Document resource
Introduction Histological severity at diagnosis of coeliac disease (CD) is heterogeneous. Identifying simple clinical predictors of extensive mucosal damage may help improve diagnostic assessment and risk stratification at presentation.Methods We conducted a retrospective observational cohort study including 237 adult patients with biopsy-confirmed coeliac disease diagnosed between 1995 and 2025 in a tertiary referral centre. Duodenal villous atrophy was assessed on diagnostic biopsies. The primary outcome was total duodenal villous atrophy. Candidate predictors available at diagnosis included age, sex, abdominal pain, diarrhoea, macroscopic duodenal findings on upper gastrointestinal endoscopy, and anti–tissue transglutaminase (tTG) status. Associations were assessed using univariate and multivariable logistic regression.Results Duodenal villous atrophy data were available in 230/237 patients (97.0%), among whom 133 (57.8%) had total villous atrophy. Complete-case multivariable analysis included 194 patients. In univariate analysis, younger age (OR 0.98 per year; p = 0.043), abdominal pain (OR 1.93; p = 0.025), and diarrhoea (OR 2.06; p = 0.021) were associated with total villous atrophy.In multivariable analysis adjusted for age, sex, macroscopic endoscopic findings, and anti-tTG status, diarrhoea at diagnosis remained independently associated with total duodenal villous atrophy (OR 2.33; 95% CI 1.21–4.48; p = 0.012), while abdominal pain showed a borderline association (OR 1.85; 95% CI 1.00–3.43; p = 0.050). Sex, endoscopic duodenal findings, and anti-tTG positivity were not independently associated. Model discrimination was modest (AUC = 0.66).Conclusions In this real-world cohort, diarrhoea at diagnosis was the only independent clinical predictor of total duodenal villous atrophy, the most extensive form of mucosal damage in coeliac disease. These findings suggest that simple clinical features may help identify patients at higher risk of severe histological involvement at presentation and reinforce the need for systematic duodenal biopsies regardless of endoscopic appearance.