September 2010 Clinical Laboratory News: Celiac Disease

CLN Banner Logo

September 2010: Volume 36, Number 9

Celiac Disease
Navigating the Maze of Laboratory Testing

By Melissa R. Snyder, PhD, and Joseph A. Murray, MD

Logo graphic
Logo graphic courtesy of

The lack of sensitive and specific biomarkers hampers diagnostic testing for some diseases. But celiac disease, a disorder of the small intestine that today is four times more common than it was 50 years ago, definitely does not fall into this category. Triggered by consumption of the protein gluten found in wheat and other grains, celiac disease is associated with numerous serological and genetic markers. The challenge, however, is knowing which test or tests to select for a patient, as well as determining what those test results mean for the continuing evaluation of the patient.

Laboratorians who are knowledgeable about the various assays used to diagnose celiac disease can provide valuable assistance to clinicians evaluating patients suspected of having the disorder. Here we present background information on the clinical manifestations and pathophysiology of celiac disease, followed by a discussion about both selecting the most appropriate tests based on a patient’s clinical presentation and interpreting the test results.

Clinical Presentation

Celiac disease is a chronic inflammatory disease that affects nearly 1% of the U.S. population (1). Caused by an inflammatory response mounted by the patient’s immune system, the disorder primarily involves the small intestine, although systemic symptoms often occur. This inflammatory response leads to expansion of intraepithelial lymphocytes, crypt hypertrophy, and atrophy of the small intestinal villi (2, 3).

Initiation of the immune response requires both a genetic predisposition and specific environmental exposure (4). The environmental component occurs from ingestion of gluten, a primary protein component of wheat, and related proteins from barley and rye. Much of the genetic susceptibility to celiac disease is related to specific alleles of the human leukocyte antigen (HLA) complex, namely the gene pairs encoding the serotype equivalents of the now defunct, but often-still-used nomenclature: HLA-DQ2 and HLA-DQ8.

Although celiac disease has a genetic component, not all the genes have been completely defined. Approximately 10% of first-degree family members of affected individuals will be diagnosed with celiac disease. However, given that only a 30% concordance exists for celiac disease diagnosis in HLA-matched siblings, other unidentified genetic factors must also play a role (5).

The classic presentation of celiac disease, which can be attributed to damage of the small intestine, includes diarrhea that may or may not be accompanied by abdominal pain (Table 1). Although strongly suggestive of a gastrointestinal disorder, these symptoms may occur in less than 50% of patients with celiac disease (2). Adult patients, particularly those with longer-standing disease, may show symptoms of malnutrition, including iron-deficient anemia and osteoporosis, while children may present with a failure to thrive. Other, more systemic clinical manifestations include neurologic, rheumatologic, and dermatologic symptoms (1). Certain laboratory abnormalities found in patients with celiac disease, some of which may be clinically silent, are hypoproteinemia, hypocalcemia, hypertransaminasemia, and various vitamin deficiencies (5).

Patients with celiac disease may also present with one of several associated clinical disorders (Table 1) (1), such as underlying selective IgA deficiency. Although still relatively rare, this condition occurs approximately 10-fold more frequently in patients with celiac disease compared to the general population. IgA deficiency is not usually clinically significant, but recognizing it in the diagnostic workup is important because many specific antibody tests used to diagnose celiac disease detect antibodies of the IgA isotype.


Table 1
Overview of Celiac Disease
Clinical Symptoms
General Laboratory Abnormalities
Associated Clinical Disorders



Abdominal pain



Secondary lactose intolerance

Vomiting (especially nocturnal)



Weight loss



Dental enamel hypoplasia



Neurological symptoms (ataxia, peripheral neuropathy, and dementia)

Failure to thrive

Delayed puberty

Reduced fertility

Iron deficiency

Anemia (microcytic, macrocytic, or dimorphic)

Blood smear with Howell jolly bodies (hyposplenism)

Zinc and copper deficiency

Folate deficiency

Vitamin deficiencies (vitamin E, K, and/or D)




Selective IgA deficiency

Autoimmune diseases

Type I diabetes

Autoimmune thyroid disease

Primary biliary cirrhosis

Chronic autoimmune hepatitis

Sjögren syndrome


Dermatitis herpetiformis

Microscopic colitis


In general, celiac disease is also associated with a predisposition to autoimmune disease and is found with increased frequency in patients with type I diabetes, autoimmune thyroid disease, primary biliary cirrhosis, and autoimmune hepatitis. Patients can also have dermatologic manifestations, such as dermatitis herpetiformis (DH). This condition results from IgA deposits in the skin and is associated with the same autoantibodies implicated in the pathogenesis of celiac disease. In fact, patients with DH often have evidence of villous atrophy on small intestinal biopsy, even in the absence of overt gastrointestinal symptoms.

