HLA-DQ2/DQ8 Testing — Body Handbook

Screening · §123

HLA-DQ2/DQ8 Testing

If you've been off gluten for years without a firm celiac diagnosis, one blood draw can end the question — and unlike every other test in a celiac workup, this one doesn't require you to go back to eating bread. HLA-DQ2/DQ8 typing reads two immune-system genes that are required for celiac disease to develop; if you don't carry either, you almost certainly cannot have the disease, full stop. The test rules celiac out with near certainty and keeps the diagnosis alive when standard tests can't be done. It earns its place in three situations: you're already on a gluten-free diet, your antibody or biopsy results are murky, or a close relative has the disease and you want a one-time answer.

Test · Once Evidence Strong Chapter Screening

Negative result, no celiac — for life. That's the headline, and the reason this test exists. A positive result is the other half of the deal: it doesn't mean you have celiac, only that you could; you'll still need antibody testing on a gluten-containing diet to confirm. One draw, no fasting, no follow-up, usually $200–$450 in the US and often covered when a doctor orders it. The catch: a positive result without context can scare a healthy person into a gluten-free diet they don't need. Read the result with a doctor.

Celiac disease isn't a wheat allergy and isn't the same as feeling bad after a sandwich. It's an autoimmune reaction in the lining of your small intestine, and your immune system can only mount that reaction if it has the right molecular machinery to grab gluten fragments and show them to T cells. That machinery is built by one of two specific genes: HLA-DQ2 or HLA-DQ8. About 30 to 40 percent of people of European ancestry carry one — and most of them never develop celiac — but more than 99 percent of confirmed celiac patients carry at least one Karell et al. 2003. Without DQ2 or DQ8, your immune cells have nothing to grab the gluten with. That's why a negative test rules celiac out so cleanly: it isn't a probability play, it's a missing part.

This also explains why having the gene isn't enough on its own. Most carriers eat gluten their whole lives without consequence. Something else — most likely some combination of gut infection, microbiome shift, and timing of gluten introduction in infancy — has to push the immune system over the line. The gene is the gate; something else turns the key. That asymmetry is the whole reason this test is used to rule out rather than rule in.

How settled this is

The gene–disease link is one of the cleanest and most replicated associations in autoimmune medicine. Decades of pooled cohorts converge on the same number: if you don't carry HLA-DQ2 or HLA-DQ8, your chance of having celiac is essentially zero. A formal diagnostic-accuracy study put the negative predictive value at nearly 100% in a referral population — the negative result is as final as any clinical test gets Hadithi et al. 2007.

Risk in carriers isn't uniform, and good genetic labs report enough detail to tell the difference. The strongest risk allele is called DQ2.5; two copies of it produce the highest absolute risk. In a prospective study of more than 6,000 children with celiac-risk genes, kids with two DQ2.5 copies developed celiac autoimmunity at about 1 in 4 by age 5, compared to roughly 1 in 30 for the lowest-risk DQ8 carriers Liu et al. 2014. The clinical use of that detail: a DQ2.5 homozygous child of celiac parents is someone you screen periodically with antibody tests; a DQ8 single-copy adult relative is someone you can mostly leave alone unless symptoms appear.

Both major specialty guidelines endorse the test for the rule-out use case. The European pediatric guideline (ESPGHAN 2020) and the American adult guideline (ACG 2023) name the same situations: patients already on a gluten-free diet, equivocal antibody or biopsy results, and selected family-screening scenarios Husby et al. ESPGHAN 2020 Rubio-Tapia et al. ACG 2023.

What's at stake if the question stays open

Two failure modes — and most people are sitting in one of them right now without knowing it.

The first: undiagnosed celiac. You stay on bread, pasta, beer. The gut lining slowly flattens. Iron drops because you can't absorb it; ferritin lab values stay stubbornly low no matter how many supplements you take. You are tired in a way that doesn't track to your sleep. Periods get irregular. Some pregnancies don't take. Bone density drops earlier than it should. A dentist notices enamel defects you weren't told to look for as a kid. Two or three decades in, your risk of small-bowel cancer and a specific lymphoma has crept up enough to bend overall mortality slightly above average Lebwohl et al. 2018. None of it is dramatic in any given year; it's a slow tax on the body that adds up.

The second is the mirror image: you went off gluten eight years ago because gluten felt bad, never got a proper diagnosis, and have been buying gluten-free bread at twice the price since then. Restaurant meals are a project. Dinner at a friend's apartment requires a phone call ahead. You read labels on canned soup. You assume, by default, that anything good was made with wheat. If you don't actually have the gene, none of this is necessary, and the years of avoidance bought you nothing.

The HLA test is the cheapest, lowest-friction way to find out which of these you're in.

How to order it

Ask your primary care doctor or a gastroenterologist for "HLA-DQ2/DQ8 typing for celiac disease." One blood draw or, in some labs, a cheek swab. No fasting, no prep, results in roughly one to two weeks. Pick a lab that reports the underlying allele combinations and not just a "DQ2 positive / DQ8 positive" line — the difference between DQ2.5 and DQ2.2 matters for how worried anyone should be, and the cheap reports don't tell you which one you have.

