Fracture Risk and Bone Mineral Density Reduction Associated with Proton Pump Inhibitors

Yuen Ting Lau, Pharm.D.; Nasiya N. Ahmed, M.D.

Pharmacotherapy. 2012;32(1):67-79.

Abstract and Introduction

Abstract

Many patients receive prolonged proton pump inhibitor (PPI) therapy for upper gastrointestinal disorders, but the long-term safety of PPIs, particularly increased risk of hip and nonhip fractures, has been questioned. To summarize the current literature on the risk of bone mineral density (BMD) reduction and fracture associated with PPI therapy, we conducted a literature search to identify all pertinent studies from 1980–February 2011. A total of 14 observational studies were included in this review. Most studies evaluated the risk of fracture associated with prolonged PPI exposure. Eight studies found an increased fracture risk at the hip, and five studies found an increased fracture risk at the spine associated with PPIs. Three studies showed reduction in fracture risk associated with PPIs after discontinuation for 1 month–1 year. Three studies evaluated the risk of BMD reduction associated with PPIs but did not find consistent changes in baseline or subsequent BMD. The current data suggest a modest increase in the risk of hip fracture and vertebral fracture associated with PPIs, although some studies showed conflicting results. Further studies will be needed to determine whether the increased risk of fracture is due to PPI exposure or residual confounding.

Introduction

Proton pump inhibitors (PPIs) irreversibly inhibit H⁺,K⁺-adenosine triphosphatase (ATPase) in the gastric parietal cells, resulting in potent suppression of gastric acid secretion; hence, PPIs are highly effective therapy for peptic ulcer disease, gastroesophageal reflux disease, pathologic hypersecretory conditions, and erosive esophagitis.^[1] Many patients receive long-term PPI therapy for these chronic conditions, leading to widespread use of PPIs; however, little is known about their long-term safety, as the duration of follow-up in clinical trials of PPIs has been limited to several months.

Recent observational data have raised concerns about adverse effects, such as Clostridium difficile-associated diarrhea, community-acquired pneumonia, and fractures, from prolonged PPI therapy.^[2] In May 2010, the United States Food and Drug Administration revised the labeling for PPIs to include information about the potential risk of hip, spinal, or radial fractures based on the results from seven epidemiologic studies.^[3] In this review, we summarize the current literature on the risk of bone mineral density (BMD) reduction and fracture associated with PPI therapy.

1 2 3 4 5 6 7 8

Next Section

Abstract and Introduction
Literature Search
Fracture Risk
Risk of Bone Mineral Density Reduction
Analysis of Study Data
Potential Mechanisms of Proton Pump Inhibitor-related Fracture
Future Research
Conclusion
References

