Association between proton pump inhibitors use and risk of hip fracture: A general population-based cohort study.

Abstract

Previous studies comparing risk of fracture among Proton Pump Inhibitors (PPIs) users with that among non-users were susceptible to confounding by indication. Epidemiologic studies of this association using an active medication as a comparator would appropriately address this bias. We conducted a propensity-score matched cohort study to compare the risk of incident fracture over five years among initiators of PPIs with initiators of histamine 2 receptor antagonist (H2RA) using data collected from The Health Improvement Network (THIN) database in the United Kingdom (2000-2016). The prevalence of PPIs prescription increased 3.8-fold from 2000 (7.9%) to 2012 (30.3%) and remained stable since then. Of the 50,265 propensity-score matched participants in each cohort (mean age was 65 years, and 57% were women), 370 (1.9/1000 person-years) incident hip fracture occurred in the PPIs initiators and 294 (1.5/1000 person-years) in the H2RA initiators during the follow-up period. The rate difference of incident hip fracture for PPIs initiation was 0.4 (95% confidence interval [CI]: 0.2-0.7)/1000 person-years compared with H2RA initiation and the corresponding hazard ratio (HR) was 1.27 (95%CI: 1.09-1.48). Compared with non-PPI use, multivariable-adjusted odds ratios (ORs) of hip fracture were 1.17 (95%CI: 0.94-1.46), 1.55 (95%CI: 1.23-1.96), and 1.67 (95%CI: 1.33-2.10) for 1-2, 3-9 and ≥ 10 prescriptions of PPIs, respectively (P for trend = 0.001). We found that PPIs prescription has been increasing rapidly over the past decade in the United Kingdom, and the initiation of PPIs was associated with a higher risk of hip fracture than initiation of H2RA.