This article is more than 5 years old. Information may no longer be current.
Secondary osteoporosis, rheumatoid arthritis improve FRAX accuracy in multiple sclerosis
Click Here to Manage Email Alerts
The Fracture Risk Assessment Tool, or FRAX, underestimated the observed 10-year risk for major osteoporotic fracture by up to 5% among adults with multiple sclerosis; however, calibration improved when including secondary osteoporosis or rheumatoid arthritis as proxies for additional risk, according to findings published in the Journal of Bone and Mineral Research.
“Our study extends the range of clinical conditions where FRAX has been shown to be of value for assessing fracture risk, but it may underestimate that risk in people with multiple sclerosis (MS),” William D. Leslie, MD, MSc, professor of medicine and radiology at the University of Manitoba, Canada, told Endocrine Today. “When using secondary osteoporosis or rheumatoid arthritis as proxies for the additional MS-associated risk, FRAX accurately predicts fracture outcomes in those with MS, providing clinicians with a readily available approach to improve the accuracy of fracture prediction in MS.”
Leslie and colleagues analyzed data from 744 adults with MS who underwent a bone mineral density screening (92.5% women; mean age, 57 years), each matched with five controls by age, sex and first BMD screening date (n = 3,915), using health claims databases linked to the Manitoba Bone Mineral Density Database. Researchers calculated 10-year probability of major fracture via FRAX at the BMD screening date and assessed incident major osteoporotic fractures occurring between BMD screening date and March 31, 2015 (mean follow-up time, 8 years). Researchers used Cox proportional hazard models to assess the effect of MS on the risk for major osteoporotic fracture, adjusting for 10-year FRAX probabilities.
Researchers found that MS was a statistically significant risk factor for major osteoporotic fracture after controlling for 10-year FRAX probability without BMD (HR = 1.67; 95% CI, 1.29-2.16), with risk falling slightly after controlling for 10-year FRAX probability with BMD (HR = 1.48; 95% CI, 1.14-1.91). After controlling for FRAX individual risk factors, MS was associated with incident major osteoporotic fracture (HR = 1.45; 95% CI, 1.12-1.89).
Researchers observed and predicted 10-year probability for major osteoporotic fracture for MS cases and controls, and found that, for controls, 10-year probability with and without BMD did not differ from the observed 10-year risk, indicating an accurate estimation of fracture risk (P > .05).
In adults with MS, however, FRAX underestimated absolute risk for major osteoporotic fracture by 3.1% for 10-year probability with BMD and by 4.4% for 10-year probability without BMD (P < .01 for both). Calibration improved when secondary osteoporosis was used to calculate FRAX without BMD, according to researchers, and was best when rheumatoid arthritis was used to calculate FRAX probability with BMD.
“When we used rheumatoid arthritis as a proxy for MS in FRAX, the 10-year probability of [major osteoporotic fracture] with and without BMD was no longer differed from the observed 10-year [major osteoporotic fracture] risk in MS cases,” the researchers wrote.
The researchers noted that adults with MS continued to have an increased risk for major osteoporotic fracture vs. controls when secondary osteoporosis was used to calculate FRAX without BMD (HR = 1.33; 95% CI, 1.02-1.72). When rheumatoid arthritis was used to calculate FRAX with BMD, MS was no longer a statistically significant risk factor (HR = 1.19; 95% CI, 0.911-1.55), according to researchers.
“Our study provides clinicians with a readily available approach to improve the accuracy of fracture prediction in MS,” Leslie said. “FRAX can then be used to identify people with MS who are at high fracture risk in order to reduce that risk through targeted intervention. Our findings need to be confirmed in other diverse populations, with sufficient numbers to examine hip fracture prediction since this fracture has the greatest adverse impact on health outcomes.” – by Regina Schaffer
Disclosures: The authors report no relevant financial disclosures.
Perspective
Back to Top
FRAX, a computer‐based algorithm primarily intended for use in primary care, calculates the 10-year probability for major osteoporotic fracture and hip fracture from easily obtained clinical risk factors. When developed, consideration was given to clinical conditions that might impact on fracture risk. A key question was whether this risk was independent of bone mineral density; and in the absence of sufficient evidence at the time, it was decided to take a conservative approach whereby secondary causes of osteoporosis were assumed to mediate their impact on fracture risk by their effect to decrease bone density. Thus, once BMD was entered to the FRAX calculation, these secondary causes carried no further risk for fracture. Rheumatoid arthritis was the exception to this assumption, as data demonstrated that its effect on fracture risk was independent of BMD in the FRAX algorithm.
Since the launch of FRAX in 2008, several studies have suggested that other diseases, most notably type 2 diabetes, may also impact on fracture risk independently of FRAX risk factors, including BMD. The study by Bisson and colleagues, utilizing the Manitoba Cohort, suggests that multiple sclerosis may also impact independently on fracture risk. This study should be regarded as an important first step in the possible incorporation of multiple sclerosis within the basket of diseases captured under the secondary causes of osteoporosis within FRAX, or even, as suggested by the authors, captured using the rheumatoid arthritis variable.
Why only a first step? The individual risk variables in FRAX underwent a thorough evaluation that included definitive demonstration of their independence from, and interaction with, the other FRAX risk factors on an international basis. There is, therefore, a need to supplement the observations reported by Bisson and colleagues in other regions and populations. Additionally, it is necessary to establish any interactions between multiple sclerosis and other variables not only in the prediction of fracture, but also in the context of FRAX, of mortality, as the probability generated by FRAX integrates the risk for fracture and the risk for death. It is to be hoped that the study by Bisson and colleagues will stimulate such research within available population cohorts.
Click Here to Manage Email Alerts