Coeliac Disease

Meet the patient

A 34-year-old woman presents with eighteen months of fatigue and an iron-deficiency anaemia that will not correct on oral iron. She has no gastrointestinal symptoms, her cycles are normal, and she has been told it is "just low iron from her periods." Her ferritin is 6, she is positive for anti-tissue transglutaminase IgA, and her duodenal biopsies show villous atrophy with crypt hyperplasia and intraepithelial lymphocytosis.^[1][2]

This is the face of modern coeliac disease — not the wasted child with steatorrhoea, but the well-looking adult with an unexplained anaemia. The single question that decides her diagnosis is the one that decides every coeliac work-up: was the serology and biopsy done while she was still eating gluten? Hold that question and the whole topic falls into place.^[1]

The coeliac iceberg — one in a hundred, most still hidden

Coeliac disease is common, immune-mediated, and vastly under-diagnosed. The true global prevalence is about 1 per cent — the Singh 2018 meta-analysis of 432 studies put seroprevalence at 1.4 per cent and biopsy-confirmed disease at 0.7 per cent — yet for every patient diagnosed, several more sit beneath the waterline.^[3]

This is the coeliac iceberg: classical malabsorption is the visible tip; the submerged mass is iron-deficiency anaemia, osteoporosis, transaminitis, dermatitis herpetiformis, infertility, and the silent cases found only by screening high-risk groups.^[2]

^[3]

Genetics is necessary but not sufficient. About 95 per cent of patients carry HLA-DQ2 (DQA105/DQB102) and most of the rest carry HLA-DQ8 — yet up to 30–40 per cent of the general population carry the same alleles and never develop the disease. Concordance runs about 75 per cent in monozygotic twins, 30 per cent in HLA-identical siblings, and 10 per cent in first-degree relatives, which is why HLA is a rule-OUT test, never a rule-in.^[1][2]

Etymology for viva gold: coeliac is from the Greek koiliakē, "of the belly" — Aretaeus of Cappadocia described the "coeliac affection" in the second century AD. The word is nineteen centuries old; the mechanism took rather longer.^[2]

The molecular cascade — tTG deamidation is the pivot

The whole disease is a five-step cascade, and one enzyme — tissue transglutaminase 2 — is the pivot. Learn the cascade and you can derive the serology, the histology, and the diet in one line each.^[1][2]

Step 1 — gliadin resists digestion. The prolamin fraction of gluten (gliadin in wheat, hordein in barley, secalin in rye) is rich in proline and glutamine, so gastric, pancreatic and brush-border proteases cannot break it down. Large immunoreactive fragments — the 33-mer of alpha-2 gliadin — reach the lamina propria intact.^[1]

Step 2 — tTG2 deamidation is the pivotal event. In the lamina propria, tissue transglutaminase-2 converts glutamine to glutamate in the gliadin peptide. The new negative charge lets gliadin slot with high affinity into the HLA-DQ2/DQ8 binding groove. The same enzyme is the autoantigen — anti-tTG and anti-endomysial antibodies recognise tTG2 itself, which is why they are so diagnostically faithful.^[1]

Step 3 — HLA-DQ2/DQ8 presents to CD4+ T cells. Deamidated gliadin is presented to gliadin-specific CD4+ T-helper cells, which polarise to a Th1 phenotype secreting interferon-gamma. IFN-gamma and its cytokine cousins (IL-21, TNF-alpha) drive the epithelial damage, crypt hyperplasia, and lamina propria inflammation.^[1]

Step 4 — B cells make the antibodies that diagnose but do not injure. The same T-cell help switches gliadin-specific B cells to produce anti-tTG and anti-endomysial IgA. These antibodies are pathognomonic diagnostic markers — but they are not the primary injurer. The tissue damage is T-cell driven; the antibodies are the fingerprint, not the weapon.^[1]

Step 5 — IL-15, intraepithelial lymphocytes, and the road to EATL. Gliadin directly upregulates interleukin-15, which expands intraepithelial lymphocytes (IELs) into cytotoxic, NKG2D-expressing killers. Persistent IL-15 signalling is the substrate of refractory coeliac type 2, in which IELs turn clonal and malignant and progress to enteropathy-associated T-cell lymphoma (EATL).^[1]

Enlarge

Why the diet works — and why it must wait until after diagnosis. Remove gluten and deamidation stops, T-cell activation stops, and antibody titres fall within months while the mucosa heals over one to two years. Reintroduce gluten and the whole cascade reignites — which is exactly why serology and biopsy must be done on a full-gluten diet, or the work-up is destroyed.^[6]