Table 1. Observational Studies on the Risk of Fracture Associated with Proton Pump Inhibitors
Study Design	Population	Exclusion Criteria	PPI Exposure	Risk of Hip Fracture	Risk of Other Fractures	Dose-Response Relationship
Nested case–control⁵	Patients aged ≥ 50 yrs from the UK GPRD (13,556 cases, 135,386 controls)	Prior hip fracture	Cumulative PPI use >1 yr	AOR 1.44 (95% CI 1.30–1.59)	NA	Increasing risk of hip fracture with increasing cumulative duration of use and ADD: >1 yr: AOR 1.22 (95% CI 1.15–1.30) >2 yrs: AOR 1.41 (95% CI 1.28–1.56) >3 yrs: AOR 1.54 (95% CI 1.37–1.73) >4 yrs: AOR 1.59 (95% CI 1.39–1.80) ≤1.75 ADD: AOR 1.40 (95% CI 1.26–1.54) >1.75 ADD: AOR 2.65 (95% CI 1.80–3.90)
Nested case–control⁶	Patients aged 50–79 yrs from UK GPRD (1098 cases, 10,923 controls)	Major medical risk factors for hip fracture with OR > 2.0	Any PPI use	RR 0.9 (95% CI 0.7–1.1)	NA	No dose-response relationship with the cumulative number of PPI prescriptions
Nested case–control⁷	Adult members of Kaiser Permanente Northern California (33,752 cases, 130,471 controls)	Prior hip fracture	Cumulative PPI use ≥2 yrs	OR 1.30 (95% CI 1.21–1.39)	NA	Increasing risk of hip fracture with increasing ADD but not with cumulative duration of use: <0.75 pills/day: OR 1.12 (95% CI 0.94–1.33) 0.75–1.49 pills/day: OR 1.30 (95% CI 1.19–1.42) ≥1.5 pills/day: OR 1.41 (95% CI 1.21–1.64)
Case–control⁸	Fractures occurred in 2000 from the Danish national registry (124,655 cases, 373,962 controls)	None	Any PPI use; last use ≤1 yr ago	AOR 1.45 (95% CI 1.28–1.65)	Any fracture: AOR 1.18 (95% CI 1.12–1.43) Forearm: AOR 0.95 (95% CI 0.82–1.11) Spine: AOR 1.60 (95% CI 1.25–2.04)	No dose-response relationship, stratified by the cumulative dose or ADD, in any fracture, hip fracture, spine fracture, or forearm fracture
Case–control⁹	Patients aged≥ 50 yrs from a population database in Taiwan (1241 cases, 1241 controls)	Prior hip, vertebral, or wrist fracture	Any PPI use	≤28 DDD: AOR 0.93 (95% CI 0.64–1.36) 29–70 DDD: AOR 1.40 (95% CI 0.91–2.15) >70 DDD: AOR 2.11 (95% CI 1.45–3.07)^a	NA	Increasing risk of hip fracture with increasing cumulative dose; a similar dose-response trend was noted in patients with any PPI use in the past and those with any PPI use in ≤1 yr or ≤4 yrs preceding the index date
Case–control¹⁰	Patients aged ≥ 50 yrs from a population database in Manitoba, Canada (15,792 cases, 47,289 controls)	Receipt of osteoprotective drugs in the year before the fracture; residents of long-term care facilities	Continuous PPI use ≥1 yr before fracture	1–4 yrs use: NS ≥ 5 yrs use: AOR 1.62 (95% CI 1.02–2.59)	Osteoporotic-related fractures: 1–6 yrs use: NS ≥ 7 yrs use: AOR 1.92(95% CI 1.16–3.18)	Increasing risk of hip fracture with increasing duration of use after yr 5 ≥ 6 yrs use: AOR 2.49 (95% CI 1.33–4.67) ≥ 7 yrs use: AOR 4.55(95% CI 1.68–12.29)
Prospective cohort¹¹	Postmenopausal women in the WHI study (130,487 women)	Prior hip fracture	Current PPI use at baseline visit	HR 1.00 (95% CI 0.71–1.40)	Clinical spine: HR 1.47 (95% CI 1.18–1.82) Forearm/wrist: HR 1.26 (95% CI 1.05–1.51) Total fractures: HR 1.25 (95% CI 1.15–1.36)	No dose-response relationship with increasing duration of PPI use in hip, spine, forearm/wrist, and total fracture
Retrospective cohort¹²	Subjects aged ≥ 65 yrs from two prospective studies conducted in the U.S. (5339 women, 5755 men)	History of bilateral hip replacement	Current PPI use, included use ≤4 wks before baseline visit	Women: RH 1.16 (95% CI 0.80–1.67) Men: RH 0.62 (95% CI 0.26–1.44)	Nonspine fracture: Women: RH 1.34 (95% CI 1.10–1.64) Men: RH 1.21 (95% CI 0.91–1.62)	NA
Retrospective cohort¹³	Postmenopausal women from a prospective study conducted in Europe (1211 women)	Baseline use of bisphosphonates or raloxifene	Current omeprazole use, included use ≤1 yr before baseline visit	NA	Vertebral fracture: RR 3.10 (95% CI 1.14–8.44)	NA
Prospective cohort¹⁴	Postmenopausal women who presented to two Japanese hospitals for evaluation and treatment of osteoporosis and upper gastrointestinal disorders (75 women)	Diabetes mellitus, thyroid disorder, metabolic bone diseases, drugs or hormones that affect bone and gastrointestinal mucosa (e.g., bisphosphonates)	RRE patients required >6 mo of PPI therapy	NA	RRE patients had more episodes of vertebral fracture than patients without RE: 2.50 ± 2.60 vs 1.11 ± 1.68, p=0.009	NA
Retrospective cohort¹⁵	Patients aged ≥ 35 yrs who began new treatment with alendronate from a Danish national registry (38,088 patients)	Prior treatment with any antiosteoporotic drug; emigrated before end of study	Any concurrent PPI use during first 36 mo of alendronate therapy	Alendronate user/PPI nonuser: HR 0.61 (95% CI 0.52–0.71) Alendronate user/PPI user: HR 0.81 (95% CI 0.64–1.01) Interaction p<0.05	Interaction between PPI use and treatment response to alendronate for spine, humerus, or forearm fracture: NS	Greater attenuation of the treatment response to alendronate with increasing cumulative dose of PPIs: 1–359 DDD: HR 0.63 (95% CI 0.55–0.72) 360–719 DDD: HR 1.00 (95% CI 0.70–1.45) >720 DDD: HR 1.24 (95% CI 0.71–2.18)
Case–control¹⁶	Adult patients in Dutch PHARMO population database (6763 cases, 26,341 controls)	None	Current PPI use within 30 days before the fracture; recent, past, and distant use of PPI	Current use: AOR 1.20 (95% CI 1.04–1.40) Recent use: AOR 0.96 (95% CI 0.83–1.12) Past use: AOR 0.97 (95% CI 0.74–1.26) Distant past use: AOR 1.24 (95% CI 1.08–1.43)	NA	Current users with the highest ADD of PPI had the largest AOR, but no consistent trend with lower doses: <1.00 DDD: AOR 1.21 (95% CI 0.93–1.57) 1.00–1.75 DDD: AOR 1.12 (95% CI 0.88–1.42) >1.75 DDD: AOR 1.35 (95% CI 1.02–1.77); Increasing continuous duration of current PPI therapy did not increase fracture risk
Retrospective cohort¹⁷	Patients aged ≥ 40 yrs from UK GPRD who received a PPI (228,121 patients)	Paget's disease or cancer	Current PPI use compared with past PPI use as control group	Without concomitant bisphosphonates: ARR 1.22 (95% CI 1.10–1.37)	Any fracture: ARR 1.15 (95% CI 1.10–1.20) Osteoporotic: ARR 1.16 (95% CI 1.09–1.24) Radius/ulna: ARR 1.09 (95% CI 0.98–1.20) Vertebral: ARR 1.40 (95% CI 1.11–1.78)	Increasing risk of osteoporotic and hip fractures with increasing DDD but not with increasing duration of current PPI use: Hip fracture: <1.0 DDD: ARR 1.13 (95% CI 0.98–1.31) Hip fracture: 1.0–1.75 DDD: ARR 1.27 (95% CI 1.12–1.45) Hip fracture: >1.75 DDD: ARR 1.45 (95% CI 1.06–1.99)

