How Celiac Disease Is Diagnosed Without Guesswork

Celiac disease is diagnosed with a blood test for tissue transglutaminase IgA (tTG-IgA), drawn together with a total IgA level, while you are still eating gluten every day. In most adults the result is then confirmed with a biopsy of the small intestine. The single most common reason the workup fails has nothing to do with the labs themselves: it is testing someone who has already cut out gluten, which can turn a genuine case into a normal-looking result. Ordered in the right sequence, the diagnosis leaves very little room for guesswork.

Key points#

• The 2023 American College of Gastroenterology (ACG) guideline names one starting test: tTG-IgA, drawn with a total IgA.
• Every antibody test and every biopsy only works while gluten is still in the diet.
• Most adults confirm the diagnosis with a duodenal biopsy; very high antibodies can support a no-biopsy path in specific situations.
• Genetic testing (HLA-DQ2/DQ8) can help rule celiac disease out, but a positive result does not rule it in.

Start with tTG-IgA, but never alone#

For almost anyone with symptoms or a risk factor that raises suspicion, the ACG guideline puts tTG-IgA first. It is sensitive and specific, which makes it a strong single screen. The catch is in the name: it measures an antibody of the IgA class, and it can only climb if the body makes IgA to begin with.

That is why a total IgA level always travels with it. Selective IgA deficiency is more common in people with celiac disease than in the general population, so if you produce little IgA your tTG-IgA can read low whether or not you have the disease. When the total IgA comes back low, the guideline switches to IgG-based tests, such as deamidated gliadin peptide IgG or tTG-IgG, which do not depend on the missing antibody class. Ordering both numbers on the first draw is what stops a false negative from slipping past unnoticed.

Keep gluten in the diet, or the tests lose their meaning#

This is the load-bearing condition of the entire workup. The antibodies the blood tests detect and the intestinal damage the biopsy grades are both products of the immune reaction to gluten. Take gluten away and the antibodies fall over the following months while the lining of your small intestine starts to heal. The findings the tests are designed to catch simply fade, and there is nothing the laboratory can do about it.

The NIH's National Institute of Diabetes and Digestive and Kidney Diseases puts it plainly: a gluten-free diet should not be started before testing, because it can change the results. This is where a well-meant trial of eating gluten-free causes the most confusion. If you feel better without gluten and then ask to be tested, you may show normal antibodies and a normal biopsy while genuinely having celiac disease.

If you have already stopped, the guideline describes a gluten challenge before retesting: adding daily gluten back in, roughly one to two slices of wheat bread per day, over a stretch of weeks so the immune signal has time to reappear. The exact length is tailored to the person, and anyone with a strong past reaction is best challenged under medical supervision.

What the biopsy still settles#

Upper endoscopy with several biopsies of the duodenum remains the confirmation step for most adults, and it does jobs that a blood test cannot. It grades how much of the intestinal lining (the villi) has flattened, it can catch seronegative disease in a person whose antibodies sit in a gray zone, and it helps separate celiac disease from other conditions that look similar under the microscope. Like the blood work, the biopsy is only trustworthy while you are still eating gluten.

When the guideline allows skipping the biopsy#

The rules differ by age. For children, the ACG guideline supports a no-biopsy diagnosis when the tTG-IgA is very high, at or above ten times the upper limit of normal, and a second, separate blood sample is positive for endomysial antibody (EMA). Requiring a second sample and a second antibody guards against a mislabeling or laboratory error on a diagnosis that commits a child to a lifelong diet.

For adults, biopsy stays the standard, with one conditional exception: in an adult who is unwilling or unable to have an endoscopy, the same pairing (tTG-IgA at or above ten times the upper limit of normal plus a positive EMA on a second sample) supports a diagnosis of likely celiac disease. Two safeguards keep this from sliding into guesswork. The ten-times threshold is anchored to a particular laboratory's own upper limit of normal, so the number only means something in the context of the assay that produced it. And the total IgA still has to be adequate, or the tTG-IgA cannot be trusted at all. A 2024 analysis in Nutrients on how total IgA levels shape non-biopsy diagnosis makes the same point: the antibody value cannot be read in isolation.

Genetic testing rules out, it does not rule in#

HLA-DQ2 and HLA-DQ8 genetic testing is often misunderstood. Nearly everyone with celiac disease carries one of these gene variants, so their absence is genuinely useful: it makes the diagnosis very unlikely and can help close the question in ambiguous cases. Their presence, however, proves little. A large share of the general population carries these genes and never develops celiac disease. Genetic testing is a tool for ruling the diagnosis out, not for confirming it.

Two habits that reintroduce guesswork#

Even a clean workup can be undone by two common shortcuts. The first is diagnosing celiac disease from symptoms plus feeling better on a gluten-free diet, with no serology or biopsy. The evidence does not support this, and it forecloses accurate testing later, because you are now off gluten. The second is over-reading genetic results, treating a positive HLA-DQ2 or DQ8 as if it confirmed the disease when it only fails to exclude it.

Celiac disease sits at the intersection of immunology, genetics, and population epidemiology, which is one reason its diagnostic pathway rewards a careful, ordered approach over pattern-matching. Kept in sequence, the workup is reliable: test on gluten, draw tTG-IgA with a total IgA, confirm by biopsy in most adults, and use genetics only to help close the door, never to open it.