Proton Pump Inhibitors and Bone Fractures: What's the Risk?

David A. Johnson, MD


September 06, 2011

This feature requires the newest version of Flash. You can download it here.

Hello. I'm Dr. David Johnson, Professor of Medicine and Chief of Gastroenterology at Eastern Virginia Medical School in Norfolk, Virginia. Proton pump inhibitors (PPI) in osteoporosis and related fractures -- is it real or not real? The US Food and Drug Administration (FDA) felt that it was real enough to issue product labeling with an advisory to patients who were on long-term use of medications, defined as over 1 year or on high-dose medications. Patients are to be made aware that there is potential risk for osteoporosis and fractures associated with the concomitant use of PPIs.

The FDA removed that risk rating for short-term use, and it was particularly intended to focus on over-the-counter medications. How do you counsel your patients? Is there evidence to suggest that there may be some causal harm?

A very nice systematic review and meta-analysis was performed by the physicians at Texas Tech in the most recent issue of the American Journal of Gastroenterology.[1] In this evaluation, the authors looked at over 1600 studies in which identification of osteoporosis and potential risks for osteoporosis were evaluated. Using very strict evidence criteria for analysis, they identified only 32 articles from that cohort of 1600, which were extracted in a vigorous way for eligibility for inclusion in this systematic review and meta-analysis. From there, they garnered 10 articles that were included in the present analysis.

The primary focus was on osteoporosis in hip fractures, but they also looked at vertebral fractures and radial fractures and they coadjusted for patients who used PPIs. There were 9 studies for hip fracture in the final analysis, and the overall odds ratio for risk was 1.25. For vertebral fractures, there were 4 studies and the overall risk was 1.5. For radial fractures, there was an odds ratio of 1.1, which was not statistically significant and was dismissed in the final analysis.

What was interesting when they started to look at the heterogeneity between these studies is that they were all across the board and impossible to look at in particular retrospective studies. Six studies were cohort controlled and 4 had retrospective data analysis. These were patient populations that they could not coadjust well for confounding variables.

Nonetheless, because the heterogeneity was there and it was obvious that the studies couldn't be put together, they were put into a random-effects model which allowed the authors to generate the numbers that I gave you earlier. Hip fracture was 1.25. Cervical spine fractures and vertebral fractures were 1.5, and for wrist fractures it was nonstatistical.

When they looked at the use of PPIs and coadjusted it for short-term use or long-term use, statistical significance was found for short-term use only. It doesn't make sense, does it? The longer you are on the medicine, the safer it gets. I don't think we can go there either. Nonetheless, the short-term risk was the only one that came out statistically significant. It was nominal, as far as the potential odds ratio risk, at 1.25 for hip fracture and 1.5 for vertebral fractures.

It doesn't make sense that the more you're exposed to a medication, for longer periods of time, your risk would actually go down. I think that would be an overextension. I would just say that no significant risk was identified for long-term exposure.

If you go back and review what the agencies that have put out warnings on this have said, it's just the opposite. They said that risk was associated with long-term exposure and not short-term exposure. Again, it's very paradoxical relative to the present concerns that have been voiced by the FDA.

It still doesn't make sense that this is a significant risk. What I think we are looking at is channeling bias. In fact, when you go back and look at confounding variables such as bisphosphonate use, one study very nicely identified that the PPI risk exposure was significantly increased after patients were started on bisphosphonates. They had symptoms, they had problems, and they had reasons to start a PPI, but they had already identified risk for osteoporosis.

So where does this leave us? I think this systematic review puts into light that there are just way too many confounding variables to put too much weight on this allegation that bone density loss is evident by PPI exposures. In fact, as noted in this systematic review and meta-analysis, the studies that have looked at longitudinal assessment of bone loss have identified none with patients when coadjusted for PPI exposure. It doesn't make sense that patients would develop osteoporosis from calcium malabsorption. The data on that are very lacking.

How could this possibly have a thread of allegation that there may be causality? This particular study pointed out that B12 absorption may be diminished. Again, theoretic studies have not shown that in prospective evaluation and that B12 may be associated with homocystinuria. Homocystine is associated with cross-linking in collagen. There may be some remote potential that it might be linked with bone fractures. But again, we're even lacking in the premise that B12 malabsorption is evident with long-term PPI use.

I really cannot put science behind this. If you come back to the overall risks that they generated from these odds ratios, the number needed to harm for hip fractures, which is the present concern, was 1 in 2600. For vertebral fractures, it was about 1 in 370. Again, very large numbers needed to harm. Nobody has looked at what would happen if the patients were not on these medications.

I leave you with the following: What do I tell my patients? I would say that there is no news as far as considerable risk. It puts more emphasis on the confounding variables that we can now rectify as reasons for some of these odds ratios that have been identified. Recommend to patients that if they don't need their medications, they should stop them. If they need them, they should stay on them.

Should they take supplemental calcium? I can't tell you from this particular study. It's not in my practice. But the systematic review and meta-analysis gave me reaffirmation that the data are extremely conflicting. Make your patients aware of this when they come to see you. I don't proactively reach out to every patient and tell them that if they're on a PPI then they need to be aware of this news. However, I think it is a healthy conversation to have so that your patients are aware of it. Make sure they don't have any questions.

The problem is patients who unilaterally stop taking medications such as PPIs based on what they have read on the Internet or heard on a news report. I think that is a big problem.

Hopefully this study will put you in better stead for your education to patients. I'm more reassured that this is really not an issue and that the concern is very much non-evidence-based. I will leave that to your application and your own interpretation to look at this review in the American Journal of Gastroenterology. I look forward to chatting with you again soon. I'm Dr. David Johnson. Thanks for listening.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: