Celiac genetic testing: what HLA-DQ2 and DQ8 results can and cannot tell you

Short answer

Celiac genetic testing looks for the DQ2 and DQ8 gene variants1. Its informative result is a negative15: NIDDK says people without them are very unlikely to have celiac disease1. A positive does not diagnose it4: NIDDK puts the share of people with DQ2 or DQ8 at about 30 percent, and only about 3 percent of them develop celiac disease2.

Published by Gluten Verdict · updated · 8 sources, each linked below

Reviewed by Mark Peter, Registered dietitian (RD), on

Cardboard package on a doorstep beside yellow rain boots and wet leaves

Key points

  • A negative HLA-DQ2/DQ8 result is the informative one, in GIG's words15: NIDDK says people without the variants are very unlikely to have celiac disease, and ACG says that if the test is negative, celiac disease is ruled out1, 3.
  • A positive result does not diagnose celiac disease: NIDDK says most people with DQ2 or DQ8 do not develop it4.
  • No source we read recommends typing the general population: Coeliac UK says typing the general population is not recommended or currently practised, and ACG says the test is not required for diagnosis in all cases5, 3.
  • ESPGHAN's 2020 guideline made HLA typing a non-obligatory criterion in its no-biopsy route for children6.
  • People who started a gluten-free diet before testing are one place ACG names HLA typing: it calls it central, and says a gluten challenge can be offered if a compatible haplotype is found3.

What HLA typing reads

NIDDK describes the test this way: in some cases a health care professional may take a blood sample or use a cheek swab, and the sample is tested for groups of gene variants called DQ2 and DQ81. NIDDK says celiac disease almost always occurs in people who have one of the two groups7.

Coeliac UK, writing for healthcare professionals, names the types in more detail. In patients with celiac disease the commonest are DQ2.5 (highest risk), then DQ8 (moderate risk), then DQ2.2 (low risk, but not no risk)8.

NIDDK says doctors most often use blood tests and biopsies of the small intestine to diagnose or rule out celiac disease9, and that genetic tests are among the additional tests they may order in some cases10.

The method matters. ESPGHAN notes that SNP-based tests recognize the common variants (DQ2.5, DQ8, DQ2.2, DQ7.5), that in-depth allele typing is needed to identify rare variants, and that the accuracy of HLA testing for excluding celiac disease still depends on the method used11.

Reading a negative result

The sources below all read a negative result as pointing away from celiac disease. Each words it a little differently, so the table keeps each source's own wording and scope.

How each source words a negative HLA-DQ2/DQ8 result
SourceWhat it says about a negative result
NIDDK (US)People without the DQ2 and DQ8 variants are “very unlikely” to have celiac disease1.
ACG guideline, 2023 (US)“If negative, CD is ruled out.” ACG places the test in selected situations, such as a serology-histology discrepancy, meaning blood and biopsy findings that do not match3.
ESPGHAN guideline, 2020 (children)If a patient tests negative for HLA DQ2 and DQ8, “the risk of CD is very low”12.
WGO guideline, 2016HLA typing has a high negative predictive value, and an absence of HLA-DQ2/-DQ8 excludes the presence of celiac disease in susceptible individuals13.
Coeliac UK (UK)Being negative “would be very reassuring”, though “never a guarantee”, and there are “very rare patients” who carry none of DQ2.5, DQ8 or DQ2.214.
Gluten Intolerance Group (GIG)GIG says negative results tell a person that it is “virtually impossible” that they have, or will develop, celiac disease15.

Where the sources put a number on the exceptions, it is small. Coeliac UK says 99.6% of patients with celiac disease carry DQ2.5, DQ8 or DQ2.216. WGO says fewer than 1% of patients lack DQ2 and DQ817. ESPGHAN cites an Australasian celiac disease population of 356 in which 99.6% were DQ2/DQ8-positive, and says it may be concluded that the higher percentage of DQ2/DQ8-negative patients in earlier publications (up to 5%) had several causes, one being the HLA testing method18.

A negative result speaks to the genes celiac disease almost always occurs with7. It does not speak to non-celiac gluten sensitivity: GIG says the test does not shed any light on whether a person has it19.

