The Diagnostic Pathway
Coeliac disease diagnosis follows a stepwise approach: serological screening, followed by confirmatory duodenal biopsy in most adults. The critical prerequisite is that the patient must be consuming gluten at the time of testing — at least 3 grams per day (roughly two slices of wheat bread) for a minimum of 6 weeks before serology, and 2 weeks before biopsy. Starting a gluten-free diet before testing is one of the most common reasons for false negative results.
First-Line Serology: tTG-IgA
Tissue transglutaminase IgA (tTG-IgA) is the recommended first-line screening test in both adults and children. It has a sensitivity of 93 to 96 percent and specificity of 96 to 98 percent for active coeliac disease with villous atrophy. A total serum IgA should be measured simultaneously because IgA deficiency — present in about 2 to 3 percent of coeliac patients (10 times the general population rate) — renders all IgA-based tests falsely negative.
For IgA-deficient patients, IgG-based alternatives are used: deamidated gliadin peptide IgG (DGP-IgG) or tTG-IgG. DGP-IgG has largely replaced the older anti-gliadin antibody tests due to superior specificity.
Endomysial Antibody (EMA)
EMA-IgA, detected by indirect immunofluorescence on primate oesophagus or human umbilical cord sections, has near-100 percent specificity for coeliac disease. It is more expensive and operator-dependent than tTG-IgA, so it is typically used as a confirmatory test when tTG-IgA is weakly positive or when biopsy-avoidance pathways are being considered.
Duodenal Biopsy and the Marsh Classification
The Marsh-Oberhuber classification grades histological severity. Marsh 0 is normal. Marsh 1 (infiltrative) shows increased IELs (more than 25 per 100 enterocytes) with normal villous architecture — this alone is not specific for coeliac disease. Marsh 2 (hyperplastic) adds crypt hyperplasia. Marsh 3 is subdivided: 3a (partial villous atrophy), 3b (subtotal), and 3c (total villous atrophy). At least four to six biopsies from the duodenal bulb and second portion are recommended, because coeliac disease can be patchy.
The No-Biopsy Pathway
Since 2012, ESPGHAN guidelines have allowed diagnosis without biopsy in symptomatic children who meet strict criteria: tTG-IgA levels more than 10 times the upper limit of normal, positive EMA on a separate blood sample, and HLA-DQ2 or DQ8 positivity. This pathway avoids endoscopy in clear-cut paediatric cases and has been validated in multiple prospective studies with positive predictive values exceeding 99 percent. Adult guidelines from BSG (2019) and ESPGHAN (2020) have cautiously extended this approach to selected adults, though biopsy remains the standard of care in most centres.
Pitfalls
Seronegative coeliac disease accounts for 2 to 5 percent of biopsy-proven cases — these patients have villous atrophy but negative serology, most commonly due to IgA deficiency, low-level gluten intake, or immunosuppressive medication. HLA-DQ2/DQ8 testing is useful here: a negative HLA result effectively excludes coeliac disease (negative predictive value above 99 percent) and can prevent unnecessary repeat biopsies.