Do PPIs Increase Clostridium Difficile Infection Risk?
They are associated with it. This 2017 systematic review and meta-analysis in the World Journal of Gastroenterology pooled 56 observational studies, 40 case-control and 16 cohort, covering 356,683 adults, and found the risk of Clostridium difficile infection almost two times higher in PPI users than in non-users. Every study in the pool was observational, so the authors state plainly that causality could not be confirmed.
Dr. Kumar’s Take
An association this consistent is worth taking seriously, even though it is not proof. The pooled estimate held up in every subgroup the authors tested: case-control and cohort, adjusted and unadjusted, single-center and multicenter, older and younger patients. That kind of consistency is hard to wave away. But consistency is not causation, and the people who get put on PPIs are not the same people who do not get put on them. The authors say further high-quality prospective studies are needed to answer the causal question, and I agree with them. My practical position has not changed: prescribe a PPI when there is a clear indication for one, and reconsider it when there is not.
What the Research Shows
The authors searched MEDLINE/PubMed and seven other databases from January 1990 through March 2017 for studies evaluating the association between PPIs and C. difficile infection. Adult case-control and cohort studies were eligible. Pooled odds ratios with 95% confidence intervals were calculated using a random effects model, heterogeneity was assessed with the I² test and Cochran’s Q statistic, publication bias was evaluated by funnel plot, and study quality was scored with the Newcastle-Ottawa scale.
Fifty-six studies met the inclusion criteria. No clinical trials were included, and no randomized evidence was available to pool.
Study Snapshot
Fifty-six studies, 40 case-control and 16 cohort, involving 356,683 patients. Published between 1990 and March 2017. All observational. Adults only.
Results in Real Numbers
- All studies combined: pooled OR 1.99 (95%CI: 1.73-2.30, P < 0.001) for PPI users versus non-users
- Case-control studies: OR 2.00 (95%CI: 1.68-2.38, P < 0.0001)
- Cohort studies: OR 1.98 (95%CI: 1.51-2.59, P < 0.0001)
- Adjusted estimates: OR 1.95 (95%CI: 1.67-2.27, P < 0.0001)
- Unadjusted estimates: OR 2.02 (95%CI: 1.41-2.91, P < 0.0001)
- Single-center studies: OR 2.18 (95%CI: 1.72-2.75, P < 0.0001)
- Multicenter studies: OR 1.82 (95%CI: 1.51-2.19, P < 0.0001)
- Age 65 and older: OR 1.93 (95%CI: 1.40-2.68, P < 0.0001)
- Under age 65: OR 2.06 (95%CI: 1.11-3.81, P < 0.01)
Tests for differences between subgroups were not significant: P = 0.93 for case-control versus cohort, P = 0.85 for adjusted versus unadjusted, P = 0.24 for single-center versus multicenter, P = 0.86 for the two age strata.
Safety, Limits, and Caveats
Heterogeneity across studies was substantial: I² = 85.4%, P < 0.001. The studies did not agree closely with each other, which widens the honest uncertainty around the pooled figure even though the direction of effect was consistent.
There was also evidence of publication bias. The funnel plot asymmetry test came back at P = 0.002, which is the pattern you see when smaller studies reporting no association are less likely to have been published. That tends to inflate a pooled estimate.
And the design ceiling matters most. All 56 studies were observational. Patients who are prescribed PPIs differ from those who are not, often in ways that independently raise infection risk. The authors’ own conclusion is that this meta-analysis provides further evidence of an association and that prospective studies are needed to assess whether it is causal.
Practical Takeaways
- Prescribe PPIs only where they are clearly indicated. That is the authors’ own recommendation to clinicians.
- Treat the association as a reason to review whether a PPI is still needed, not as a reason to stop one that is doing necessary work.
- Be aware of the potential association when a patient on a PPI develops diarrhea, and evaluate accordingly.
- Reassess long-standing PPI prescriptions that were started for a problem that has since resolved.
Related Studies and Research
- Role of Gastric Acid in Preventing Foodborne Disease
- Pharmacology of Proton Pump Inhibitors
- Trends in Use of Proton Pump Inhibitors Among Adults in the United States
- Acute Interstitial Nephritis Due to Proton Pump Inhibitors
- Episode 25: The Great GERD Mistake - How Medicine Made Heartburn Worse and How to Fix It
FAQs
How large is the association between PPIs and C. difficile infection?
The pooled odds ratio across all 56 studies was 1.99 (95%CI: 1.73-2.30), which the authors describe as a risk almost two times higher in PPI users than in non-users.
Does this study prove PPIs cause C. difficile infection?
No. Every study included was observational, and the authors state that causality could not be confirmed. They call for further high-quality prospective studies to answer that question.
Was the association weaker in older patients?
No. The odds ratio was 1.93 (95%CI: 1.40-2.68) for patients 65 and older and 2.06 (95%CI: 1.11-3.81) for those under 65, and the test for a difference between the two groups was not significant at P = 0.86.
How reliable is the pooled number?
It carries real caveats. Heterogeneity between studies was high at I² = 85.4%, and the funnel plot asymmetry test suggested publication bias at P = 0.002.
Should I stop my PPI because of this?
That is a decision to make with your own physician, weighing why the PPI was started against your individual risk. This analysis cannot tell you whether stopping the drug changes your risk, because it could not establish cause.
Bottom Line
Across 56 observational studies and 356,683 patients, PPI use was associated with roughly double the odds of Clostridium difficile infection, pooled OR 1.99 (95%CI: 1.73-2.30). The association was consistent across every subgroup tested, but the evidence is observational, heterogeneity was high, and publication bias was detected, so this establishes an association rather than a cause. The authors’ recommendation to clinicians is to prescribe PPIs only where clearly indicated.