Three masks, one disease — classical is now the minority

The malabsorptive picture of the textbooks is now the minority presentation. Most adult diagnoses today arrive through atypical, extraintestinal, or screening pathways, and a final-prof candidate who pictures only "diarrhoea and weight loss" will miss most of the disease.^[2]

^{[1] [2]}

The classic trap: reaching for "IBS" in a young woman with bloating and tiredness without first sending anti-tTG IgA plus a total IgA. Coeliac hides behind the IBS label in roughly 3–5 per cent of such patients — screen before you ever diagnose irritable bowel.^[1]

The bedside round — anaemia first, then the rash

Examination in suspected coeliac rarely hands you the diagnosis; it hands you the deficiencies, the associated autoimmunity, and — if you are lucky — the rash.^[1]

Run the round in this order:^[1]

• Growth and nutrition — BMI, wasting, failure to thrive, short stature, delayed puberty in children.
• Haematology — pallor, koilonychia, angular stomatitis, glossitis from iron and folate; bruising and petechiae from vitamin K or hyposplenic thrombocytopenia.
• Metabolic — tetany or proximal myopathy from hypocalcaemia and vitamin D deficiency.
• Abdomen — distension from gas and dilated loops; a palpable mass raises lymphoma.
• Skin and mouth — the extensor surfaces (elbows, knees, buttocks, scalp) for the itchy grouped vesicles of dermatitis herpetiformis; the mouth for aphthous ulcers and dental enamel defects.
• Autoimmune stigmata — thyroid goitre, vitiligo, arthralgia, the features of Down or Turner syndrome, signs of type 1 diabetes.^[1]

Everyone forgets: the blood film. Howell-Jolly bodies, target cells, thrombocytopenia and leucopenia mark functional hyposplenism — common in coeliac, and the reason these patients are vulnerable to fulminant encapsulated-organism sepsis. Order the film, and act on it with vaccination.^[1]

The serology ladder — tTG screens, EMA confirms, HLA rules out

Three tests do almost all of the diagnostic serology, and each has one job. Anti-tTG IgA screens (most sensitive), EMA confirms (most specific), and HLA-DQ2/DQ8 rules out (high negative predictive value). Run them — every one of them — while the patient is still on a full-gluten diet.^[1]

^[1]

The total IgA is non-negotiable. Selective IgA deficiency overlaps with coeliac in about 3 per cent of patients (and is 5–10 times more common in coeliac than in the general population). Send an IgA-based screen alone in an IgA-deficient patient and you get a falsely negative result — which is why the total IgA travels with every anti-tTG request, and anti-DGP IgG takes over when IgA is low.^[1]

Marsh 3 closes the case — and the bulb biopsy saves it

The duodenal biopsy is the gold standard, Marsh 3 is the diagnostic lesion, and the bulb biopsy is the detail that stops you missing patchy disease. Take at least six biopsies — four from the distal duodenum and two from the duodenal bulb — because coeliac can be patchy and bulb-only.^[1][6]

The Marsh-Oberhuber grades every duodenal biopsy is reported against:^[1][2]

^[1]

Marsh 3 (a, b, or c) is the diagnostic flat lesion — villous atrophy with crypt hyperplasia and IEL over 30 per 100 enterocytes. Marsh 1 — IEL alone — is non-specific: Helicobacter pylori, NSAIDs, infections, and inflammatory bowel disease all raise IEL counts, so never diagnose coeliac on Marsh 1 alone.^[1]

Enlarge

Enlarge

^{[1] [11]}

The gluten-free diet — the only cure, and it is for life

There is no drug that cures coeliac disease. The strict, lifelong gluten-free diet is the entire definitive treatment. Everything else — deficiency correction, vaccination, monitoring — is layered onto that one non-negotiable foundation.^[1][2]

Enlarge

^[1]

What "gluten-free" actually means:^[1]