Celiac Disease Pathogenesis

Although much is known about celiac disease, scientists have not fully elucidated its pathogenesis (2, 4). What is known is that when individuals eat food containing wheat, barley, or rye, digestion of the gluten protein is incomplete. The ethanol-soluble fraction of gluten, referred to as gliadin, contains a 33-amino acid fragment that resists digestion and is thought to contain the immunodominant peptide responsible for initiating the inflammatory response. In individuals with a genetic predisposition to the disease, antigen-presenting cells present the peptide on the cell surface in the context of the HLA-DQ2 and/or HLA-DQ8 molecules.

These HLA molecules can bind the gliadin peptide, but a post-translational modification—specifically, deamidation of glutamine by tissue transglutaminase (TTG)—enhances the interaction. TTG converts the neutral glutamine residues to negatively-charged glutamate residues. Due to the nature of the peptide-binding groove, there is enhanced interaction between the negatively-charged deamidated gliadin peptide and the HLA-DQ2 and HLA-DQ8 molecules. This peptide/HLA complex then interacts with the T-cell receptor on CD4 T cells, which likely initiates the inflammatory response. The CD4 T cells interact with B cells to induce antibody production and with other lymphocytes to initiate cytokine secretion and cytotoxic activity, all of which contribute to the small intestinal atrophy that is characteristic of celiac disease.

Establishing Diagnosis

Diagnostic evaluation for celiac disease is warranted in patients with classic gastrointestinal and non-classic systemic symptoms, in individuals with a family history of celiac disease, and in asymptomatic patients with associated co-morbid conditions. Clinicians establish a presumptive diagnosis of celiac disease based on positive autoantibody serology and an intestinal biopsy with evidence of villous atrophy. After a presumptive diagnosis has been established, patients are put on a gluten-free diet, optimally with the guidance of an experienced dietician.

Once gluten has been successfully abolished from the diet, the patient’s clinical symptoms should begin to resolve, often accompanied by a decrease in antibody titers and reconstitution of the small intestinal villi. Based on these additional findings, clinicians can definitively establish the diagnosis.

Laboratory Tests

Laboratory evaluation, particularly serology and sometimes HLA typing, plays a key role in diagnosing celiac disease. The primary antibodies associated with the disorder are TTG antibodies, deamidated gliadin antibodies (DGA), and endomysial antibodies (EMA), with testing available for both IgA and IgG isotypes (Table 2). These antibodies are physiologically and pathologically related.


Table 2
Primary Antibodies Associated with Celiac Disease
Endomysial antibody
Tissue transgluta-minase Indirect 
Sensitive and specific for celiac disease, although variability is an issue due to subjective nature of immunofluorescence
Although may be useful for patients with IgA deficiency, testing is rarely performed and has been replaced by TTG-IgG and DGA-IgG
Anti-tissue transgluta-minase antibody
Tissue transgluta-minase Enzyme 
Based on sensitivity and specificity, useful as first-level of diagnostic testing for celiac disease
Useful primarily for patients with IgA deficiency due to lower sensitivity in comparison to IgA isotypes
Anti-gliadin antibody
Unmodified gliadin Enzyme 
Not recommended for the evaluation of celiac disease due to lower sensitivity and specificity
Not recommended for the evaluation of celiac disease due to lower sensitivity and specificity
Anti-deamidated gliadin antibody
Deamidated gliadin Enzyme 
May be useful in diagnosis of celiac disease, although appears to be slightly less sensitive than TTG-IgA
Useful primarily for patients with IgA deficiency due to lower sensitivity in comparison to IgA isotypes


EMA, one of the first antibodies associated with celiac disease, is so-named because it recognizes an antigen in the endomysium, the connective tissue that surrounds smooth muscle fibers. Researchers later identified the specific antigen target of EMAs as TTG and developed solid-phase immunoassays specific for antibodies against TTG. Today labs typically use immunofluorescent assays to detect EMAs. These semi-quantitative assays use a tissue substrate with abundant smooth muscle, such as monkey esophagus or human umbilical cord. Positive results are generally reported as a titer.

The third antibody associated with celiac disease, DGA, recognizes an antigen related to dietary gluten and is responsible for initiating inflammation in celiac disease. Early DGA immunoassays tested for antibodies against unmodified gliadin peptides. Given that deamidation of gliadin seems to result in a more clinically relevant antigen, the new generation of gliadin antibody assays has replaced unmodified gliadin peptides with deamidated gliadin.

Numerous studies have addressed the sensitivity and specificity of the various serologic tests for celiac disease. EMA and TTG antibodies of the IgA isotype are comparable, although there is some variation depending on the specific methodology and patient population. Both antibodies generally have specificities > 95% and sensitivities > 90% (6). The sensitivity of EMA testing varies somewhat more between labs, possibly due to the subjective nature of the immunofluorescence interpretation.