One important sequencing note. If you are currently eating gluten and don't have a reason the test is faster than antibodies, do the antibody test first instead. It's cheaper, it directly measures the disease (not just the predisposition), and a positive antibody is far more actionable than a positive gene. The HLA test earns its place when the antibody pathway is blocked — usually because the patient is already gluten-free.

Who specifically benefits

This test pays off most for four groups. Outside these, the antibody test is usually a better first move.

Anyone already on a gluten-free diet without a firm diagnosis. This is the test's headline use case. Once you've been gluten-free for more than a few weeks, antibody levels normalize and biopsy results soften — the standard celiac workup quietly stops working. The choice becomes either a long gluten challenge (six to eight weeks of eating bread on purpose, which a lot of people who once felt awful on gluten flatly refuse) or the gene test. If the gene test is negative, you're done forever. If positive, the question is still open, and a doctor can help you decide whether the gluten challenge is worth it.
First-degree relatives of someone with celiac. Pooled data put celiac prevalence in the parents, siblings, and children of a confirmed patient at roughly 1 in 13 — about ten times the general-population rate, and siblings carry the highest risk Singh et al. 2015. A one-time HLA test in each relative either ends the worry permanently (negative) or flags them for periodic antibody screening (positive).
People with type 1 diabetes, Down syndrome, Turner syndrome, or autoimmune thyroid disease. All of these conditions share genetic territory with celiac. Celiac prevalence runs 3 to 12 percent in these groups depending on the condition, and clinical guidelines recommend either antibody screening or HLA-based risk stratification Rubio-Tapia et al. ACG 2023. The HLA test identifies the subset for whom future antibody monitoring is meaningful versus the subset who can be left alone.
Anyone with antibody or biopsy results that don't fit the picture. A weakly positive antibody, a biopsy that looks suspicious but not classic, a child whose serology and symptoms disagree — these are the muddy cases where the gene test pulls in the right direction. A negative HLA in this setting almost always means celiac isn't what's going on, and the workup should turn elsewhere.

What about just doing the antibody test?

The first-line test for celiac in anyone eating gluten is not the gene test. It's a blood test for tissue transglutaminase antibodies (tTG-IgA), and it's cheaper, faster, and more directly diagnostic — both sensitivity and specificity above 95% in patients on a normal diet Lebwohl et al. 2018. If you're already eating gluten and willing to keep eating it for another month or two, do that test first. The HLA test isn't a replacement; it's a tool for the cases where antibody testing has stopped being an option.

The historical reference standard is the small-bowel biopsy. A gastroenterologist looks at duodenal samples under a microscope and grades the damage to the gut lining — flattened villi, dense immune-cell infiltration, the patterns clinicians call Marsh 2 or Marsh 3 Bao and Bhagat 2012. The biopsy also requires you to be eating gluten, and adds an endoscopy under sedation to the workup. It remains the tiebreaker when antibodies and clinical picture disagree, but it isn't where the workup starts.

And then there's the alternative most people skip: a formal gluten challenge. If you're gluten-free and want to know for certain, the rigorous version is eating about three slices of regular bread daily for six to eight weeks, then repeating antibodies and biopsy. Two weeks at a slice and a half daily is the bare minimum that produces detectable changes Leffler et al. 2013. A lot of people who suspect celiac have a strong physical reason they went gluten-free in the first place and refuse this option outright. The HLA test exists in part for them.

What gets confused

Three things are worth being clear about, because the wrong reading of this test has real consequences.

A positive result is not a celiac diagnosis. Roughly one in three people of European ancestry carries one of these genes and never gets the disease. Going gluten-free on the basis of a positive HLA test alone — without antibodies, without symptoms, without context — buys you nothing and costs you the rest of your eating life. The number that matters in the positive direction is the antibody titer, not the gene.

The gene test cannot replace the rest of the workup in adults. A confident celiac diagnosis still rests on antibodies, biopsy, or the symptom-and-serology pattern that meets the European pediatric short-cut rule. The gene test never confirms; it only fails to rule out.

The role in children has shrunk, not grown. The 2020 European pediatric guideline actually removed the gene test from the standard no-biopsy diagnostic pathway in kids. Today, a child with a very high tissue transglutaminase antibody (more than ten times the upper normal limit) plus a positive endomysial antibody on a separate sample can be diagnosed without biopsy and without HLA testing Husby et al. ESPGHAN 2020. The gene test is still useful in pediatrics, but for narrower reasons — equivocal antibodies, screening relatives, ruling out celiac in kids with autoimmune comorbidities — not as a routine confirmation step.

