can’t or won’t do a challenge, ambiguous biopsy results, and deciding which family members of a diagnosed patient need ongoing antibody screening versus a one-time all-clear. It’s a gatekeeping tool, and, used that way, a genuinely valuable one, since your genes don’t change no matter what you eat.

What actually happens during the celiac biopsy

The word “biopsy” alarms people out of proportion to the experience. Celiac confirmation uses an upper endoscopy: a thin, flexible tube with a camera passes through your mouth, past the stomach, into the duodenum: the first stretch of small intestine. You’ll typically receive sedation, the procedure itself takes about 15 to 30 minutes, and most people remember little of it. There are no incisions and no stitches.

Because celiac damage can be patchy, healthy tissue sitting beside flattened areas like a lawn with bald spots, guidelines call for multiple samples, usually four to six, including at least one from the duodenal bulb. Each sample is smaller than a grain of rice; the intestine doesn’t feel it and heals quickly.

A pathologist then examines the tissue under a microscope for three things: whether the villi are shortened or flattened, whether the crypts (the regenerative pits between villi) are enlarged from overwork, and whether immune cells called intraepithelial lymphocytes are crowding the surface layer. The findings get graded on the Marsh scale, covered in the next section.

Two practical notes. First, and by now you can predict it, you must still be eating gluten daily when the biopsy happens, or the tissue may have healed enough to look normal. Second, plan for someone to drive you home; sedation lingers longer than the procedure does. Serious complications are rare, and for adults, this remains the step that turns “probably” into a diagnosis you can build on.

The '5 stages of celiac': what Marsh scores really mean

Search for “5 stages of celiac” and you’re really asking about the Marsh classification: a pathologist’s grading system for biopsy tissue, running from 0 through 4. It describes damage severity, not how sick you feel or how long you’ve had the disease.

• Marsh 0: Normal tissue. Villi tall, immune cell counts ordinary. Celiac is unlikely, unless a gluten-free diet has already allowed healing.
• Marsh 1: Architecture still normal, but intraepithelial lymphocytes are increased. Suggestive, not specific; infections and some medications can produce the same picture.
• Marsh 2: Lymphocytes increased plus enlarged, overactive crypts. Uncommon as an isolated finding, often seen in dermatitis herpetiformis.
• Marsh 3: The hallmark of active celiac disease, villous atrophy, subdivided into 3a (partial flattening), 3b (subtotal), and 3c (total). Most diagnoses land here.
• Marsh 4: Rare, severe total atrophy with a flattened, hypoplastic lining, historically linked to refractory disease.

Two clarifications spare a lot of anxiety. A higher Marsh score does not mean worse symptoms, some Marsh 3c patients feel fine while some Marsh 1 patients feel dreadful, and studies consistently show a loose correlation between histology and misery. And the stages are not a one-way conveyor belt: on a strict gluten-free diet, most intestines heal, moving down the scale over months to a couple of years. The score is a snapshot, useful as a baseline, not a prophecy.

Already gluten-free? The gluten challenge, honestly explained

If you cut gluten before testing, you have three realistic paths, and it’s worth weighing them clearly rather than being talked into any of them.

Option one: test anyway, eyes open. If you’ve been gluten-free only a week or two, or your diet has been leaky, antibodies may still be detectable. A positive result is meaningful; a negative one proves nothing.

Option two: the gluten challenge. This means deliberately reintroducing gluten before testing. NHS guidance suggests eating gluten in more than one meal every day for at least six weeks before blood testing, roughly the equivalent of a couple of slices of ordinary bread daily, though your own doctor sets the specifics. Some research protocols use shorter challenges (as little as two weeks) when a biopsy is the endpoint, because tissue changes can reappear before antibodies fully rebound. The honest downside: people who genuinely have celiac disease often feel unwell during a challenge, which is exactly why finishing testing before the diet is so much kinder.

Option three: genetic testing first. A negative HLA-DQ2/DQ8 result makes a challenge pointless: you can stop worrying about celiac disease specifically. A positive result means a challenge is still the only route to a definitive answer.

What you should not do is stay in limbo, half-avoiding gluten and half-wondering. An unconfirmed self-diagnosis makes it harder to know how strict to be, whether relatives need screening, and whether your symptoms were ever gluten’s fault at all.