• Exclude completely — wheat (including durum, semolina, spelt, kamut, couscous), barley, rye, triticale. All sources: bread, pasta, biscuits, cakes, flour-thickened sauces, beer, lager, ale.
• Oats are tolerated by most if pure and uncontaminated; introduce cautiously and monitor serology.
• The legal threshold is under 20 ppm (Codex Alimentarius, FDA, UK and EU law) — that is "gluten-free".
• Hidden gluten is everywhere — processed foods, soups, sausages, soy sauce, stock cubes, communion wafers, lipsticks, some medications. Teach label-reading.
• Cross-contamination — separate toasters, utensils, cooking oil, and workspace. The diet treats the cause; this layer treats the damage already done. Every newly diagnosed patient gets a deficiency screen, replacement where deficient, pneumococcal vaccination, and an assessment of bone health — none of it is optional.^[1][11]

Screen and replace what is missing — by agent:^[1]

• Iron deficiency — replace iron orally; reserve parenteral iron for severe malabsorption or intolerance of oral iron, and recheck ferritin on treatment.

• Folate deficiency — replace folate orally.

• Vitamin B12 deficiency — screen and replace B12 if low. B12 deficiency is uncommon in coeliac because the ileum is usually spared, but screen and treat.

• Calcium and vitamin D — bone disease is a recognised manifestation of coeliac disease; make sure calcium and vitamin D intake is adequate.^[1] Pneumococcal risk and bone:^[11]

• Vaccinate against pneumococcal infection — coeliac disease carries an increased risk of pneumococcal infection, so preventive pneumococcal vaccination should be considered. Counsel the patient to present urgently with fever.

• Bone disease is a recognised manifestation — assess bone health at diagnosis and treat osteoporosis along the standard pathway, with adequate calcium and vitamin D.^[1]

Annual review — the bundle that keeps the patient safe:^[1]

• Symptoms, BMI, growth in children, and adherence.
• Annual bloods — FBC, ferritin, folate, vitamin B12, calcium, vitamin D, LFTs, TSH, and anti-tTG IgA titre as the adherence marker.
• Repeat DEXA every 2–5 years as guided by baseline.
• Repeat biopsy is NOT routine if responding well; re-biopsy only for non-response, refractory disease, or serology-histology discordance.^[1]

Dermatitis herpetiformis — coeliac on the skin

Dermatitis herpetiformis is the cutaneous face of coeliac disease, and virtually every patient with it has coeliac-type enteropathy. It is the one skin rash that should trigger a coeliac work-up on sight.^[1]

The rash is intensely itchy, symmetric, grouped vesicles and crusts on the extensor surfaces — elbows, knees, buttocks, scalp, sacrum. It affects about 10–25 per cent of coeliac patients, and the enteropathy beneath is often mild (Marsh 1–2).^[1]

The two procedural rules that examiners test:^[12]

• Biopsy perilesional skin, NOT the lesion itself — direct immunofluorescence showing granular IgA deposits at the dermal-epidermal junction is the most specific diagnostic test for dermatitis herpetiformis.
• Treat with a strict gluten-free diet plus dapsone. The diet treats the enteropathy and the skin; dapsone brings the rash and itch under control quickly while the diet takes effect.^[12]

Check G6PD before the first dapsone dose — dapsone causes haemolysis and methaemoglobinaemia in deficient patients — and monitor FBC and LFTs on treatment. The pearl that confirms the diagnosis at the bedside: dapsone relieves the itch within 24–48 hours.^[1]

The seronegative atrophy trap — olmesartan, tropical sprue, and friends

Villous atrophy is not synonymous with coeliac. When the serology is negative but the biopsy shows flat mucosa, the diagnosis is a wide-open question — and the two traps examiners love are both drug-and-region stories.^[2][8]

^{[7] [8]}

The drug-history trap: olmesartan. A patient on an ARB with severe watery diarrhoea and villous atrophy but negative serology has olmesartan sprue-like enteropathy until proven otherwise — stop the drug, and the enteropathy resolves. Lebwohl and Ludvigsson named this in 2014, and it remains the single most-tested seronegative mimic.^[7]

The regional mimic: tropical sprue. In a patient from or living in the tropics, both B12 and folate are low (coeliac lowers folate but usually spares B12, because the ileum is spared), the villous atrophy is partial with less crypt hyperplasia and fewer IELs, and the disease responds to tetracycline plus folic acid. Sharma 2019 is the head-to-head comparison to cite.^[8]

A confident coeliac diagnosis needs four things together: consistent serology, characteristic histology, an HLA-DQ2 or DQ8 background, and a clinical and histological response to gluten withdrawal. Any one alone is not enough.^[2]

Refractory coeliac and EATL — the road to the lethal complication

Most coeliac patients heal on a strict gluten-free diet. The ones who do not are on the road to the disease's lethal complication, and the fork in that road is the type of refractory coeliac.^[2]