Where this goes wrong in practice

The classic mistakes around this test, in roughly the order they cause problems:

Going gluten-free on a positive result alone. The single most common harm. A patient sees "positive for HLA-DQ2" on a report, sometimes from a direct-to-consumer kit with no clinical guidance, reads the lab's vague comment about celiac risk, and clears the kitchen of bread. They never get tested with antibodies. They have no idea whether they have the disease. The diet is permanent because they're afraid to test off it.
Ordering the gene test as the first step in someone eating gluten. Wastes the cheaper, more informative antibody test. The clinical sequence in a patient with normal eating is: total IgA and tTG-IgA first; HLA only if those leave a real question.
A short gluten challenge after a positive HLA. If you've been gluten-free for years and then eat one slice of bread a day for a week before retesting antibodies, your antibodies haven't had time to rise. False negative, wrong conclusion. The minimum effective challenge is a slice and a half a day for two weeks; the conventional version is three slices a day for six to eight weeks Leffler et al. 2013.
Treating "DQ2 positive" as a single thing. Some lab reports lump the much higher-risk DQ2.5 form with the much lower-risk DQ2.2 form. Two patients can both be "DQ2 positive" and have very different absolute risks Megiorni and Pizzuti 2012. Ask for the underlying alleles.
Overriding a strong clinical picture with a negative gene result. About four people in a thousand with biopsy-confirmed celiac don't carry the textbook genes Karell et al. 2003. Most turn out, on review, to have a different gut disease that was mistaken for celiac — but a tiny residue is real. If symptoms and antibodies and biopsy all point hard at celiac and the gene test is negative, the right move is to repeat the workup rather than dismiss the picture.

Cost, where to get it, what to expect

In the US, list price for clinical HLA-DQ2/DQ8 typing runs roughly $200 to $450 at the major reference labs. Insurance generally covers it when a doctor documents a clinical indication — usually one of the rule-out, family-screening, or equivocal-results scenarios. In European public-health systems the cost is often a fraction of that and bundled into existing celiac care pathways Rubio-Tapia et al. ACG 2023. Direct-to-consumer kits report the same alleles for less but without clinical interpretation — useful for a curious person, not enough on its own to drive a diet decision.

Logistics are simple. One blood draw at a regular outpatient lab or a buccal cheek-swab kit by mail in some markets. No fasting. Turnaround is one to two weeks. The result is yours for life — no monitoring, no retesting, no refreshing every five years. The person who gets the most out of this test is the one who orders it once, files the result with the rest of their medical records, and never thinks about it again.

What changes when you have the answer

Two clean futures branch off this test.

If the result is negative. Within a few days you've eaten bread again. Within a couple of weeks the gluten-free routine you've been running for years is just gone — restaurants are easy, dinner at a friend's apartment doesn't need a phone call, the cabinet stops being a labeled fortress. Your grocery bill drops. A real source of low-grade food anxiety closes for the rest of your life. Your kids stop hearing about it. If you have other gut symptoms, the workup turns to the actual cause — non-celiac gluten sensitivity often among them — instead of circling celiac forever.

If the result is positive and a full workup confirms celiac. The first few weeks of strict gluten-free eating are a learning curve — labels, cross-contamination, what to say at a restaurant. By a few months in, energy starts coming back; the iron supplements you've been taking forever finally hold; the dermatitis-herpetiformis rash you may have been treating as eczema fades. The mental fog a lot of celiac patients describe — the can't-quite-focus, can't-quite-find-the-word state — often lifts in the same window, and people only notice it was there once it's gone Lebwohl et al. 2018. At one to two years, the small intestine has largely healed in most adults Lebwohl et al. 2018, the long-term cancer risk has dropped back toward the general population, and the next generation in your family knows to get tested early. The diet is permanent, but it's permanent for a reason, and you stop guessing about whether it was worth it.

A few adjacent things this entry doesn't cover:

Non-celiac gluten sensitivity. Real for some people, no HLA association, no antibody test, diagnosed by exclusion. The gene test says nothing useful about it.
Wheat allergy. A different immune mechanism (IgE-mediated, like a peanut allergy), tested with skin-prick or specific-IgE panels. Separate problem.
Dermatitis herpetiformis. The skin version of celiac — same HLA-DQ2/DQ8 association, same gluten-free-diet treatment, but the diagnosis runs through a skin biopsy.
Living gluten-free. What to actually buy, what to ask at restaurants, where the hidden gluten is. A whole skill set, downstream of the diagnosis.

Related in the handbook

Diagnosis

— This gene test is one piece of the celiac picture — most useful when antibody or biopsy results come back murky.

Alternatives

— If a food makes you feel bad, this gene test is the validated way to clear celiac — unlike the IgG panels that just flag whatever you eat.

— This is a targeted clinical gene test — the kind that actually changes a diagnosis, not an ancestry spit kit.
— If you lack DQ2 and DQ8, celiac is essentially excluded, which lands you squarely in the non-celiac group instead.
— Celiac is autoimmune and clusters with thyroid disease and others, especially in women — one positive often means screening for more.
— Like HLA-B27 for spine disease, this HLA test only earns its place once the picture already fits.
— Celiac is a classic hidden cause of empty iron stores; this gene test helps decide whether celiac is even worth chasing.

