Osteoporotic proximal humerus fractures constitute about 10 percent of all geriatric fractures, noted Anshu Singh, MD, of Kaiser Permanente San Diego, during the American Shoulder and Elbow Surgeons (ASES) 2013 Specialty Day session on shoulder trauma and arthroplasty. But the impact of osteoporosis treatment on these fractures has not been measured, nor have such fractures been taken into account when assessing the value of osteoporosis screening and treatment.
Dr. Singh and colleagues identified 524,612 Kaiser Permanente enrollees who were at least 60 years old as of Jan. 1, 2002, and who were at risk for sustaining incident proximal humerus fracture. Of these, 34 percent were screened for osteoporosis and 30 percent received some sort of treatment.
Patients were considered to have been screened for osteoporosis if a bone mineral density (BMD) test had been conducted during the study period. Regardless of the BMD test results, a prescription for disphosphonates, calcitonin, or estrogens, selective estrogen receptor modulators, miscellaneous hormones, and sex hormone combinations was considered pharmacologic intervention for osteoporosis.
Overall, 1 percent of the patients (5,111) sustained a new proximal humerus fracture during the study period (2002–2008). Researchers found that proximal humerus fractures were more common among women (1.4 percent vs. 0.5 percent for men, P < 0.001) and among white patients (1.4 percent vs. 0.6 percent for nonwhite patients, P < 0.001).
Proximal humerus fracture occurrence was most common in the oldest patients and was highest among those 80 years or older. Patients not screened for osteoporosis had proximal humerus fractures in greater proportions than did patients who had been screened by DXA scan (1.1 percent vs. 0.8 percent, P < 0.001). Proximal humerus fractures were also more common among patients diagnosed with osteoporosis, compared with those without the diagnosis (2.6 percent vs. 0.7 percent, P < 0.001).
Does screening help?
Patients in the screened and treated groups were more likely to be female, white, diagnosed with osteoporosis, and to have sustained previous hip or distal radius fractures, compared with the unscreened and untreated groups. But when researchers considered screening and treatment jointly, they found the following differences among the four groups:
- The group that was both screened and treated was more likely to be female, white, diagnosed with osteoporosis, and have a history of prior distal radius fractures, compared with the other combinations of screening and treatment.
- Those who were treated but not screened were more likely than all other groups to have had a prior hip fracture (5.8 percent).
Over the course of the study period, 1.1 percent of patients not treated with pharmacologic interventions for osteoporosis sustained incident proximal humerus fractures, compared with 1.5 percent of the treated group who sustained a fracture (P < 0.001). However, the treated group appeared to sustain fractures less frequently than the untreated group until about 3.25 years into the study, at which time the treated group sustained more fractures.
Finally, Dr. Singh said, looking at screening/treatment combinations, approximately 0.7 percent of the patients who were screened but not treated sustained proximal humerus fractures. Those who were screened and treated had similar experience early in the study period, but by approximately 3.5 years of follow-up, this group began to sustain more proximal humerus fractures than did the screened/untreated group (1.0 percent and 0.7 percent, respectively). The unscreened but treated group sustained incident proximal humerus fractures most frequently throughout the study period, with 2.9 percent of that group ultimately having a fracture.
Adjusting for age, sex, race, history of diabetes, history of prior hip or distal radius fracture, and diagnosis of osteoporosis, “we found that having been screened for osteoporosis was strongly associated with a reduction in the instantaneous risk of incident proximal humerus fracture, relative to the unscreened group,” Dr. Singh said.
“Receipt of pharmacologic intervention for osteoporosis was also strongly associated with a reduction in the risk of proximal humerus fracture, relative to patients who had not received medication treatment.” Having a prior history of hip fracture was not associated with risk of sustaining an incident proximal humerus fracture, he added.
“The novel finding in this study is that screening an individual for osteoporosis decreased the hazard ratio for fracture to 0.17 versus unscreened controls. Medical treatment decreased the hazard ratio to 0.55 versus patients who were not medically treated,” Dr. Singh said. “This is the only study that we know that shows that treatment can actually decrease the risk of a shoulder fracture. This may decrease the number-needed-to-treat and have a significant impact on the cost savings of osteoporosis management programs, because the shoulder has not factored into the calculus of such treatment to date.”
Co-authors with Dr. Singh are Annette Adams, PhD, MPH; Raoul Burchette MS; Richard M. Dell, MD; Tadashi Funahashi, MD; and Ronald A. Navarro, MD.
Disclosure information: Mr. Burchette—JALEVA. Dr. Navarro—Accumed; Arthrex; AAOS; American Board of Orthopaedic Surgery. Dr. Adams—no data available. Drs. Singh, Dell, and Funahashi—no conflicts.
Terry Stanton is senior science writer for AAOS Now. He can be reached at firstname.lastname@example.org
- Osteoporotic proximal humerus fractures constitute about 10 percent of all geriatric fractures.
- Only about a third of patients at risk for sustaining incident proximal humerus fracture are screened or treated for osteoporosis.
- In this study population, the treated group appeared to sustain fractures less frequently than the untreated group until about 3.25 years into the study, at which time the treated group sustained more fractures.
- However, screening an individual for osteoporosis decreased the hazard ratio for fracture to 0.17 versus unscreened controls; treatment decreased the hazard ratio to 0.55 versus patients who were not treated.