Proton-Pump Inhibitors Versus H2-Receptor Antagonists on the Risk of Major Osteoporotic Fractures: A Global Propensity-Score-Matched Cohort Study

Danling Jiang, Xi Li, W. Susan Cheng, Bing Li, Guoqing Lv, Ying LiResearch Square 2025 · DOI

This research paper, a large-scale retrospective cohort study, investigates the long-term skeletal safety of two common classes of acid-suppressing drugs. By comparing over 260,000 patients globally, the authors demonstrate that Proton-Pump Inhibitors (PPIs) carry a significantly higher risk of major osteoporotic fractures compared to H2-Receptor Antagonists (H2RAs), with a particularly pronounced effect in Asian populations.

TL;DR

Long-term use of Proton-Pump Inhibitors (PPIs) is associated with a 23% higher risk of major osteoporotic fractures compared to H2-receptor antagonists (H2RAs). This risk jumps to 77% in Asia-Pacific populations, suggesting that clinicians should favor H2RAs or shorter PPI durations for patients at high risk of bone loss.

What Problem Does This Paper Address?

Millions of people worldwide take acid-suppressive medications like omeprazole (a PPI) or famotidine (an H2RA) for gastro-esophageal reflux disease (GERD). While PPIs are highly effective, concerns have simmered for years regarding their impact on bone health.

The challenge in previous research has been "confounding by indication": do PPI users have more fractures because of the drug, or because they are generally sicker than people who don't need acid suppression? Furthermore, most data comes from Western populations. This paper addresses these gaps by using an active comparator (comparing two drugs used for the same condition) and focusing specifically on the understudied East-Asian population, where dietary calcium intake and drug metabolism (via the CYP2C19 enzyme) differ significantly from Western norms.

Methods

The researchers conducted a retrospective cohort study using the TriNetX Global Collaborative Network, a massive database of electronic health records.

To ensure a fair "head-to-head" comparison, they employed Propensity Score Matching (PSM). They identified adults who had received at least 12 prescriptions for either a PPI or an H2RA between 2015 and 2025. They then matched 131,698 PPI users to 131,698 H2RA users based on age, sex, race, and a wide array of comorbidities (like diabetes and kidney disease).

Crucially, they included a negative control outcome—malignant melanoma. Since there is no biological reason for acid blockers to cause skin cancer, a lack of correlation here serves as a "sanity check" that their statistical matching successfully removed general health biases.

Datasets

The study utilized the TriNetX Global Collaborative Network, which aggregates de-identified data from 147 healthcare organizations across North America, Europe, the Middle East, and the Asia-Pacific (APAC) region. This provided a total sample of over two million person-years of follow-up data.

Key Insights and Ideas

  • The "Drug-Class Specific" Toxicity: Because the increased fracture risk persisted even when comparing PPIs to another acid-suppressor (H2RAs), the authors argue the damage isn't just from low stomach acid. Instead, PPIs likely cause specific skeletal toxicity by significantly elevating gastrin levels and interfering with mineral absorption more aggressively than H2RAs.
  • The APAC Signal: The most striking finding was the regional disparity. While the global risk increase for major fractures was 23%, it skyrocketed to 77% in the Asia-Pacific subgroup. The authors hypothesize this is due to lower baseline dietary calcium in Asian diets and genetic variations that make PPIs more potent in these populations.
  • Fracture vs. Diagnosis: Interestingly, there was no significant difference in the rate of newly diagnosed osteoporosis between the two groups. This suggests that PPIs might compromise bone quality and structural integrity (leading to fractures from minor trauma) rather than just reducing bone mineral density, which is what standard scans measure.

Key Results

  • Major Osteoporotic Fractures (MOF): PPI users had a significantly higher risk (Hazard Ratio [HR] 1.23).
  • Specific Injuries: The risk was even higher for hip repair/arthroplasty (HR 1.46) and vertebral compression fractures (HR 1.16).
  • Asia-Pacific Results: PPI use in the APAC region was linked to a 77% increase in MOF risk (HR 1.77) and a 39% increase in vertebral fractures.
  • Validation: The negative control (melanoma) showed no association (HR 1.10, non-significant), validating the robustness of the propensity matching.