Refractory coeliac disease (RCD) is persistent or recurrent malabsorptive symptoms with villous atrophy despite a strict gluten-free diet, after you have excluded ongoing gluten exposure (dietitian-led) and the other causes of atrophy above. Flow cytometry, immunohistochemistry and T-cell-receptor rearrangement studies separate the two types on the IEL phenotype:^[10]

• Type 1 RCD — a normal (polyclonal) intraepithelial lymphocyte population. Better prognosis.
• Type 2 RCD — an aberrant, clonal IEL population. This is pre-malignant: type 2 runs on to enteropathy-associated T-cell lymphoma (EATL), so image the small bowel with capsule endoscopy plus CT or MR enterography at diagnosis. Both types get corticosteroids — most commonly open-capsule budesonide, or prednisone where budesonide is unavailable — as first-line therapy, and patients who do not respond to steroids need referral to a specialist centre or a clinical trial.^[10]

EATL is the classic coeliac lymphoma — a T-cell (not B-cell) lymphoma arising from intraepithelial lymphocytes in RCD2. It presents as relapse of malabsorption, abdominal pain, fever, weight loss, perforation or obstruction, and the prognosis is poor. Any coeliac patient with new weight loss, fever or abdominal pain on a strict diet needs urgent re-biopsy and imaging.^[2]

Small-bowel adenocarcinoma is the other feared malignancy — risk markedly elevated in coeliac, presenting with obstruction, bleeding or weight loss. The Emilsson 2020 nationwide cohort quantified the small-bowel adenocarcinoma, adenoma and carcinoid risk in coeliac, and a strict gluten-free diet substantially reduces malignancy risk toward baseline, especially after five years of adherence.^[5]

Who you screen and why — special populations

Screening is bidirectional: coeliac travels with a list of autoimmune and chromosomal conditions, and you test for both directions. These are the groups an examiner expects you to name, and each has its own nuance.^[1]

• Type 1 diabetes — coeliac in 3–8 per cent; screen at diagnosis and every 1–2 years, or with new glycaemic instability.
• Autoimmune thyroid disease, autoimmune hepatitis, primary biliary cholangitis, Addison's disease — screen bidirectionally.
• Down syndrome, Turner syndrome, Williams syndrome — screen at diagnosis.
• First-degree relatives — 10 per cent prevalence; screen with anti-tTG IgA plus total IgA from age 3. An HLA-DQ2/DQ8-negative result rules out coeliac for life.
• Selective IgA deficiency — about 3 per cent of coeliac; use anti-DGP IgG, never IgA-based serology alone.
• Children — failure to thrive, delayed puberty, short stature, distended abdomen, irritability, dental enamel defects; the ESPGHAN 2020 no-biopsy pathway applies when criteria are met.
• Pregnancy — untreated coeliac causes infertility, recurrent miscarriage, intrauterine growth restriction, preterm delivery and low birth weight; screen at-risk women and enforce a strict gluten-free diet pre-conception and through pregnancy.^[6][1]

The mortality question — and the thyroid nodule. Lebwohl's JAMA 2020 cohort found a small but real increase in all-cause mortality (hazard ratio about 1.21 over a median 12 years), driven by cardiovascular disease, malignancy and respiratory disease, and strict adherence reduces it substantially. Coeliac patients also carry a higher risk of thyroid carcinoma — the Ludvigsson 2013 nationwide cohort is the reference to cite, which is why TSH sits on the annual panel and any thyroid nodule is investigated.^[4][9]

The mantra, and the mnemonic

^[1]

The mantra: serology and biopsy on full gluten — then, and only then, the diet.^[1][6]

Ward-round test — three stems, thirty seconds each

Stem 1 — the woman from the top of the topic (answer)ShowHide

The 34-year-old with fatigue and iron-deficiency anaemia refractory to oral iron. Anti-tTG IgA is positive, total IgA is normal, and duodenal biopsies show Marsh 3. What is the diagnosis, and what is the first thing you say to her? Model: This is coeliac disease presenting as isolated iron-deficiency anaemia — a common adult presentation. The first thing you say is: do not start a gluten-free diet yet — the serology and biopsy are already done on a full-gluten diet, so the diagnosis is secure, but confirm with EMA and HLA if serology-histology discordance arises, correct her iron deficiency, check folate, B12, calcium, vitamin D and TSH, assess bone health, consider pneumococcal vaccination, and only then start the strict lifelong gluten-free diet with dietitian and Coeliac Society support.^[1][11]

