Major Finding: Among patients taking proton pump inhibitors for at least 2 years, the risk of having a hip fracture was increased by 30%, but only when at least one other fracture risk factor (smoking, dementia, arthritis, visual impairment) was present. Among PPI users with none of these other risk factors, the odds ratio for fracture was 0.66.
Data Source: A nested case-control study using data from an integrated health services organization, including 33,752 cases and 130,471 controls.
Disclosures: Dr. Corley disclosed receiving research funding, unrelated to this study, from Wyeth Pharmaceuticals, which makes a proton pump inhibitor. The remaining authors disclose no conflicts of interest. The study was funded by Kaiser Permanente and a grant from the U.S. National Institutes of Health.
Hip fracture patients were 30% more likely to have a long-term history of proton pump inhibitor use, compared with controls, Dr. Douglas A. Corley and his colleagues reported.
Moreover, the association was found to be stronger with higher doses of PPIs, and the link diminished after PPI discontinuation, the researchers wrote.
“These findings do not recommend against acid suppression for persons with clear indications for treatment,” wrote Dr. Corley, a researcher with Kaiser Permanente Northern California (Gastroenterology July [doi: 10.1053/j.gastro.2010.03.055
Dr. Corley and his colleagues looked at 33,752 adult members of the Kaiser Permanente Northern California integrated health care delivery system who had an incident diagnosis of a hip fracture between January 1995 and September 2007. To be included in the study, patients had to have been in the Kaiser Permanente Northern California system for at least 2 years prior to their fracture. Patients who had a previous hip or femur fracture diagnosis were excluded.
Patients were roughly matched in a 4:1 ratio with demographically comparable controls, also from the Kaiser Permanente Northern California system. Controls had no history of hip fracture, and had also been in the Kaiser system for at least 2 years.
Patients were predominantly women (65.7%), 70 years of age or older (69.4%), and white (79.6%), according to the authors. Roughly 40% had received a prescription for a proton pump inhibitor while in the Kaiser Permanente Northern California system.
According to Dr. Corley, patients whose records indicated “long-term” use of PPIs (defined by the authors as greater than 2 years) had an odds ratio of having a fracture within the study period of 1.30, compared with nonusers (95% confidence interval, 1.21-1.39).
However, all of the increased risk for fracture was present only in patients who had at least one other risk factor for fracture, such as smoking, dementia, arthritis, or visual impairment. Indeed, among patients with none of these risk factors, the odds ratio for fracture among PPI users was 0.66 (95% CI, 0.38-1.12).
The researchers also found a trend toward increased fracture risk among subjects taking higher daily doses of PPIs. For example, among patients taking an average of 0.01-0.74 pills/day, for a duration between 2 and 3.9 years, the OR for a fracture, compared with nonusers, was 1.23 (95% CI, 1.08-1.39); among users taking 0.75-1.49 pills/day, for the same duration of time, the OR was 1.43 (95% CI, 1.28-1.60), and for more than 1.49 pills/day the OR was 1.41 (95% CI, 1.21-1.64).
Despite the association with PPI dosage, there was no link between duration of PPI use and fracture risk.
The researchers also found that “the strength of the association between PPI use and hip fracture was greatest among current users and diminished after discontinuation of PPI use.” For example, while the OR for current users was 1.30 (95% CI, 1.21-1.41), it was 1.24 for patients whose most recent prescription was 1.0-1.9 years before the index date (95% CI, 0.90-1.72), and dropped to 1.09 for patients whose last PPI prescription was 2.0-2.9 years before the index date (95% CI, 0.64-1.85).
Dr. Corley proposed several mechanisms by which acid inhibition could influence fracture risk. For one, he said, acid inhibition could directly influence calcium absorption: he pointed to a small, randomized trial in which omeprazole decreased the absorption of radio-labeled calcium pills by 61%, compared with placebo (Am. J. Med. 2005;118:778-81).
“Second,” he wrote, “acid inhibition may induce hyperparathyroidism, which directly decreases bone mineral density, through hypergastrinemia, although this is controversial.”
Finally, he suggested that fracture risk may be mediated by interference by PPIs with bone remodeling. However, he added, “none of these mechanisms are proven.”
Fracture risk also was increased in patients using histamine2-receptor antagonists, another class of drugs that inhibit acid secretion (OR, 1.18).
The authors concluded that acid inhibition might raise fracture risk in persons already at risk for osteoporosis, although confounding cannot be excluded.