In contrast, assays for antibodies against unmodified gliadin are less sensitive, approximately 80%, and less specific, 80–90% (6). Assays that detect antibodies against deamidated gliadin are significantly better than those that use unmodified gliadin. While initial studies found that DGA-IgA performs similarly to TTG-IgA (7), in a recent meta-analysis, researchers concluded that although there is no difference in specificity, TTG-IgA is significantly more sensitive (93%) than DGA-IgA (88%) (8). Therefore, labs should no longer use older gliadin antibody tests for evaluation of celiac disease.

Researchers have also studied the clinical utility of assays for IgG isotype antibodies in diagnosing celiac disease. EMA-IgG and TTG-IgG assays show specificity in excess of 98%, although this may only apply to patients with IgA deficiency. IgG assays, however, are considerably less sensitive, generally around 40% (6). On the other hand, although DGA-IgA assays may be slightly less sensitive than TTG-IgA assays, DGA-IgG seems to be more sensitive than its TTG counterpart (65–70%) (7). DGA-IgG’s sensitivity is still inferior to IgA isotypes assays; therefore, labs may want to used the IgG assay only for patients who are IgA deficient.

Currently, labs perform IgA and IgG assays for TTG and DGA almost exclusively by plate-based enzyme immunoassay, although some still use radioimmunoassays. Newer technologies, specifically fluorescence-based multiplex immunoassays, may become an option for celiac disease serology testing in the future (9) because they offer the advantage of testing for multiple antigen specificities in a single reaction. This format may offer significant process improvements for labs, as well as being more cost-effective, but the clinical implications for large-scale diagnostic screening using multi-antibody panels will need to be carefully evaluated.

The other type of testing used for celiac disease is HLA typing. The HLA-DQ2 allele is found in 90-95% of individuals with celiac disease, with the remaining 5–10% possessing the HLA-DQ8 encoding allele (3, 5). Of the HLA-DQ2 alleles, DQA1*05xx, DQB1*0201 is most closely associated with celiac disease risk. In addition, homozygosity for DQA1*0201, DQB1*0202 is also associated with an increased risk, while heterozygosity for these alleles imparts only a very weak predisposition to celiac disease. For HLA-DQ8, the risk alleles for celiac disease are DQA1*03xx, DQB1*0302.

Most labs use DNA-based techniques, such as allele-specific hybridization, to identify the genes encoding the α- and β- chains that combine to form the heterodimeric DQ2 and DQ8 molecules. Since HLA-DQ2 and/or HLA-DQ8 are found in virtually all patients with celiac disease, the absence of these gene pairs virtually excludes celiac disease as a diagnosis. Most people who have the DQ2 or DQ8 alleles, however, will not develop celiac disease. In other words, the HLA gene pairs are necessary, but not sufficient, for an individual to be predisposed to celiac disease. Although this HLA typing cannot establish a diagnosis of celiac disease, it is useful for ruling out the disease in cases where neither gene pair is detected.

An Approach to Lab Evaluation

At this time, labs should not rely upon serologic testing alone to establish a diagnosis of celiac disease. However, serology, in conjunction with HLA typing, is useful for identifying individuals at high risk for celiac disease who should undergo a small intestine biopsy. This testing strategy also prevents patients for whom celiac disease is an unlikely diagnosis from having to go through the invasive procedure.

Given the variety of available tests, choosing the tests that are most appropriate for a particular patient and interpreting the results can prove challenging for both clinicians and labs. Researchers have proposed several testing panels designed to simplify lab testing while still maintaining high sensitivity and specificity (10, 11). At Mayo Clinic, we developed a diagnostic algorithm based on input from our physicians and scientists (Figure 1, below).

Celiac figure

Click here for Figure 1

Due to the prevalence of IgA deficiency in patients with celiac disease, this algorithm begins with an assessment of total IgA levels. If the IgA levels are within the age-adjusted reference range, the most appropriate second level of testing, given its sensitivity and specificity, is for TTG-IgA antibodies. If this test result is positive, a diagnosis of celiac disease is possible, and the recommendation is to proceed to a biopsy of the small intestine. If the result for the TTG-IgA antibody test falls in an equivocal or indeterminate range, further serologic testing may be of useful, specifically the DGA-IgA and EMA-IgA antibodies. Patients should have a biopsy of the small intestine if either the EMA-IgA or DGA-IgA tests are positive. For individuals with IgA deficiency, a multi-antibody approach may also be useful.

For individuals with a low but detectable concentration of total IgA, our testing scheme evaluates TTG and DGA, both the IgA and IgG isotypes. In patients who have no detectable IgA, testing should include TTG and DGA, but only the IgG isotype. Any positive result, regardless of the antigen specificity, would be consistent with a possible diagnosis of celiac disease, and a small intestine biopsy would be necessary.