Who should get tested, even without gut symptoms

Celiac disease clusters in predictable places, and testing guidance reflects that. You have a solid case for asking about a celiac blood test if any of the following applies:

• A first-degree relative with celiac disease. Parents, siblings, and children of a diagnosed patient carry roughly a 1 in 10 risk, ten times the general population, and can have silent disease. Most experts recommend screening them even without symptoms.
• Type 1 diabetes. The two conditions share genetic roots; celiac disease appears in a meaningful minority of people with type 1 diabetes, often silently, which is why many diabetes clinics screen routinely.
• Autoimmune thyroid disease, autoimmune liver disease, Down syndrome, or Turner syndrome, all of which carry elevated rates.
• Unexplained findings: iron-deficiency anemia without an obvious source, early or unexplained osteoporosis, persistently elevated liver enzymes, or infertility without another identified cause.
• Dermatitis herpetiformisthe blistering rash is so tightly linked to celiac disease that a confirmed skin diagnosis (by skin biopsy) is generally treated as celiac disease of the skin.

What about screening everyone? Population-wide screening isn’t currently recommended in the United States; the evidence hasn’t shown that finding truly asymptomatic cases in low-risk people improves outcomes enough to justify it. Targeted testing of higher-risk groups, by contrast, has a strong rationale: the disease is common in those groups, the blood test is cheap and accurate, and the consequences of missing it accumulate quietly over decades.

Testing children: where the rules genuinely differ

Children follow the same first step, tTG-IgA with total IgA, while eating gluten, but with meaningful wrinkles worth knowing before a pediatric appointment.

Timing matters at the young end. Antibody responses are unreliable before roughly age 2, so for toddlers, labs often add DGP antibody tests, and doctors interpret everything more cautiously. Testing also can’t precede gluten itself: a child has to have been eating gluten-containing foods regularly for the tests to mean anything.

The biggest difference is the biopsy question. European pediatric guidelines (from the society known as ESPGHAN) allow a no-biopsy diagnosis in children when the tTG-IgA level is more than ten times the upper limit of normal, confirmed by a positive EMA test on a second, separately drawn blood sample. The logic: at those antibody heights, biopsy confirmation adds almost nothing, and sparing a child sedation and endoscopy is a real benefit. Practice varies, many US centers still prefer biopsy confirmation, and adults generally aren’t offered this shortcut because their false-positive patterns differ.

Watch for the pediatric-specific warning signs, which skew toward growth: falling off the growth curve, delayed puberty, a distended belly with thin limbs, irritability, and dental enamel defects on adult teeth. Children of a parent or sibling with celiac disease deserve screening even when they seem entirely well; guidelines generally suggest starting around age 2 to 3, once gluten has been in the diet for a while, sometimes with repeat testing later since the disease can develop at any age.

When to see a doctor

Some situations warrant an appointment rather than more reading. Book one, while continuing to eat gluten, if you have:

• Diarrhea, bloating, or abdominal pain persisting beyond two weeks
• Unintentional weight loss
• Stools that are persistently pale, greasy, and foul-smelling
• Unexplained fatigue plus any lab finding of iron-deficiency anemia
• An intensely itchy, blistering rash on elbows, knees, buttocks, or scalp
• A first-degree relative newly diagnosed with celiac disease
• A child who has stopped gaining weight or dropped across growth percentiles

Seek care promptly, same day or urgent care, for blood in the stool, black tarry stools, severe abdominal pain, persistent vomiting, or signs of dehydration. These need evaluation regardless of what’s causing them; none should wait for a celiac workup.

When you go, bring specifics. Doctors can act on “loose stools four to five times daily for six weeks, worse after pasta and sandwiches” far better than “my stomach’s been off.” Note any family history of celiac or autoimmune disease, any past anemia or bone density results, and, critically, whether you’ve reduced gluten and since when. That single detail changes how every test result should be read.

One request worth making explicitly: ask for a copy of your actual lab values, not just “normal” or “positive.” A borderline tTG-IgA and a sky-high one are different conversations, and you’re entitled to know which one you’re having.

After the tests: what a diagnosis, or a negative result, means

A confirmed diagnosis comes with a genuinely good prognosis and a clear job. The treatment is a strict, lifelong gluten-free diet, no medication required for the disease itself, and on that diet, most people’s antibody levels fall, symptoms ease over weeks to months, and the intestinal lining heals over months to a couple of years. Sensible early steps include a session with a registered dietitian (label-reading and cross-contact are skills, not instincts), baseline checks for iron, vitamin B12, vitamin D, and folate, a bone density scan where indicated, and screening blood tests for first-degree relatives. Follow-up antibody tests, typically at six and twelve months, track whether the diet is truly working.

A negative result deserves equal honesty. If you were eating gluten normally when tested, a negative tTG-IgA with normal total IgA makes celiac disease unlikely, but your symptoms still deserve an explanation, whether that’s irritable bowel syndrome, lactose intolerance, or something else worth pursuing. And a negative test today doesn’t grant lifetime immunity; celiac disease can develop at any age, so high-risk people (like relatives of patients) may need periodic re-testing.

