The HUNT Study: Proton Pump Inhibitors and Fracture Risk

The HUNT Study: Proton Pump Inhibitors and Fracture Risk
Kaylin Maddy, PharmD, Park Nicollet Health Services

Background: Proton pump inhibitors (PPIs) are broadly prescribed for gastrointestinal conditions. These medications have been correlated with an increased risk of fractures in multiple previous studies, however a definite causative link is not well understood. The FDA requires PPI packaging information to include a warning about increased hip, wrist, or spine fracture risk with long term and high dose PPI use. Many mechanisms for this risk have been proposed including reduced calcium and magnesium absorption, though these have not been adequately demonstrated in humans.

Purpose: The Nord-Trøndelag Health Study (HUNT) study aimed to examine a possible association between the use of PPIs and risk of fracture in a large group of Norwegian individuals.

Study Design: A total of 15,017 women and 13,241 men aged 50 to 85, with a mean age of 65 years old were analyzed. Data was retrospectively reviewed from a large Norwegian Health Study (HUNT3), the Norwegian Fracture Registry, and the Norwegian Prescription Database. The fracture registry collects information on hip and forearm fractures. History of exposure to PPIs, anti-osteoporotic drugs (AODs), and oral glucocorticoids (GCs) were pulled from the prescription database. Individuals were followed from their date of participation in the health study in 2006-2008 until the date of their first fracture, death, or study completion in 2012. PPI exposure was defined as a minimum of 90 days and on a daily dose equivalent to or greater than omeprazole 20 mg, lansoprazole 30 mg, esomeprazole 30 mg, or pantoprazole 40 mg. Four different COX proportional hazard models were used for analysis and included variations of time dependent exposure to PPIs as well as age, FRAX score, milk intake, and exposure to AODs and GCs.

Results: In total, 17.9% of women and 15.5% of men had PPI exposure, with 11.7% of women and 10% of men exposed for greater than one year. The mean duration of PPI therapy was 3.8 years, with a range from 6 months to 7 years. Over a median of 5.2 years, 266 women and 134 men had a hip fracture; 662 women and 127 men had a forearm fracture. Overall, after adjusting for age, they found no increased risk of fracture with PPI exposure (Hazard Ratio - Women: 0.82 [95% CI 0.67-1.01], Men: 1.05 [95% CI 0.72-1.52]). After adjusting for use of AODs and FRAX score, there was an overall decline in hazard ratio but still no increased risk identified (Women: 0.80 [95% CI 0.65-0.98], Men: 1.00 [95% CI 0.69-1.45]). The lack of association between PPIs and fractures was true across all of the hazard model groups. There was also no difference found between high and low dose PPI users, which looked at doses above and below the defined dose equivalencies of the different PPIs. It was noted that PPI users were older, had a higher BMI, higher FRAX score, more previous fractures, and were on more GCs and AODs. For women, use of additional medication use was higher in the PPI group than the no PPI group (AOD 14.5% vs. 8.5%, GC 27.2% vs. 10.6%).

Conclusions: The use of PPIs over an average of 3.8 years of exposure was not associated with increased risk of fractures in a large retrospective, population based study of  individuals aged 50-85 years old in Norway. Interestingly, PPI users seemed to be higher risk but were also more protected from fracture due to more prevalent AOD treatment. In the adjusted model, the hazard ratio was actually significant for reduced fracture risk in women, though the authors did not identify reduced risk as a finding since they were looking to prove increased risk. It should be noted that the fracture registry only had the ability to collect information on hip and forearm fractures. Additionally, many previous studies linked rabeprazole to fracture risk, however rabeprazole was not included in this study as it is not approved in Norway. The findings of this study interestingly contradict many previous studies about fracture risk, however the retrospective design and lack of vertebral fracture inclusion limit its application to practice.

Key Point: The link between PPIs and increased fracture risk may be less correlated than previously thought. However, pharmacists should continue assessing the need for PPIs and de-escalating therapy as appropriate.


  1. Hoff M, Skovlund E, Skurtveit S et al. Proton pump inhibitors and fracture risk. The HUNT study, Norway. Osteoporos Int. 2019;31(1):109-118. doi:10.1007/s00198-019-05206-0