1. Substance and claimed effects

HLA-DQ2/DQ8 testing is a one-time genetic test that types two HLA class II alleles — HLA-DQ2 (most importantly the DQ2.5 heterodimer encoded by DQA1*05 + DQB1*02) and HLA-DQ8 (encoded by DQA1*03 + DQB1*03:02) — to determine whether a person carries the genetic substrate required to develop celiac disease. The test is performed once per lifetime on a blood or buccal sample using PCR-based sequence-specific oligonucleotide hybridization or sequence-specific primer methods Megiorni and Pizzuti 2012. Claimed effects covered by this entry: (i) high negative predictive value as a rule-out test for celiac disease — a negative result makes celiac essentially impossible Karell et al. 2003 Hadithi et al. 2007; (ii) preserved diagnostic certainty when a patient is already on a gluten-free diet and gluten challenge is impractical or refused Husby et al. ESPGHAN 2020 Rubio-Tapia et al. ACG 2023; (iii) risk stratification for first-degree relatives and high-prevalence comorbid populations (type 1 diabetes, Down syndrome, Turner syndrome, autoimmune thyroid disease) Pietzak et al. 2009 Liu et al. 2014; (iv) downstream consequences of the diagnostic call: a confirmed celiac diagnosis commits the patient to a lifelong gluten-free diet, removes a major risk channel for refractory disease and small-bowel lymphoma, and triggers family screening, while a confident rule-out frees the patient from indefinite dietary restriction. The test does not diagnose celiac disease; a positive result only keeps the possibility open. Out of scope: non-celiac gluten sensitivity, wheat allergy, and dermatitis herpetiformis (touched briefly).

2. Evidence by addressing question

mechanism

Celiac disease is a CD4+ T-cell–mediated enteropathy driven by gluten peptides presented on HLA-DQ2 or HLA-DQ8 molecules. Ingested gliadin is partially resistant to gastric and pancreatic proteases, leaving long proline-rich fragments that reach the lamina propria intact. Tissue transglutaminase 2 (tTG2) deamidates specific glutamine residues to glutamate, generating negatively charged peptides that fit with high affinity into the positively charged P4/P6/P7 anchor pockets of DQ2.5 and DQ8. The DQ2/DQ8–peptide complex activates gluten-specific CD4+ T cells, which drive interferon-γ–mediated villous atrophy and B-cell help that produces the IgA anti-tTG and anti-endomysial antibodies clinicians measure Sollid and Jabri 2013. Crucially, in the absence of DQ2.5, DQ2.2, or DQ8, no presentation surface exists for deamidated gliadin and the autoimmune cascade cannot initiate — the genetic prerequisite is mechanistically necessary, not merely correlated Sollid and Jabri 2013 Megiorni and Pizzuti 2012. The DQ2.5 heterodimer presents the highest number of immunogenic gluten epitopes and confers the largest risk; DQ8 presents fewer and confers intermediate risk; DQ2.2 alone (DQA1*02 + DQB1*02 without the DQA1*05 partner) presents the fewest stable peptides and confers the lowest risk among carriers Megiorni and Pizzuti 2012 Wolters and Wijmenga 2008. Gene dosage matters: DQ2.5 homozygotes carry two presentation surfaces and reach an order-of-magnitude higher absolute risk than heterozygotes by age 5 in the prospective TEDDY cohort Liu et al. 2014.

evidence

The HLA association is one of the strongest single-gene–region associations known in any complex autoimmune disease. The landmark European Genetics Cluster on Celiac Disease analysis pooled 1,008 biopsy-confirmed celiac patients across nine European populations: 93.7% carried HLA-DQ2 (DQ2.5), an additional 5.5% carried HLA-DQ8, and of the remaining 0.8% — almost all carried at least the DQB1*02 allele on its own. Only 4 of 1,008 patients (0.4%) carried neither DQ2 nor DQ8 nor a half-DQ2 allele Karell et al. 2003. The Hadithi et al. prospective diagnostic-accuracy study in 463 patients reported HLA-DQ typing sensitivity of 96% and a negative predictive value of 100% for biopsy-confirmed celiac disease in a referral population, supporting the test's use as a rule-out tool Hadithi et al. 2007. Population-attributable risk estimates put the HLA-DQ region at roughly 35–40% of the heritable component of celiac susceptibility; over 40 non-HLA loci identified by GWAS contribute the remainder but none individually approach HLA's effect size Sollid and Jabri 2013. Prospective absolute-risk data come from the TEDDY birth-cohort study of 6,403 HLA-selected children: by age 5, DQ2.5/DQ2.5 homozygotes had a cumulative incidence of celiac disease autoimmunity of 26%; DQ2.5/DQ8 heterozygotes 11%; DQ2.5/X heterozygotes 9%; DQ8/DQ8 homozygotes 6%; DQ8/X heterozygotes 3% Liu et al. 2014. Both the ESPGHAN 2020 pediatric guideline and the ACG 2023 adult guideline endorse HLA testing specifically for ruling out celiac in patients on a gluten-free diet, equivocal histology or serology, Down syndrome, and selected family-screening scenarios Husby et al. ESPGHAN 2020 Rubio-Tapia et al. ACG 2023.