One important caveat about autoantibody testing is that negative serology does not completely rule out the possibility of celiac disease. For some patients, this may reflect the natural course of the condition, as patients with mild clinical symptoms and partial villous atrophy may have lower antibody concentrations or may be completely seronegative. In other cases, the negative serology may be due to initiation of a gluten-free diet prior to testing where removal or reduction of the environmental trigger abrogates the inflammatory response, ultimately leading to decreased autoantibody production. Alternatively, if serology is negative, but celiac disease is highly suspected based on clinical presentation or a family history, the patient should have the HLA typing tests. Although a positive result is not diagnostic, the absence of the DQ2 and DQ8 alleles excludes celiac disease from the differential diagnosis and would prevent an unnecessary biopsy.

Following the biopsy, clinicians can evaluate the pathology results in the context of the serology data. If the biopsy shows evidence of villous atrophy and the serology is positive for one or more specific antibodies, then a presumptive diagnosis of celiac disease is established. However, if the patient underwent the biopsy based on positive serology and the pathology results are negative, the results are inconclusive. For these cases, HLA typing is also appropriate if the patient has not been tested previously. In the event that the HLA typing is negative for both DQ2 and DQ8, then both the biopsy and HLA result indicate that celiac disease is not a probable diagnosis, and the serology result was likely a false positive.

On the other hand, if the patient is positive for DQ2 or DQ8, then celiac disease remains a possibility. In this instance, both the serology and HLA results are positive, but the biopsy is negative. For example, the patient may be symptomatic and the biopsy may be a false negative, perhaps due to inadequate sampling or improper histology. In asymptomatic patients, the negative biopsy may accurately reflect the fact that the patient’s small intestinal villi have not been exposed to an inflammatory insult. But given that the patient’s serology was positive and a compatible HLA type was identified, clinicians will want to watch for possible future development of disease.

Selecting the Best Approach to Testing

There are many possible approaches to diagnostic testing for suspected celiac disease. For an individual patient, the selection of the most appropriate lab tests should be guided by the specific clinical scenario including positive family history, adoption of a gluten-free diet, and relevant co-morbid conditions. The results from all lab testing, both serology and HLA typing, should be interpreted within the clinical context to determine if a biopsy is needed. Finally, if a biopsy is performed, clinicians should consider the results of the entire evaluation to determine if the patient already has or is at risk for developing celiac disease in the future.


  1. Barton SH and Murray JA. Celiac disease and autoimmunity in the gut and elsewhere. Gastro Clin N Amer 2008;37:411–28.
  2. Green PHR and Cellier C. Celiac disease. New Engl J Med 2007;357:1731–43.
  3. Farrell RJ and Kelly CP. Celiac sprue. New Engl J Med 2002;346:180–8.
  4. Jabri B and Sollid LM. Mechanisms of disease: immunopathogenesis of celiac disease. Nat Clin Prac Gastro Hepat 2006;3:516–25.
  5. Green PHR and Jabri B. Celiac disease. Annu Rev Med 2006;57:207–21.
  6. Rostom A, Dube C, Cranney A, Saloojee N, et al. The diagnostic accuracy of serologic tests for celiac disease: a systematic review. Gastro 2005;128:S38–46.
  7. Rashtak S, Ettore MW, Homburger HA, Murray JA. Comparative usefulness of deamidated gliadin antibodies in the diagnosis of celiac disease. Clin Gastro Hepat 2008;6:426–32.
  8. Lewis NR and Scott BB. Meta-analysis: deamidated gliadin peptide antibody and tissue transglutaminase antibody compared as screening tests for coeliac disease. Aliment Pharmacol Ther 2009;31:73–81.
  9. Rashtak S, Ettore MW, Homburger HA, Murray JA. Combination testing for antibodies in the diagnosis of coeliac disease: comparison of multiplex immunoassay and ELISA methods. Aliment Pharmacol Ther 2008;28:805–13.
  10. Hopper AD, Cross SS, Hurlston DP, McAlindon ME, et al. Pre-endoscopy serological testing for coeliac disease: evaluation of a clinical decision tool. BMJ 2007;334:729–33.
  11. McGowan KE, Lyon ME, Butzner, JD. Celiac disease and IgA deficiency: complications of serological testing approaches encountered in the clinic. Clin Chem 2008;54:1203–09.

Melissa Snyder
Melissa R. Synder, PhD, is assistant professor of laboratory medicine/pathology in the Division of Clinical Biochemistry and Immunology, Department of Laboratory Medicine and Pathology at the Mayo Clinic, Rochester, Minn. Email:

Joseph A. Murray
Joseph A. Murray, MD, is professor of medicine in the Division of Gastroenterology and Hepatology, Department of Internal Medicine at the Mayo Clinic. Email:

Page Access: