A new assessment index based on six easily accessible variables effectively predicts the risk of nonvertebral fractures in postmenopausal osteoporotic women, a study has shown.
Because of its efficacy and clinical convenience, the new tool can be used to help in the identification and management of high-risk patients, reported lead investigator Dr. Christian Roux of Université de Paris Descartes, and colleagues.
The effect of a given osteoporosis treatment on nonvertebral fracture risk is a potential determinant of treatment choice. As such, the assessment of nonvertebral fracture risk in postmenopausal women with osteoporosis has important management implications, the authors wrote. Because risk assessment in this population can be difficult, in that multiple variables such as type and mechanism of fracture can impact an individual's risk profile, the investigators sought to identify significant predictive factors from which to develop a clinical useful fracture-risk-assessment tool.
Toward this end, they analyzed 3 years of follow-up data from 2,546 postmenopausal osteoporotic women who had been in the placebo groups of three multicenter, randomized controlled trials of the biphosphonate risendronate.
At baseline, the mean age of the study participants was 72 years, the mean femoral T score was −2.5, and 60% and 53% had prevalent vertebral and nonvertebral fractures, respectively.
All of the participants received 1,000 mg of calcium daily and up to 500 IU of vitamin D daily if baseline serum hydroxyvitamin D was less than 16 ng/mL (Ann. Rheum. Dis. 2007;66:931–35).
During the 3-year follow up, 222 nonvertebral fractures were observed in 206 patients. Of 14 variables included in logistic regression analysis of the fracture group, 6 emerged as independent predictors of nonvertebral fracture risk: age, height, prior nonvertebral fracture, number of prevalent vertebral fractures, femoral neck T score, and serum 25-hydroxyvitamin D, the authors reported.
With respect to vitamin D, “In the 48 patients who had a baseline serum hydroxyvitamin D less than 16 ng/mL, the incidence of nonvertebral fractures was 14.58%,” compared with 8.25% among the patients whose baseline measure was 16 ng/mL or higher, the authors wrote.
“Although patients received supplements of calcium and vitamin D, the low baseline level of 25-hydroxyvitamin D was still a significant risk factor for fracture, which may be related to a long-term effect of sarcopenia due to vitamin D deficiency,” they stated.
Based on the odds ratios for fracture associated with each of these variables, the investigators calculated fracture-risk index values for intervals within each variable, adjusting the minimal possible index contribution for each variable to zero (see index value chart).
For example, the risk index value for patients younger than 65 years is 0, while the index value for patients 80 years or older is 0.8.
The area under the receiver operating characteristic (ROC) curve for the final model based on the six predictors was 0.66. To minimize the absolute difference between sensitivity and specificity in the ROC curve, the investigators selected a cut-off value of 0.086 (or 8.6% nonvertebral fracture risk at baseline) for the predicted probability of nonvertebral fracture.
According to the model, patients with an index value of 2.1 or higher belong to a subgroup at high risk for nonvertebral fracture, the authors wrote.
In the 998 women from the original study population with an index value of at least 2.1, “the incidence of nonvertebral fractures was 13.2%, 1.5 times higher than the average of the population,” they noted.
One of the study's primary limitations is the lack of information on falls, which was not available.
This “ignored parameter” can be of crucial importance in explaining the fractures and may be useful in explaining the low value of the ROC curve, according to the researchers.
Still, the study findings suggest that “among osteoporotic women, a proportion of patients with a high risk of nonvertebral fractures can be selected and an index is a convenient tool for this selection in clinical practice,” the authors concluded.
ELSEVIER GLOBAL MEDICAL NEWS