Reading a positive result

NIDDK is direct about it: having DQ2 or DQ8 alone does not mean you have celiac disease, most people with these gene variants do not develop it, and a doctor may recommend additional tests to check for or rule out celiac disease4.

The reason is how common the gene types are. The sources give different figures, so the table reports each as its owner states it.

The gene types and celiac disease, in each source's own figures
SourceGene types in the general populationCeliac disease, as the source puts it
NIDDK (US)About 30 percent of people have DQ2 or DQ82Only about 3 percent of people with DQ2 or DQ8 develop celiac disease2
WGO, 201630–35% in the general population20Prevalence varies between roughly one in 100 and one in 300 in most parts of the world21
Coeliac UK (UK)DQ2 and DQ8 haplotypes found in up to 40%22Estimated prevalence in the UK is 1%22
GIGAbout a third of the general population carries the genes15Only one percent develops celiac disease15

Coeliac UK says that for a patient with suspected celiac disease who is not responding to a gluten-free diet, a positive result does not diagnose celiac disease and is “not very helpful”, because so many people in the general population are DQ positive without having it23. For a child with no symptoms and an at-risk condition such as type 1 diabetes, it says a positive result only shows the child has the type that may permit celiac disease; antibody testing matters if the child is eating gluten, and if the antibody test is also positive, celiac disease is likely and some children need an endoscopy and biopsy to confirm24.

Gluten stays part of the picture. WGO says there is no celiac disease without gluten, even in the presence of a genetic predisposition25, and NIDDK says not all people who have DQ2 or DQ8 and eat gluten develop the disease26.

Where the guidelines place the test

No source we read recommends HLA typing for the general population. Coeliac UK says typing the general population is not recommended or currently practised, and that practice is to use it in specific at-risk scenarios, primarily looking to exclude the condition5.

Where each source places HLA typing
SourceWhere it places the test
NIDDK (US)Genetic tests are among the additional tests doctors may order “in some cases”10.
ACG guideline, 2023 (US)Not required for diagnosis in all cases; may be helpful in selected situations such as a serology-histology discrepancy; central to the approach for people who started a gluten-free diet before evaluation3.
ESPGHAN guideline, 2020 (children)Not required in a patient with positive TGA-IgA who qualifies for diagnosis with biopsies, or who has high TGA-IgA (10 times the upper limit of normal or more) and positive EMA-IgA12; may be useful in other circumstances27.
WGO guideline, 2016Its list of situations names exclusion of celiac disease, an uncertain diagnosis (negative serology but suggestive histology, where typing helps only if negative), siblings, and people with other autoimmune diseases or some genetic disorders28. It also says genotyping may be important in seronegative disease and before a formal gluten challenge29.
Coeliac UK (UK)Not routinely recommended in the diagnosis pathway, but it can have an important role in ruling out people who will not likely (or never) need future testing, and may allow a risk assessment for those with the appropriate HLA types30.

Why ESPGHAN no longer requires HLA typing in the children’s no-biopsy route

ESPGHAN's 2012 recommendations for a no-biopsy approach included testing for HLA-DQ2 and DQ8 in individuals with very high TGA-IgA titres and EMA-IgA positivity31. The 2020 guideline says HLA DQ2/DQ8 determination, like symptoms, is not an obligatory criterion6.

The guideline gives its reasons. It reviews one study of 645 paediatric patients in which 399 met the no-biopsy criteria, all 399 were positive for DQ2 and/or DQ8, and it was concluded that HLA typing did not add to the certainty of the diagnosis in these patients32. It also says HLA testing is not universally available and is quite costly in some countries, so if it does not improve the no-biopsy diagnosis it should be omitted33.

Coeliac UK reported the change as genetic testing no longer being required for children with high antibody levels34. It quotes Dr Peter Gillett, a consultant paediatric gastroenterologist: “Diagnosis can now be based on two antibody blood tests without the need for genetic testing”35. ACG's 2023 guideline says the updated 2020 ESPGHAN guidelines proposed that HLA testing is no longer necessary36.

The change is about the antibody-based route for children with high antibody levels. It does not take HLA typing out of the situations in the previous section: ESPGHAN's statement still says testing for HLA DQ2 and DQ8 may be useful in other circumstances27.