protocol

A single blood draw (EDTA tube) or buccal swab is sent to a clinical genetics or specialty lab. Test menus vary: minimal panels report only DQ2 and DQ8 positivity; better panels report the underlying DQA1 and DQB1 alleles, distinguish DQ2.5 from DQ2.2, and flag DQ2 homozygosity. Results are categorical and lifelong — no retesting needed. The clinically actionable readouts are: (a) negative for both DQ2 and DQ8 (and the DQB1*02 half-allele) → celiac disease ruled out with NPV approaching 100% in published cohorts Hadithi et al. 2007 Karell et al. 2003; (b) positive for DQ2.5 or DQ8 → celiac remains possible; proceed with antibody testing on a gluten-containing diet and small-bowel biopsy if serology is positive Rubio-Tapia et al. ACG 2023; (c) positive specifically for DQ2.5 homozygosity → elevated absolute risk; consider closer surveillance in pediatric first-degree relatives Liu et al. 2014. Decision logic: order HLA-DQ2/DQ8 testing when the patient is (i) already on a gluten-free diet without a prior firm diagnosis, (ii) refusing or unable to complete a gluten challenge, (iii) a first-degree relative of a celiac patient who wants a one-time rule-out, (iv) has equivocal serology/biopsy results, or (v) belongs to a high-prevalence comorbid group. Order standard serology (tTG-IgA + total IgA) first if the patient is willing and able to consume gluten — serology is cheaper and more directly diagnostic Rubio-Tapia et al. ACG 2023.

contraindications

No medical contraindications exist for the test itself — it is a single blood draw or cheek swab measuring fixed germline DNA. Editorial caveats apply to interpretation, not safety: a positive result must not be communicated as a celiac diagnosis (~30–40% of European-descent populations carry DQ2 or DQ8 yet only ~1% develop the disease), and HLA testing should not be used as a population screening tool because its positive predictive value is low at population pre-test probability Wolters and Wijmenga 2008 USPSTF 2017. Insurance coverage and genetic-information disclosure (GINA-like protections in some jurisdictions, none in others) is a real consideration for some patients; the result is otherwise low-stakes from a privacy standpoint compared to BRCA or APOE testing.

misconceptions

Three persistent confusions among patients and even some primary-care clinicians:

"Positive HLA-DQ2/DQ8 means I have celiac." No — it means celiac is possible. Roughly 30–40% of Northern European-descent populations carry at least one risk allele, but population celiac prevalence is ~1% Singh et al. 2018 Wolters and Wijmenga 2008.
"I can do HLA testing instead of going back on gluten." Partially true. A negative result rules celiac out reliably and frees the patient. A positive result does not confirm celiac — to confirm requires antibody testing and usually biopsy, which require ongoing gluten exposure of at least 3 grams/day for 2 weeks at minimum, and ideally 10 grams/day for 6–8 weeks, to produce reliable serology and histology changes Leffler et al. 2013 Rubio-Tapia et al. ACG 2023.
"HLA testing replaces biopsy." Only partially in pediatrics, and the role has narrowed. The ESPGHAN 2020 update actually removed routine HLA typing from the pediatric no-biopsy pathway — pediatric celiac can now be diagnosed without biopsy when tTG-IgA exceeds 10× the upper limit of normal and a separate EMA test is positive, regardless of HLA status. HLA testing remains valuable in pediatric scenarios where the diagnosis is uncertain or family screening matters Husby et al. ESPGHAN 2020.

audience

The test's utility is highly population-dependent. Subgroups where it adds disproportionate value:

First-degree relatives of celiac patients. Pooled meta-analysis pegs first-degree relative prevalence at ~7.5% — roughly an order of magnitude above the general population. Siblings carry the highest risk (~8.9%), children of celiac parents next, then parents Singh et al. 2015. A negative HLA test in a first-degree relative provides lifelong reassurance; positive relatives are then candidates for periodic serology.
Type 1 diabetes. Celiac prevalence in T1DM cohorts is ~3–10%, partly because T1DM and celiac share DQ2/DQ8 risk alleles. ACG 2023 recommends serology screening at T1DM diagnosis; HLA typing can stratify which T1DM patients warrant repeated serology over time Rubio-Tapia et al. ACG 2023.
Down syndrome and Turner syndrome. Celiac prevalence is roughly 6–12% in Down syndrome and ~4–8% in Turner syndrome. HLA testing identifies the subset for whom future serology surveillance is meaningful Rubio-Tapia et al. ACG 2023.
Patients already on a gluten-free diet. Self-started gluten-free diets are common; symptoms improve, the patient never gets a firm diagnosis, and serology turns negative. A negative HLA test in this group definitively rules out celiac and allows return to gluten without long-term concern Husby et al. ESPGHAN 2020 Rubio-Tapia et al. ACG 2023.

alternatives

HLA-DQ2/DQ8 testing is one of three diagnostic pillars for celiac disease and is rarely used alone:

