Incidence and predictors of osteoporotic fractures in patients with Barrett's oesophagus: a population-based nested case-control study

Background: Proton pump inhibitors (PPI) are inconsistently associated with osteoporotic fractures. Barrett's oesophagus (BO) patients are treated with high PPI doses for prolonged periods, but there are limited data on the incidence of osteoporosis and fractures in this group pf patients. Aim: To estimate the incidence of (and risk factors for) low bone mass (osteoporosis and/or osteopenia) related fractures in a population-based BO cohort. Methods: All subjects with BO and a diagnosis of osteoporosis and fractures were identified using Rochester Epidemiology Project resources. The incidence rates of all and osteoporotic fractures in these subjects were compared to an age- and gender similar population in Olmsted County to determine standardised incidence ratios (SIR). Predictors were assessed using Cox proportional hazards models. Results: Five hundred and twenty-one patients were included (median [IQR] age 61 [52, 72] years; 398 [76%] men) of whom 113 (21.7%) had fractures, and 46 (8.8%) had osteoporotic fractures. The incidence of all fractures and osteoporotic fractures was comparable to that of an age- and gender-matched population (SIR 1.09; 95% CI 0.92-1.29: SIR 1.05; 95% CI 0.85-1.29). PPI use, dose or duration of use was not associated with osteoporotic fracture risk (HR 0.87; 95% CI 0.12-6.39). Independent risk factors for osteoporotic fractures included older age, female gender and higher co-morbidity index. Conclusions: The incidence of osteoporotic fractures was not increased in BO patients compared to the general population. In addition, PPI use was not associated with increased fracture risk regardless of the duration of therapy or dose.