Relatives and at-risk groups

NIDDK says blood relatives of people with celiac disease, and people with type 1 diabetes, should talk with their doctor about their chances of getting the disease to see if they should be tested37.

Two sources put figures on families. GIG says prevalence in first-degree relatives ranges from 4% to 16%, and is about 2.5% in second-degree relatives38. WGO says approximately 70% of first-degree relatives are HLA-DQ2- and/or HLA-DQ8-positive, against 30–35% in the general population20.

The sources differ in how they bring genes in. WGO's section on first-degree relatives includes the line “HLA typing for all first-degree relatives; do no further work-up on those who are negative for HLA-DQ2/8”39. Coeliac UK says HLA DQ2 and 8 typing is suggested in guidance as helpful in excluding people from future testing in at-risk groups, with first-degree relatives and Down syndrome as examples40. ESPGHAN says that for some at-risk groups, such as type 1 diabetes, typing may not be cost-effective because of the high percentage of HLA positives, but may be in others, for example first-degree relatives or Down syndrome41.

For parents who have celiac disease and want to know whether their children could develop it, GIG, in a 2019 FAQ, says positive genetic tests mean monitoring should be ongoing and negative tests mean no further screening is necessary42. ACG's guideline separately says patients with a first-degree family member who has a confirmed diagnosis should be tested for celiac disease43. How a family combines gene and antibody tests is a decision for their clinician.

When gluten has already been dropped

NIDDK says doctors do not recommend starting a gluten-free diet before diagnostic testing, because the diet can affect test results9. WGO says celiac disease should be excluded by a negative test for celiac-specific antibodies before anyone starts the diet for any reason44. GIG says testing before giving up gluten is always advisable, but that this does not always happen45.

For people who did stop first, the sources name HLA typing. ACG calls it central to the approach, and says that in the presence of a compatible haplotype a gluten challenge can be offered3. WGO says genotyping may be important in seronegative disease and to exclude celiac disease before a formal gluten challenge29. Coeliac UK describes a patient who went gluten-free without a diagnosis and then cannot tolerate reintroducing gluten, or starts a challenge, where the medical team performs HLA typing on clinical suspicion, and says some patients choose a commercial test and look for an explanation of the results46. For a patient with suspected celiac disease who is not responding to a gluten-free diet, it says that if the patient tests negative for DQ2.5, DQ8 or DQ2.2, it could be stated with confidence that the patient does not have celiac disease47.

A positive result in this situation leads to the next step, not to a diagnosis: ACG's text offers a gluten challenge when a compatible haplotype is found3.

Who orders it and what the report shows

The only source we read that describes ordering is Coeliac UK, and it describes the UK:

  • Typing is widely available, cost varies considerably from laboratory to laboratory, and the format of the reporting is not standardised48.
  • Such testing generally goes through secondary care, by paediatricians or gastroenterologists with an interest in celiac disease or by specialist dietitians48.

A result sheet can be read against those points: which of DQ2.5, DQ8 and DQ2.2 was reported8, which method the lab used11, and whether the result is being used to rule out or to assess risk30. For the difference between a consumer genetic test and clinical HLA typing, see 23andMe vs clinical HLA typing; for the antibody tests, see the celiac blood test guide.

Questions people ask

Does a positive celiac gene test mean I have celiac disease?

No. NIDDK says having DQ2 or DQ8 alone does not mean you have celiac disease, and that most people with these gene variants do not develop it4.

Can I have celiac disease with a negative gene test?

The sources treat it as very unlikely: NIDDK says “very unlikely”1 and ACG says celiac disease is ruled out3. Coeliac UK notes there are very rare patients who carry none of DQ2.5, DQ8 or DQ2.214.

Is genetic testing needed to diagnose celiac disease?

Not in every case. ACG says it is not required for diagnosis in all cases3, ESPGHAN says HLA typing is not an obligatory criterion for its no-biopsy diagnosis in children6, and Coeliac UK says it is not routinely recommended in the diagnosis pathway30.

Does HLA typing show gluten sensitivity?

GIG says non-celiac gluten sensitivity is not thought to have a genetic component49, and that the test does not shed any light on whether a person has it19.

More in Celiac genetics and family risk

Sources