Risk tools do not outperform BMD for predicting hip fractures among late-life adults
Click Here to Manage Email Alerts
Key takeaways:
- Femoral neck BMD was better at predicting hip fractures for women aged 80 years and older than two fracture risk assessment tools.
- Both the Garvan Fracture Risk Calculator and BMD outperformed FRAX among men.
Two fracture risk assessment tools incorporating femoral neck bone mineral density did not outperform femoral neck BMD alone for predicting hip fractures among adults aged 80 years and older, according to study data.
In an analysis of data from three prospective cohort studies, the use of femoral neck BMD was more accurate than the Fracture Risk Assessment Tool (FRAX) and the Garvan Fracture Risk Calculator for predicting 5-year hip fracture probabilities among women aged 80 years and older, whereas both the Garvan calculator and femoral neck BMD outperformed FRAX for predicting 5-year hip fracture probabilities among men aged 80 years and older.
“Until better fracture risk assessment tools are available, clinicians should prioritize consideration of hip BMD, life expectancy and patient preferences in decision-making regarding drug treatment for hip fracture prevention in late-life adults,” Kristine E. Ensrud, MD, MPH, a professor in the division of epidemiology and community health, department of medicine at University of Minnesota, told Healio.
Ensrud and colleagues obtained data from 8,890 adults aged 80 years and older who participated in one of three prospective cohort studies where femoral neck BMD was measured and a clinical risk factor assessment was conducted from 1997 to 2016. Predicted 5-year hip fracture probabilities were calculated using FRAX and the Garvan Fracture Risk Calculator with femoral neck BMD included in the models. Risk was also predicted using femoral neck BMD alone. Self-reported hip fractures were collected every 4 or 6 months after baseline and later confirmed by radiographic reports. Follow-up continued for 5 years after baseline or until the occurrence of hip fracture or death.
The findings were published in JAMA Network Open.
BMD best for predicting fractures
There were 4,906 women and 3,984 men included in the study. Mean estimated 5-year hip fracture probabilities ranged from 4.4% to 16.8% for women and 2% to 5.6% for men.
During follow-up, 6.5% of women and 3.1% of men sustained a hip fracture, and 16.7% of women and 23.1% of men died before a hip fracture. Area under the curve for predicting hip fractures among women was highest using BMD alone (AUC = 0.72) compared with FRAX (AUC = 0.69; P = .01) and the Garvan calculator (AUC = 0.69; P = .01). For men, both the Garvan calculator (AUC = 0.77) and BMD alone (AUC = 0.76) were better at predicting hip fractures than FRAX (AUC = 0.71; P < .001 for both).
When individual components of FRAX were assessed, women with a prior fracture had a higher risk for a hip fracture than those who did not have a prior fracture (HR = 1.35; 95% CI, 1.06-1.72) and the risk for a hip fracture increased with each 1 standard deviation (SD) decrease in femoral neck BMD T-score (HR = 2.08; 95% CI, 1.78-2.43). Among men, each 5-year increase in age (HR = 1.61; 95% CI, 1.2-2.15), the presence of rheumatoid arthritis at baseline (HR = 2.16; 95% CI, 1.15-4.05) and each 1 SD decrease in femoral neck BMD T-score (HR = 2.62; 95% CI, 2.1-3.26) were associated with a higher risk for a hip fracture.
In an assessment of Garvan Fracture Risk Calculator components, women with two prior fractures (HR = 1.44; 95% CI, 1.07-1.95) or three or more prior fractures (HR = 1.48; 95% CI, 1.06-2.06) had a higher risk for a hip fracture than those with no prior fracture. For men, the risk for a hip fracture increased with each 5-year increase in age (HR = 1.55; 95% CI, 1.16-2.08), with three or more prior fractures compared with no previous fractures (HR = 2.3; 95% CI, 1.15-4.59) and for men with two falls in the past year compared with no falls (HR = 2.21; 95% CI, 1.41-3.46). Both women (HR = 2.21; 95% CI, 1.92-2.53) and men (HR = 2.63; 95% CI, 2.15-3.22) had a higher risk for a hip fracture with each 1 SD decrease in femoral neck BMD T-score.
Few adults receive BMD testing
Ensrud said the findings are important because many practices use the risk assessment tools to help make treatment decisions to prevent fractures among older adults.
“Our results are highly relevant to the primary care practice setting with its major time constraints, numerous demands, scarce resources and great need to prioritize risk factor assessment,” Ensrud said. “Among U.S. women aged 80 years and older, 87% had no evidence of BMD testing in the last 2 years. The rate of BMD testing among late-life men is even lower.”
Ensrud said more studies are needed on fracture prevention for adults aged 80 years and older, including research assessing osteoporosis drug therapy.
“Randomized trials of osteoporosis drug treatment have not included sufficient numbers of this age group to evaluate efficacy in fracture prevention, especially in late-life adults with multiple medical conditions who are typically excluded from trial participation,” Ensrud said.
For more information:
Kristine E. Ensrud, MD, MPH, can be reached at ensru001@umn.edu.
Collapse
Click Here to Manage Email Alerts