We investigated cardiovascular risk due to proton pump inhibitor (PPI) treatment using a self-controlled case series (SCCS) study design, a type of case-only design and an approach to overcome between-person confounding in which individuals act as their own control.
We conducted an SCCS study using the National Health Insurance Service-Health Screening cohort in Korea (2002–2015). The cohort included 303,404 adult participants without prior cardiovascular events, who were followed up until December 2015. The primary outcome was a composite of stroke or myocardial infarction. The SCCS method estimated the age-adjusted incidence rate ratio between periods with and without exposure to PPI among patients with primary outcomes. As sensitivity analysis, conventional multivariable Cox proportional regression analyses were performed, which treated the exposure to PPI and H2 blocker during follow-up as time-dependent variables.
In the SCCS design, 10,952 (3.6%) patients with primary outcomes were included. There was no association between PPI exposure and primary outcome (incidence rate ratio 0.98, 95% confidence interval [CI] 0.89–1.09). In the time-dependent Cox regression analyses, both PPI (adjusted hazard ratio 1.36, 95% CI 1.24–1.49) and H2 blocker (adjusted hazard ratio 1.46, 95% CI 1.38–1.55) were associated with an increased risk of the primary outcome.
Negative findings in the SCCS design suggest that association between increased cardiovascular risk and PPI, frequently reported in prior observational studies, is likely due to residual confounding related to conditions with PPI treatment, rather than a true relationship.