A new study has found that a meaningful share of people who believed they had celiac disease likely never had it — because they lack the genetic markers the condition requires to develop. At the same time, thousands of adults with real, active celiac disease had no idea they were sick. The findings, published in Clinical Gastroenterology and Hepatology by Hjort, Andersen, and Lukina, come from one of the largest celiac screening projects ever conducted: the HUNT4 study, which tested 52,588 Norwegian adults.
Researchers identified 846 total cases — 376 who already carried a celiac diagnosis, 465 newly confirmed through biopsy, and 163 potential cases. Then they looked at the DNA. What they found raises hard questions about diagnostic accuracy, both for people who were told they have celiac and for the far larger group who were never told at all.
What This Means for You
Every confirmed celiac case must carry specific genetic variants — either HLA-DQ2 or HLA-DQ8. These genes do not cause celiac disease on their own, but the disease cannot develop without them. They are a necessary condition, not a sufficient one.
All 465 people newly identified through screening carried DQ2 or DQ8. That is exactly what the science predicts. But among the 376 people who said they already had a celiac diagnosis, 21 tested negative for both variants. Since those genes are required for celiac to develop, their absence is strong evidence that these individuals do not — and never did — have celiac disease.
Researchers then re-examined intestinal biopsy tissue from those 21 people. Only 5 had findings consistent with a real celiac diagnosis. The other 16 were most likely misdiagnosed at some point in the past.
That figure — roughly 5.6% of previously diagnosed cases — may look small. But for any person who spent years on a strict gluten-free diet, convinced they had celiac, only to learn the diagnosis was probably wrong, it represents something significant: their actual condition was never identified. Whatever was causing their symptoms went untreated.
For families navigating a celiac diagnosis, this research raises a practical question: how confident are we in the original workup? For anyone diagnosed before genetic testing became standard practice — or based primarily on symptoms and older, less sensitive blood tests — a conversation with a gastroenterologist about confirming the diagnosis is worth having. This does not mean abandoning a gluten-free diet without medical guidance. It means making sure the diagnosis rests on solid ground.
The flip side of this study is equally striking. The HUNT4 screeners found 465 people with biopsy-confirmed celiac disease who had never been diagnosed. Celiac families know this story too — years of unexplained symptoms, dismissed or attributed to something else, while intestinal damage quietly accumulates. Population-level research on celiac prevalence has long suggested that most celiac disease goes undetected, and this study puts hard numbers on that gap in a large, well-characterized population.
Key Takeaways
- About 5.6% of people with a prior celiac diagnosis lacked the genetic markers required for the disease — pointing to historical misdiagnosis.
- All newly identified celiac cases in this large screening study carried either the DQ2 or DQ8 genetic variant, confirming these genes are necessary for celiac to develop.
- Carrying two copies of the DQ2.5 variant carries the highest celiac risk — more than 16 times the odds of those without risk genes.
- Hundreds of real celiac cases were discovered in people who had never received a diagnosis, underscoring how many patients remain undetected.
- If an existing celiac diagnosis was never confirmed with HLA genetic testing and intestinal biopsy, it may be worth asking a doctor about re-evaluation.
The Science
Want to understand how this actually works? We’ll walk you through the technical details below and define every term. No medical degree required.
What Are HLA-DQ Allotypes?
HLA stands for human leukocyte antigen — a family of proteins on cell surfaces that help the immune system distinguish self from threat. The HLA-DQ proteins specifically present fragments of digested food to immune cells. In most people, gluten passes through this process without incident. In people with specific HLA-DQ variants, gluten fragments get flagged as dangerous, triggering the immune attack that defines celiac disease.
The variants relevant to celiac are HLA-DQ2.5, HLA-DQ2.2, and HLA-DQ8. These are called allotypes — specific forms of the HLA-DQ protein encoded by different versions of the HLA-DQA1 and HLA-DQB1 genes. Prior work, including an analysis of how HLA-DQ5 and DQ9 present gliadin fragments to immune cells, has expanded understanding of which variants can trigger the immune cascade and how different combinations shape risk.
About 30–40% of the general population carries at least one of these variants. Only roughly 1–3% of the population develops celiac disease. Having the gene is necessary but not sufficient — other factors, including environmental exposures, gut microbiome composition, and early-life infections, also appear to push a genetically susceptible person toward the disease.
The Gene-Dose Effect
One of the more important findings in HUNT4 is the gene-dose effect: carrying two copies of a risk gene raises risk considerably more than carrying one copy.