Serology (tTG-IgA + total IgA, with EMA reflex): The default first-line test on a gluten-containing diet. Sensitivity and specificity for biopsy-confirmed celiac are both ~95%+ for tTG-IgA. Cheaper than HLA, but requires gluten exposure and can normalize on a gluten-free diet Lebwohl et al. 2018 Rubio-Tapia et al. ACG 2023.
Small-bowel biopsy: The historical gold standard. Marsh 2–3 lesions (villous atrophy, intraepithelial lymphocytosis, crypt hyperplasia) on duodenal biopsy confirm celiac. Requires endoscopy and adequate gluten exposure Bao and Bhagat 2012.
Gluten challenge then re-test serology + biopsy: The alternative when a patient is already gluten-free. Effective but burdensome: 3 g/day for 2 weeks induces serology and histology changes in most patients, but symptom-tolerated challenges in adults often run 6–8 weeks at 10 g/day to maximize sensitivity. Many patients refuse outright because of symptoms Leffler et al. 2013.

HLA testing complements rather than replaces these. The decision tree: if a patient can eat gluten, do serology first; if a patient cannot or will not, HLA testing is the lowest-friction path to a confident rule-out or to keeping the diagnosis open.

failure-modes

The classic clinical mistakes around HLA-DQ2/DQ8 testing:

Overreading a positive. Clinicians or labs that report "HLA-DQ2 positive: consistent with celiac disease" without context create real patient harm — unnecessary biopsies, premature gluten-free diet commitments, family alarm. A positive result has positive predictive value of only ~2–3% at population pre-test probability Wolters and Wijmenga 2008.
Using HLA as a screening test in low-risk populations. USPSTF found insufficient evidence to recommend universal celiac screening in asymptomatic adults; HLA testing fares worse than serology as a first-line screen because of its low PPV USPSTF 2017.
The ~0.4% truly HLA-negative celiac patient. Karell et al. documented a tiny residual: 4 of 1,008 biopsy-confirmed cases carried neither DQ2 nor DQ8 nor the DQB1*02 half-allele. Most of these on re-review were misdiagnosed (giardiasis, autoimmune enteropathy, common variable immunodeficiency) — but the residual is real and a strongly clinical picture should override a negative HLA result Karell et al. 2003.
Inadequate gluten challenge after a positive HLA. Patients who reintroduce gluten for only a few days before serology often produce false negatives. The minimum effective challenge is ~3 g/day (one and a half slices of bread) for 2 weeks; serology takes weeks to rise Leffler et al. 2013.
Reporting variability. Lab menus differ in whether DQ2.2 is distinguished from DQ2.5; some labs report only "DQ2 positive" without specifying the heterodimer. DQ2.2 alone carries much lower risk than DQ2.5, and lumping them can mislead risk stratification Megiorni and Pizzuti 2012.

practicalities

The test is widely available through clinical genetics labs, major reference labs (LabCorp, Quest), and increasingly direct-to-consumer panels. US list price is typically $200–$450; insurance coverage in the US depends on documented clinical indication (celiac evaluation, family screening, equivocal serology). European public-health systems often include HLA-DQ celiac typing in covered care pathways at substantially lower cost (~€50–150). Turnaround is typically 1–2 weeks. One blood draw or buccal swab; no fasting required. The result is lifelong — once typed, the patient and their physicians never need to repeat the test. Multi-locus celiac panels that include serology + HLA in one order are sometimes available but rarely cost-effective; serology should generally precede HLA testing in patients who can eat gluten Rubio-Tapia et al. ACG 2023.

stakes

Two failure scenarios the test prevents. First, undiagnosed celiac on continued gluten exposure: untreated celiac disease produces malabsorption, iron and vitamin D deficiency, osteoporosis, infertility and recurrent miscarriage, peripheral neuropathy, transaminitis, and an elevated risk of small-bowel adenocarcinoma and enteropathy-associated T-cell lymphoma. Long-term mortality in untreated celiac is modestly elevated; mortality on a strict gluten-free diet returns toward population baseline Lebwohl et al. 2018. Second, the inverse failure: lifelong gluten-free dieting on a false suspicion. The diet is socially burdensome, costs 50–250% more for staple foods, is associated with lower intake of fiber, B vitamins, and iron, and removes a major source of food enjoyment. A negative HLA result frees an undiagnosed gluten-avoider from this burden permanently. Family stakes also follow: a confirmed celiac diagnosis triggers screening of first-degree relatives where 1 in ~13 will also have undiagnosed disease Singh et al. 2015.

payoff

For the patient with confirmed celiac who tests HLA-positive, the payoff is downstream: a confident commitment to a strict gluten-free diet, with mucosal healing in 60–80% of adults by 2 years and a return toward population-baseline mortality and cancer risk Lebwohl et al. 2018. For the negative-HLA patient who was gluten-avoiding "just in case," the payoff is immediate: the entire question is closed, the diet ends, and a major source of food anxiety is resolved within weeks of returning to normal eating. For first-degree relatives, a negative HLA result removes a lifelong worry — they never need celiac screening again, regardless of vague gastrointestinal symptoms in the future Husby et al. ESPGHAN 2020.

out-of-scope

Adjacent topics this entry does not cover but a reader may want to pursue: non-celiac gluten sensitivity (no HLA association, no serology, diagnosis of exclusion); wheat allergy (IgE-mediated, distinct from celiac); dermatitis herpetiformis (the cutaneous manifestation of celiac — same HLA-DQ2/DQ8 association); the long-term gluten-free diet itself (nutritional management, cross-contamination); refractory celiac disease (rare, ~1–2% of celiac patients, requires specialty care).

