Eating Disorders in Individuals With Celiac Disease and Their Biological First-Degree Relatives

Importance Patients with celiac disease (CD) may face elevated risk of eating disorders, potentially due to genetic susceptibility and psychosocial factors of a gluten-free diet. However, up-to-date population-based evidence is scarce, and whether any excess risk extends to biological first-degree relatives (bFDRs) remains unclear. Objective To investigate the risk of eating disorders in patients with CD and their bFDRs, including parents, full siblings, and offspring. Design, Setting, and Participants A nationwide multigenerational cohort study of patients with celiac disease in Sweden, observed from January 2002 to December 2023. Data were analyzed from December 2025 to August 2026. From Sweden’s 28 histopathology departments, patients with biopsy-confirmed CD diagnosed 2002 to 2023 were identified. Each patient was matched with fewer than or equal to 5 general population reference individuals; bFDRs of patients with CD and reference individuals were also identified. Main Outcomes and Measures The primary outcome was any eating disorder. The adjusted hazard ratio (aHR) for eating disorders diagnosed at age 6 years or older were estimated using Cox regression and flexible parametric survival models for the overall and time-varying estimates, respectively. Results This study included 37 853 patients with CD (23 704 [62.6%] female; median [IQR] age at index date 28.7 [11.7-53.0] years ) and 185 159 reference individuals. Over 10.9 years median follow-up (IQR, 5.9-15.8 years), 485 patients with CD and 1578 reference individuals were diagnosed with any eating disorder. CD was associated with an increased hazard for eating disorders (aHR, 1.39; 95% CI,1.27-1.53), most apparent during the first 5 years of follow-up. Over 20 years of follow-up, 1 extra eating disorder case was expected per 222 patients with CD. In contrast, there was no hazard elevation for any eating disorder among CD-bFDRs (53 435 parents: aHR, 0.82 [0.61-1.12]; 48 743 siblings: aHR, 1.11 [0.99–1.24]; 38 695 offspring: aHR, 1.05 [0.92-1.20]). Sensitivity analyses mitigating information bias and residual confounding largely aligned with main findings. Conclusions and Relevance In this cohort study of patients with CD, a moderately increased eating disorders risk particularly during the first 5 years of follow-up was observed, although the absolute risk was low. This increased risk did not extend to bFDRs of patients with CD. Clinical vigilance toward disordered eating is important among patients with newly diagnosed CD.