Then there’s the gray zone: people who test negative yet feel better off gluten. Non-celiac gluten sensitivity is a real clinical description, though it has no biomarker and remains scientifically unsettled, some evidence points to other wheat components, such as certain fermentable carbohydrates, as the actual culprit. The practical takeaway holds either way: get the celiac question answered properly first, because that answer changes what you do next for the rest of your life.

Frequently asked questions

What are the warning signs of celiac disease?

Digestive signs include persistent diarrhea or constipation, bloating, gas, abdominal pain, and pale, foul-smelling stools. Just as important are the non-gut signs: unexplained iron-deficiency anemia, chronic fatigue, early bone loss, mouth ulcers, tingling in the hands or feet, and an itchy blistering rash called dermatitis herpetiformis. In children, watch for poor growth and delayed puberty. Many adults present without any digestive complaints at all.

What are the four types of celiac disease?

Clinicians describe classical (diarrhea, weight loss, malabsorption), non-classical (anemia, fatigue, bone loss, or neurological symptoms without prominent diarrhea), silent (positive tests and intestinal damage but no noticeable symptoms), and potential or latent (positive blood antibodies with a normal biopsy). A fifth term, refractory celiac disease, describes the rare situation where the intestine fails to heal despite a verified strict gluten-free diet.

What do celiac stools look like?

When celiac disease causes significant fat malabsorption, stools become pale or clay-colored, bulky, greasy, unusually foul-smelling, and may float or be hard to flush: a pattern called steatorrhea. However, many people with confirmed celiac disease have normal stools, and some are constipated, so appearance alone can neither confirm nor rule out the disease. Persistent stool changes lasting more than two weeks deserve a medical evaluation.

What are the 5 stages of celiac disease?

The “stages” usually refer to the Marsh classification pathologists use to grade biopsy tissue: Marsh 0 is normal; Marsh 1 shows increased immune cells with normal architecture; Marsh 2 adds enlarged crypts; Marsh 3 shows the hallmark villous flattening (subdivided into 3a, 3b, and 3c); and Marsh 4, which is rare, shows severe total atrophy. The score describes tissue damage, not symptom severity, and typically improves on a gluten-free diet.

Can I get tested for celiac disease if I've already stopped eating gluten?

Not reliably. Blood antibodies fall and the intestine heals on a gluten-free diet, so tests can read normal even in someone with genuine celiac disease. Your options are a gluten challenge, eating gluten daily, typically for at least six weeks before blood testing under medical guidance, or a genetic test first, since a negative HLA-DQ2/DQ8 result makes celiac disease so unlikely that a challenge becomes unnecessary.

How long do I need to eat gluten before a celiac blood test?

NHS guidance recommends eating gluten in more than one meal every day for at least six weeks before blood testing, roughly a couple of slices of ordinary bread daily. Shorter challenges, sometimes as brief as two weeks, are occasionally used when a biopsy is the planned endpoint, because tissue changes can reappear before antibodies fully rebound. Your own doctor should set the duration and amount for your situation.

Can a blood test alone diagnose celiac disease?

In adults, generally no: a positive blood test is confirmed with a small-intestine biopsy taken during an upper endoscopy, which also provides a healing baseline. Children are the notable exception: under European pediatric guidelines, a tTG-IgA level more than ten times the upper limit of normal, confirmed by a positive EMA test on a second blood sample, can establish the diagnosis without a biopsy. Practice varies between countries and centers.

Does a negative celiac test mean I'm in the clear forever?

No. A negative test is only trustworthy if you were eating gluten regularly at the time, and even then it reflects the present, not the future, celiac disease can develop at any age. People at elevated risk, such as first-degree relatives of patients or those with type 1 diabetes, may need periodic re-testing. The one durable all-clear is a negative HLA-DQ2/DQ8 genetic test, which makes future celiac disease extraordinarily unlikely.

What does a positive celiac genetic test mean?

Only that celiac disease is possible for you, not that you have it or will get it. Between 25 and 40 percent of the general population carries HLA-DQ2 or DQ8, yet only about 1 percent develops celiac disease. A positive genetic test should never trigger a gluten-free diet on its own. Its real power runs the other direction: a negative result rules out celiac disease with better than 99 percent confidence.

What is dermatitis herpetiformis and how is it tested?

Dermatitis herpetiformis is the skin form of celiac disease: an intensely itchy, blistering rash that typically appears symmetrically on the elbows, knees, buttocks, or scalp. It is diagnosed with a skin biopsy taken from just beside a lesion, looking for characteristic IgA deposits. Because the link to celiac disease is so strong, a confirmed diagnosis generally leads to celiac blood testing and the same lifelong gluten-free management.

References

• MedlinePlus: Celiac Disease Screening
• NIDDK (NIH): Diagnosis of Celiac Disease
• NHS: Coeliac Disease: Diagnosis
• Cleveland Clinic: Celiac Disease

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.