Limitations and Future Work

The authors acknowledge several limitations inherent to retrospective "big data" studies:

  1. Missing Lifestyle Factors: The database lacked granular data on smoking, alcohol consumption, physical activity, and over-the-counter (OTC) supplement use, all of which affect bone health.
  2. Prescription vs. Ingestion: The study tracked prescriptions, not whether patients actually took the pills as directed.
  3. Tertiary Care Bias: TriNetX data primarily comes from large hospital centers, which may not perfectly represent patients treated in smaller community clinics.

Future Work: The authors call for prospective randomized controlled trials to test "step-down" strategies—moving patients from PPIs to H2RAs once their initial symptoms are controlled—to see if this can mitigate fracture risks in aging populations.

Summary updated

Abstract

<title>Abstract</title> <bold>Background</bold> Long-term suppression of gastric acid is crucial for various conditions; however, proton-pump inhibitors (PPIs) may negatively impact skeletal health more than histamine-2 receptor antagonists (H2RAs). There is limited head-to-head data, particularly within Asian communities that have unique calcium consumption and drug metabolism patterns. <bold>Methods</bold> We conducted a retrospective cohort study that matched participants using propensity scores, utilizing the TriNetX integrated claims and electronic health-record database. Adults aged between 18 and 80 years who had at least 12 prescriptions for either a sole PPI or H2RA therapy from 2015 to 2025 were included. Following a 1:1 nearest-neighbor matching (caliper = 0.1), 131,698 individuals remained in each cohort. The outcomes measured included major osteoporotic fracture (MOF; primary), hip repair/arthroplasty, vertebral compression fracture, newly diagnosed osteoporosis, and malignant melanoma (negative control). The application of Cox proportional-hazards models allowed for the assessment of hazard ratios (HRs) and their linked 95% confidence intervals (CIs). <bold>Results</bold> Over a span of more than two million person-years, PPIs were linked to a higher incidence of MOF compared to H2RAs (HR 1.23, 95% CI 1.18–1.28; p &lt; 0.0001). Significant dangers were pointed out regarding hip repair/arthroplasty (HR 1.46, 1.35–1.58) and vertebral compression fractures (HR 1.16, 1.09–1.23), while newly diagnosed osteoporosis showed no notable changes (HR 1.01, 0.98–1.03). The negative-control outcome indicated no correlation. In the Asia-Pacific subgroup, the risk of MOF increased by 77% (HR 1.77, 1.56–2.01). Consistent findings emerged from subgroup and sensitivity analyses. The increased fracture risk remained despite comparable demographics, comorbidities, and consistent acid suppression, suggesting that there may be drug-class-specific skeletal toxicity—potentially due to impaired mineral absorption, elevated gastrin levels, and increased bone turnover. The more pronounced effect in East-Asian populations highlights the interaction between dietary and genetic influences. <bold>Conclusions</bold> Ongoing PPI treatment, in contrast to H2RA therapy, leads to significant increases in the risk of fragility fractures. These findings advocate for a reduction in PPI usage duration, favoring H2RAs for patients at greater fracture risk, and call for proactive measures to safeguard bone health while future studies explore causality and effective strategies for mitigation.
BibTeX
@inproceedings{jiang2025proton,
  author = {Danling Jiang and Xi Li and W. Susan Cheng and Bing Li and Guoqing Lv and Ying Li},
  title = {{Proton-Pump Inhibitors Versus H2-Receptor Antagonists on the Risk of Major Osteoporotic Fractures: A Global Propensity-Score-Matched Cohort Study}},
  booktitle = {Research Square},
  year = {2025},
  doi = {10.21203/rs.3.rs-7725160/v1},
}
Figure