Background:
Celiac disease (CD) contributes to secondary osteoporosis and increased fracture risk, yet most affected adults remain undiagnosed. CD affects approximately 1-2% of the general population, raising the possibility that fracture clinics could offer a pragmatic opportunity for case-finding. However, the diagnostic yield and real-world utility of routine CD screening in this setting are unknown.
Methods:
In a prospective study at a UK fracture clinic (April 2023-July 2025), ambulatory adults presenting with recent bone fractures (≤3 months) were screened using IgA-tissue transglutaminase (IgA-tTG). Individuals with known CD were not retested. IgA-tTG-positive participants were referred for duodenal biopsies. Intention-to-test analyses retained all enrolled participants in the denominator; missing samples, blood-processing errors, and known CD were classified as seronegative. Equivocal IgA-tTG results were reviewed by a consultant gastroenterologist and subsequently classified as negative. Per-protocol analyses included only participants with valid serology and no protocol deviations. Prevalence estimates were calculated using Wilson 95% confidence intervals (CIs).
Results:
Among 1200 participants, 9 were newly serology-positive, yielding a diagnostic (seropositive) prevalence of 0.75% (95% CI 0.34-1.42%). Six participants had biopsy-confirmed CD, giving a true diagnostic yield of 0.50% (95% CI 0.23-1.09%). Biopsy uptake among serology-positive individuals was 100%. The positive predictive value of IgA-tTG was 66.7% (6/9). Eight participants (0.7%) reported pre-existing CD, resulting in an overall CD prevalence of 1.17% (0.70-1.95%). PP analyses showed similar results (0.70% new seropositive; 0.43% biopsy-confirmed), indicating minimal influence of missing samples or protocol deviations.
Conclusion:
In this large prospective cohort of ambulatory adults with recent fractures, universal IgA-tTG screening identified few new cases of CD and did not exceed the expected 1-2% background prevalence. Routine screening at the point of fracture care therefore offers limited diagnostic value. These findings support maintaining selective testing based on established clinical indications rather than adopting universal screening in fracture services.