Stem 2 — the seronegative atrophy (answer)ShowHide

A 62-year-old man on olmesartan for hypertension has severe watery diarrhoea and weight loss. Anti-tTG IgA, total IgA and EMA are all negative; duodenal biopsies show villous atrophy. What is the likely diagnosis, and what do you do? Model: This is olmesartan sprue-like enteropathy — the classic seronegative enteropathy trap. The drug history is the clue, the negative serology is the pattern, and the test of time is the response. Stop the olmesartan (switch antihypertensive class with the GP), and the enteropathy resolves over weeks. Do not commit him to a lifelong gluten-free diet on negative serology.^[7]

Stem 3 — the relapse on a strict diet (answer)ShowHide

A 48-year-old with biopsy-proven coeliac, previously well-controlled, presents with new weight loss, abdominal pain and fever despite a strict gluten-free diet for years. Anti-tTG has risen. What are you worried about, and what is your next step? Model: This is refractory coeliac disease with possible EATL until proven otherwise. Confirm adherence with a dietitian-led review, re-biopsy the duodenum with flow cytometry and TCR clonality of the IELs (looking for type 2 RCD — loss of surface CD8), and image with capsule endoscopy or CT enterography for ulcerative jejunoileitis or lymphoma. Type 2 RCD is pre-malignant and EATL is the lethal complication — this is an urgent specialist-centre referral, not a routine outpatient review.^[2][5]

References12ShowHide

1. [1]Rubio-Tapia A, Hill ID, Kelly CP, Calderwood AH, Murray JA. ACG clinical guidelines: diagnosis and management of celiac disease Am J Gastroenterol, 2013.PMID 23609613
2. [2]Lebwohl B, Sanders DS, Green PHR. Coeliac disease Lancet, 2018.PMID 28760445
3. [3]Singh P, Arora A, Strand TA, et al. Global Prevalence of Celiac Disease: Systematic Review and Meta-analysis Clin Gastroenterol Hepatol, 2018.PMID 29551598
4. [4]Lebwohl B, Green PHR, Söderling J, Roelstraete B, Ludvigsson JF. Association Between Celiac Disease and Mortality Risk in a Swedish Population JAMA, 2020.PMID 32259229
5. [5]Emilsson L, Semrad C, Lebwohl B, et al. Risk of Small Bowel Adenocarcinoma, Adenomas, and Carcinoids in a Nationwide Cohort of Individuals With Celiac Disease Gastroenterology, 2020.PMID 32679218
6. [6]Husby S, Koletzko S, Korponay-Szabó IR, Kurppa K, Mearin ML, Ribes-Koninckx C. European Society Paediatric Gastroenterology, Hepatology and Nutrition Guidelines for Diagnosing Coeliac Disease 2020 J Pediatr Gastroenterol Nutr, 2020.PMID 31568151
7. [7]Lebwohl B, Ludvigsson JF Editorial: sprue-like enteropathy due to olmesartan and other angiotensin receptor blockers--the plot thickens Aliment Pharmacol Ther, 2014.PMID 25303379
8. [8]Sharma P, Baloda V, Gahlot GP, et al. Clinical, endoscopic, and histological differentiation between celiac disease and tropical sprue: A systematic review J Gastroenterol Hepatol, 2019.PMID 30069926
9. [9]Ludvigsson JF, Lebwohl B, Kämpe O, et al. Risk of thyroid cancer in a nationwide cohort of patients with biopsy-verified celiac disease Thyroid, 2013.PMID 23281890
10. [10]Green PHR, Paski S, Ko CW, Rubio-Tapia A. AGA Clinical Practice Update on Management of Refractory Celiac Disease: Expert Review Gastroenterology, 2022.PMID 36137844
11. [11]Simons M, Scott-Sheldon LAJ, Risech-Neyman Y, Moss SF, Ludvigsson JF, Green PHR. Celiac Disease and Increased Risk of Pneumococcal Infection: A Systematic Review and Meta-Analysis Am J Med, 2018.PMID 28801224
12. [12]Antiga E, Maglie R, Quintarelli L, Verdelli A, Bonciani D, Bonciolini V, Caproni M. Dermatitis Herpetiformis: Novel Perspectives Front Immunol, 2019.PMID 31244841