PPI = proton pump inhibitor; UK GPRD = United Kingdom General Practice Research Database; AOR = adjusted odds ratio; CI = confidence interval; NA = not available; ADD = average daily dose; OR = odds ratio; RR = relative risk; DDD = defined daily dose; NS = not statistically significant; WHI = Women's Health Initiative; HR = hazard ratio; RH = relative hazard; RRE = refractory reflux esophagitis; ARR = adjusted relative risk.
^a Adjusted for matching variable and comorbidities.

Table 2. Observational Studies on the Effect of Proton Pump Inhibitors on Bone Mineral Density
Study Design	Population	Exclusion Criteria	PPI Exposure	BMD at Hip	BMD at Spine	BMD at Radius	BMD of Total Body
Prospective cohort¹¹	Postmenopausal women in WHI (10,833 women from three study centers)	Prior hip fracture	Current PPI use at baseline visit	No difference in baseline BMD; 3-yr BMD change: 0.62 ± 0.45% vs 1.36 ± 0.25%, p=0.05^a	No difference in baseline BMD; 3-yr BMD change: NS	NA	No difference in baseline BMD; 3-yr BMD change: NS
Retrospective cohort¹²	Women and men ≥65 yrs from two prospective studies conducted in U.S. (5339 women, 5755 men)	History of bilateral hip replacement	Current PPI use including use within 4 wks before baseline visit	Baseline BMD in women: NS; baseline BMD in men: 0.946 g/cm² vs 0.958 g/cm², p=0.05 No difference in annualized BMD change in men or women	NA	NA	NA
Cross-sectional and retrospective cohort¹⁸	Patients who had BMD measurements recorded in a population database in Manitoba, Canada from 2000–2007 (2193 hip cases, 5527 hip controls; 3956 spine cases, 10,257 spine controls; 2549 in cohort)	<5 yrs of complete drug history; concurrent osteoprotective drugs in case–control analysis	Case–control: cumulative PPI use during the 5 yrs before BMD measurement compared with no PPI use; Cohort: ADD of PPI > 0.5 dose/day between 2 BMD measurements vs <0.5 dose/day	Osteoporosis: NS annualized change in BMD: NS	Osteoporosis: NS annualized change in BMD: NS	NA	NA

PPI = proton pump inhibitor; BMD = bone mineral density; WHI = Women's Health Initiative; NS = not statistically significant; NA = not available; ADD = average daily dose.
^a 3-year change in BMD data were reported as mean ± standard error in PPI users versus nonusers.
^bBaseline BMD data were reported as mean in PPI users versus nonusers.

processing....

Fracture Risk and Bone Mineral Density Reduction Associated with Proton Pump Inhibitors

Abstract and Introduction

Abstract

Introduction

Tables

Tables

References

Authors and Disclosures

Authors and Disclosures

Email This

Feedback