Background and aims: A lifelong gluten-free diet (GFD) is standard treatment for coeliac disease (CD), but responses to gluten exposure are heterogeneous. We evaluated mucosal damage and anti-transglutaminase IgA (anti-tTG IgA) status in adults with voluntary GFD non-adherence.
Methods: This retrospective single-centre cohort included adults with guideline-confirmed CD reporting intermittent gluten consumption. Gluten exposure was assessed by an expert dietitian using a validated modified food-frequency questionnaire covering 12 months. It recorded frequency, timing, meal context and quantity of gluten-containing foods and included open-ended questions to capture occasional exposure. Foods were grouped into five predefined categories. Weekly gluten intake was estimated using food-specific conversion factors and expressed as grams/week. Duodenal histology was classified according to Marsh–Oberhüber criteria: Marsh 0–2 was non-atrophic and Marsh 3a–3c atrophic. Anti-tTG IgA was classified as elevated or non-elevated.
Results: Seventy-four patients were included; 51 (68.9%) were women. Median follow-up was 20.5 years, duration of non-adherence 17 years and gluten intake 1.52 g/week. Overall, 41 patients (55.4%) were non-atrophic and 33 (44.6%) had villous atrophy. Anti-tTG IgA was elevated in 23/74 patients (31.1%) and was less frequent in non-atrophic than atrophic patients (19.5% vs 45.5%; p=0.023). Thirty-three patients (44.6%) had both non-elevated anti-tTG IgA and no villous atrophy. Non-atrophic patients had longer GFD non-adherence than atrophic patients (18 vs 14 years; p=0.048), despite comparable gluten intake. Symptoms, BMI, malabsorption-related abnormalities and bone impairment did not differ significantly.
Conclusions: More than half of adults with longstanding intermittent gluten exposure had no villous atrophy, and nearly half had both absence of villous atrophy and non-elevated anti-tTG IgA. These findings suggest a temporarily remitting or gluten-tolerant phenotype. Prospective multicentre studies should confirm its stability, identify predictors and define safe monitoring before controlled GFD liberalisation is considered.