3. The credibility range

Optimist case

HLA-DQ2/DQ8 testing is one of the few genetic tests in routine clinical practice with a near-100% negative predictive value for a specific disease and decades of replicated population data behind it. The mechanistic case is unusually clean: gliadin presentation literally cannot occur without DQ2 or DQ8, so the negative result has biological force, not just statistical association Sollid and Jabri 2013. The test resolves the single hardest diagnostic dilemma in celiac evaluation — the patient who has self-started a gluten-free diet, often years ago, and refuses or cannot tolerate a 6-week gluten challenge for repeat serology and biopsy. It also gives first-degree relatives a one-time decision rather than indefinite repeat-serology surveillance. The cost (~$200–450, one time, no follow-up) is favorable against the alternative of multi-week gluten challenges and endoscopy under sedation.

Skeptic case

The test's positive predictive value is poor in unselected populations: ~30–40% of European-descent populations carry DQ2 or DQ8 yet only ~1% develop celiac disease, so a positive result conveys very little diagnostic information Wolters and Wijmenga 2008. Clinicians who order HLA testing reflexively, without first attempting serology on a gluten-containing diet, generate uninterpretable results and patient anxiety. The role in pediatric diagnosis has narrowed: ESPGHAN 2020 explicitly removed routine HLA typing from the no-biopsy pathway, where it was previously required Husby et al. ESPGHAN 2020. The very rare HLA-negative celiac patient (~0.4% in Karell et al.'s pooled cohort) means the test is not absolutely 100% sensitive, and a strongly clinical picture should override a negative result Karell et al. 2003. Direct-to-consumer marketing of HLA typing as a "celiac test" without serology context risks both false reassurance (in the positive case, "I have the gene therefore I must avoid gluten") and false alarm.

Author's call

The evidence for HLA-DQ2/DQ8 testing as a high-NPV rule-out test for celiac disease is settled and uncontroversial; the test belongs in any celiac diagnostic algorithm that includes a patient already on a gluten-free diet, equivocal serology/histology, or systematic family screening. The controversy is editorial, not scientific: how aggressively to deploy the test, and how to communicate a positive result without triggering inappropriate diet changes. The entry lands as a confident test recommendation with strong evidence and modest controversy, scoped narrowly to the rule-out and family-screening use cases rather than a general celiac workup tool.

4. Stakeholder and incentive map

Pediatric and adult GI societies (ESPGHAN, ACG, BSPGHAN, NICE) drive the evidence-based indications. ESPGHAN 2020 narrowed the test's pediatric role; ACG 2023 preserved a broader adult role. Generally aligned on rule-out and family-screening utility Husby et al. ESPGHAN 2020 Rubio-Tapia et al. ACG 2023.
Reference and specialty genetics labs (LabCorp, Quest, EuroImmun, Prometheus) market HLA-DQ celiac typing as a clinical assay. Modest commercial incentive but constrained by insurance coverage rules that require documented clinical indication.
Direct-to-consumer genomics (23andMe, AncestryHealth) report DQ2/DQ8 as part of broader panels. Incentive to inflate clinical interpretation of positive results; this pressure drives some of the misconception problem.
Patient advocacy groups (Celiac Disease Foundation, Coeliac UK, Beyond Celiac) advocate broad family screening including HLA testing where serology is impractical. Aligned with the rule-out use case.
USPSTF and primary-care guidelines push back against population screening of asymptomatic adults USPSTF 2017, which limits indiscriminate use.
Gluten-free food industry benefits from any diagnosis (real or self-imposed) but is downstream of testing decisions rather than a driver of them.

5. Population variability

The bulk of HLA-DQ2/DQ8 data comes from Northern European, Mediterranean, and US populations of European descent. Approximate DQ2.5 + DQ8 carrier prevalence: Northern Europe ~30–40%, Mediterranean Europe ~25–35%, North America (admixed) ~25–35%, North African populations ~30–40%, Latin American populations ~15–25% (with relatively higher DQ8). Sub-Saharan African and East Asian populations have much lower DQ2.5 prevalence (often <5–10%), and celiac disease is correspondingly rare though not absent Singh et al. 2018. The Saharawi population in North Africa has the highest documented celiac prevalence globally (~5–6%), tracking very high DQ2.5 frequency. Within carriers, gene dosage matters: DQ2.5 homozygotes show 5–10× the absolute risk of single-copy heterozygotes by age 5 Liu et al. 2014. Female sex roughly doubles celiac risk. Co-existing autoimmunity (T1DM, autoimmune thyroid disease, Down syndrome, Turner syndrome) markedly elevates both DQ2/DQ8 prevalence and conditional celiac risk. The interpretive lesson: a positive HLA result in a low-prevalence-ancestry patient with no risk factors carries even less PPV than the European baseline; a negative result remains a strong rule-out across all populations.