The study measured this using odds ratios (OR) — a way of expressing relative risk. An OR of 16.7 means people with that genetic combination are more than 16 times more likely to have celiac than people with no risk genes. The hierarchy in the HUNT4 data:
- DQ2.5/DQ2.5 (two copies of DQ2.5): OR 16.7 — highest risk
- DQ8/DQ8 (two copies of DQ8): OR 8.8
- DQ2.5/DQX (one copy of DQ2.5, any other variant): OR 5.7
- DQ8/DQX (one copy of DQ8): OR 2.6
- DQ2.2/DQ2.2: OR 3.4 (wide confidence interval — estimate is uncertain)
- DQ2.2/DQX (one copy of DQ2.2): OR 0.8 — essentially no increased risk
That last number matters. DQ2.2 alone, without DQ2.5 on the paired chromosome, does not appear to meaningfully elevate celiac risk. This distinction is clinically relevant: having a DQ2 gene does not automatically signal elevated risk unless that gene is confirmed as DQ2.5 rather than DQ2.2.
Why Were People Misdiagnosed?
In HUNT4, 21 people with prior celiac diagnoses tested negative for both DQ2 and DQ8. When researchers performed histological reassessment — re-examining intestinal biopsy samples under a microscope — they confirmed celiac in only 5 of those 21. The remaining 16 had most likely been misdiagnosed.
The reasons this happens are not hard to understand. Historically, celiac diagnosis relied heavily on symptoms and less sensitive serological tests. Older biopsy criteria were less standardized. The clinical picture — fatigue, gastrointestinal distress, nutrient deficiencies — overlaps with many other conditions: non-celiac gluten sensitivity, irritable bowel syndrome, inflammatory bowel disease. Before HLA genetic testing became routine, distinguishing between these possibilities was harder, and some diagnoses did not hold up.
The HUNT4 data also showed that among newly identified cases, the odds ratios for DQ2.5 and DQ8 were somewhat higher than among previously diagnosed cases. This pattern may partly reflect historical misdiagnosis diluting the genetic signal in the known-cases group, or it may reflect that people who sought a clinical diagnosis for symptoms — without genetic testing — were a more heterogeneous group than those identified through rigorous population screening.
The DQ8 Finding in a Nordic Population
In Norway, DQ2.5 dominates as the primary celiac risk gene — 87.8% of confirmed cases in HUNT4 carried it. What drew researchers’ attention was the proportion of DQ8 cases: 8.5% of confirmed cases were DQ8-positive but DQ2.5-negative. For a Nordic population, where DQ2.5 typically dominates, that proportion is relatively high.
This matters because DQ8-associated celiac can present differently and may be underrecognized in clinical settings where DQ2.5 is the primary focus. Earlier research on HLA risk alleles across different populations has shown that the distribution of celiac-risk genes varies meaningfully by ancestry — and that no single allele tells the whole story. The HUNT4 data reinforces that point with a large, well-powered dataset.
As the parent of a child with celiac disease, I take diagnostic accuracy seriously. My son’s diagnosis shapes the terms of his daily life — every meal, every school trip, every birthday party. If there is any doubt that a diagnosis is correct, that is worth resolving. And for the hundreds of people in this study who were living with undetected celiac disease, the damage was accumulating without any of the protections a diagnosis would have triggered.
The message is not to panic, and it is not to discard a gluten-free diet without medical guidance. The message is that diagnosis should be built on solid evidence — HLA genetic testing and biopsy confirmation — and that when doubt exists, the right move is to revisit with a specialist. If you have concerns about a celiac diagnosis for yourself or a family member, bring this research to your gastroenterologist and ask about HLA testing.
Related Coverage
- HLA-DQB1*02, HLA-DRB1*03 alleles and HLA-DQB1*02-HLA-DRB1*03 haplotype confer a significant risk of celiac disease in the Mauritanian population
- In Silico Analysis of Functional Overlap and Complementarity of HLA-DQ5 and DQ9 in Gliadin Presentation: Expanding the Immunopeptidomic Landscape of Celiac Disease
- How Common Is Celiac Disease? What the Major Prevalence Studies Tell Us
References
Hjort R, Andersen IL, Lukina P. “HLA-DQ allotypes in adults screened for celiac disease reveal historical misdiagnosis — results from the HUNT study.” Clinical Gastroenterology and Hepatology: the official clinical practice journal of the American Gastroenterological Association. Published August 16, 2026. PubMed