- Proton pump inhibitors (PPIs) were associated with both all-cause and cause-specific mortality.
- However, the change in HRs by increasing adjustment and between comparator groups indicate that residual confounding is likely to explain the association between poor health outcomes and PPI use.
Why this matters
- PPIs are globally one of the most frequently used classes of drugs.
- Concern over the safety of PPIs has grown, given associations observed in non-interventional studies between PPI use and a large number of adverse health outcomes.
- This cohort study compared 733,885 new users of PPIs with 124,410 new users of H2 receptor antagonists (H2RAs) using data from the UK Clinical Practice Research Datalink (CPRD) GOLD database.
- Main outcomes: all-cause and cause-specific mortality.
- Funding: None.
- PPIs were associated with an increased risk of all-cause mortality, with HRs decreasing considerably by increasing adjustment (unadjusted HR, 1.65; 95% CI, 1.62-1.69; propensity score weighted HR [wHR], 1.38; 95% CI, 1.33-1.44).
- By more specific cause-of-death category, mortality was higher in PPI vs H2RA users from (wHR; 95% CI):
- neoplasms (1.74; 1.63-1.86);
- cardiovascular/circulatory causes (1.17; 1.10-1.25);
- chronic respiratory diseases (1.40; 1.22-1.62);
- liver cirrhosis (1.95; 1.10-3.46);
- digestive causes other than cirrhosis (1.43; 1.20-1.69); and
- diabetes, urogenital, blood and endocrine causes (1.27; 1.06-1.51).
- Short-term associations were seen with mortality from causes where a causal short-term association is unexpected (e.g. lung cancer mortality: wHR at 6 months, 1.77; 95% CI, 1.39-2.25).
- In secondary analysis, the risk of all-cause mortality was greater in PPI user (n=689,602) vs non-users of H2RAs (n=1,361,245) (wHR, 1.96; 95% CI, 1.94-1.99).
- Similarly, cause-specific mortality was significantly higher in PPI users vs non-users from (wHR; 95% CI):
- neoplasms (3.74; 3.64-3.84);
- liver cirrhosis (4.10; 3.36-5.01); and
- gastric cancer (14.59; 11.16-19.08).
- Risk of misclassification of acid suppression drug usage.
- Risk of residual confounding.
- Causality could not be established.