6. Knowledge gaps

The exact mechanism of HLA-negative celiac. The 0.4% of biopsy-confirmed cases in Karell et al. who carry neither DQ2 nor DQ8 nor a half-DQ2 remain incompletely explained — some are misdiagnosed enteropathies, but whether a small residue has a genuinely DQ-independent celiac-like disease is unresolved Karell et al. 2003.
Quantifying lifetime celiac risk in DQ2.5 homozygotes beyond age 5. TEDDY data extend to childhood; adult-onset celiac in DQ2.5 homozygotes is less precisely quantified, and the question of whether all homozygotes eventually develop disease if they live long enough remains open Liu et al. 2014.
Non-HLA modifiers. Over 40 non-HLA loci contribute to celiac risk but explain a small fraction of the heritability; integration of HLA + non-HLA polygenic risk scores for clinical use is research-stage, not clinical-stage.
Cost-effectiveness in low-prevalence population screening. Whether HLA-first screening of high-risk groups (T1DM, Down syndrome) outperforms serology-first surveillance over a lifetime has not been settled with full economic modeling.
Direct-to-consumer interpretation harms. The downstream behavior of patients who learn they are "DQ2 positive" through a consumer test, without clinical context, has not been studied systematically.

Scope and brief alignment. The brief named three downstream consequences: diagnostic certainty when gluten challenge isn't feasible, lifelong diet decisions, and family screening. All three are covered end-to-end — the rule-out use case anchors the dek, mechanism, and protocol sections; lifelong diet implications drive the stakes and payoff framing; family screening has its own audience bullet plus the relative-risk number from Singh et al. 2015. No silent narrowing.

Hard call on the ESPGHAN 2020 shift. The European 2020 guideline removed HLA testing from the routine pediatric no-biopsy pathway, which complicates the simple "this is what you do for kids" story. Chose to address it head-on in the misconceptions section rather than pretend the test's role hasn't narrowed, because direct-to-consumer marketing and older patient resources still push the older framing. The adult ACG 2023 indications remain broader, which is what the article foregrounds.

Sleep scored 0 deliberately. Untreated celiac causes restless nights via GI symptoms and iron deficiency in some patients, but the link is too indirect and downstream-conditional to support a non-zero score against an evidence gate. Beauty-direct also 0 — no short-term cosmetic effect from a genetic test itself.

Beauty-cumulative kept at 1. Dermatitis herpetiformis, telogen effluvium, and pallor reverse on a gluten-free diet for the celiac subset, but the entry's substance is the test, not the diet. A 1 honestly reflects that the cumulative beauty effect is real but conditional on disease + adherence.

Category placement. Picked screening over gut-digestion or medical because the test's editorial function is screening/risk stratification, and most readers will browse for it in that bucket. If the catalogue later grows a dedicated genetics subcategory, revisit.

Controversy scored 2, not 1. The ESPGHAN/ACG difference in pediatric vs adult guidance plus the direct-to-consumer interpretation debate are real if minor disagreements. Not a battleground; not a quiet consensus either.

Future links to wire. When these entries exist, cross-link from this one: tissue transglutaminase antibody test (tTG-IgA), gluten-free diet, dermatitis herpetiformis, non-celiac gluten sensitivity, wheat allergy, celiac disease (overview), small bowel biopsy / endoscopy, iron-deficiency anemia workup, first-degree relative screening protocols.

Separate-entry candidates surfaced during research. The gluten-free diet warrants its own entry (nutritional management, cost, social cost, cross-contamination). Celiac disease (overview) as a condition-literacy entry would be the natural parent. Dermatitis herpetiformis is narrow enough to be its own entry. The HLA test entry is intentionally kept tight to the test itself rather than expanding into a celiac primer.

Rating note on evidence = 5. Met the "name 2+ rigorous trials" bar without strain: Karell et al. 2003 pooled cohort, Hadithi et al. 2007 diagnostic accuracy, Liu et al. 2014 TEDDY prospective cohort, plus two aligned specialty guidelines. The mechanism (Sollid and Jabri 2013) is fully characterized at the molecular level. This is the rare clinical genetic test where a 5 is honest.

Voice note on stakes/payoff. Tried hard to anchor stakes on the typical reader — the long-gluten-free no-diagnosis adult and the slowly-malabsorbing on-bread adult — rather than the extreme case (lymphoma, infertility crisis). Mentioned the cancer and miscarriage risks because they're real and load-bearing for the longevity score, but kept them inside a "slow tax" framing rather than fear-mongering.