Do proton pump inhibitors increase upper gastrointestinal cancer risk?

The study did not find the temporal pattern expected if proton pump inhibitors were major long-term causes of upper gastrointestinal cancer. Apparent risk was strongest just before diagnosis, when early cancer symptoms may prompt acid-suppressing treatment. Earlier exposure moved toward neutral or lower risk after accounting for timing and gastrointestinal conditions. Persistent or new symptoms still require clinical evaluation.

Bottom line: the medication may appear associated with cancer because early, undiagnosed cancer caused the symptoms that led to treatment. This is called reverse causation.

Why earlier studies could be misleading

Proton pump inhibitors (PPIs) are often prescribed for reflux, abdominal pain, ulcers, gastritis, and other upper gastrointestinal symptoms. Some of those symptoms can also be early manifestations of cancer.

If exposure is counted without regard to timing, prescriptions written during the diagnostic lead-up can make the medicine look like the cause of a disease that was already developing.

How the study handled timing

The matched case-control analysis separated PPI exposure into multiple periods before cancer diagnosis and included those windows in the same model. It also considered gastrointestinal diagnoses that could explain why treatment was prescribed.

The strongest apparent associations occurred in the months immediately before diagnosis. By contrast, exposure one to ten years earlier shifted toward neutral or reduced risk. That is not the pattern expected if cumulative PPI exposure were a major driver of cancer development.

What the result does and does not mean

The findings are reassuring about a large causal effect from appropriately used PPIs, but they do not prove that every duration, dose, or clinical situation is risk-free. Observational analyses can still have misclassification and residual confounding.

The result also contains an important clinical warning: symptom relief does not explain the cause of a symptom. New, persistent, progressive, or otherwise concerning gastrointestinal symptoms should be evaluated rather than indefinitely masked.

Medication should not be stopped abruptly or continued unnecessarily on the basis of a web page. The indication, dose, duration, and alternatives belong in a clinician–patient review.

Primary sources