Proton Pump Inhibitor Use and Incident Cardiovascular Disease in Older Postmenopausal Women

Background: Epidemiological studies have been inconsistent regarding an association between proton pump inhibitor (PPI) use and risk of primary cardiovascular disease (CVD) events. Methods: We studied 85,189 postmenopausal women (mean age 63 years at baseline) without known CVD at enrollment into the Women's Health Initiative Observational Study (1993–1998). PPI use was determined from medication inventories at baseline and Year-3. CVD events were physician adjudicated and defined as a composite of coronary heart disease, stroke, and CVD mortality. Follow up was from baseline to September 2010. Multivariable Cox proportional hazard models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for incident CVD according to baseline PPI use (no/yes), use duration (non-user, < 1 year, 1–3 years, > 3 years), and time-varying based on updated Year-3 information. Propensity score adjustment was used to control for residual confounding. Results: At baseline, 1747 (2.1%) women reported using PPIs. During a mean follow-up of 11 years, 5778 (6.8%) cases of primary CVD were identified. PPI users had significantly higher risk of CVD compared with non-users in the fully adjusted model (HR: 1.21, 95% CI: 1.02–1.43), and after propensity score adjustment (HR: 1.27, 95% CI: 1.21–1.32). Longer PPI use duration was associated with incrementally higher CVD risk (HRs: < 1 year: 1.11, 1–3 years: 1.27, > 3 years: 1.33; p for trend = 0.02). Conclusions: PPI use was associated with higher risk of incident primary CVD in older postmenopausal women. These findings underscore the importance of guideline-directed PPI use to avoid unwanted adverse events.