Skip to navigation bar Skip to breadcrumbs Skip to page content
clear place holder
Envelope icon E-mail Updates Teal square Text size:  a A A

Evidence Summary

Other Supporting Document for Osteoporosis to Prevent Fractures: Screening

By Meera Viswanathan, PhD; Shivani Reddy, MD, MSc; Nancy Berkman, PhD; Katie Cullen, MEd; Jennifer Cook Middleton, PhD; Wanda K. Nicholson, MD, MPH, MBA; and Leila C. Kahwati, MD, MPH

The information in this article is intended to help clinicians, employers, policymakers, and others make informed decisions about the provision of health care services. This article is intended as a reference and not as a substitute for clinical judgment.

This article may be used, in whole or in part, as the basis for the development of clinical practice guidelines and other quality enhancement tools, or as a basis for reimbursement and coverage policies. AHRQ or U.S. Department of Health and Human Services endorsement of such derivative products may not be stated or implied.

This article was published in the Journal of the American Medical Association on June 26, 2018.

Abstract

Importance: Osteoporotic fractures cause significant morbidity and mortality.

Objective: To update the evidence on screening and treatment to prevent osteoporotic fractures for the US Preventive Services Task Force.

Data Sources: PubMed, the Cochrane Library, EMBASE, and trial registries (November 1, 2009, through October 1, 2016) and surveillance of the literature (through March 23, 2018); bibliographies from articles.

Study Selection: Adults 40 years and older; screening cohorts without prevalent low-trauma fractures or treatment cohorts with increased fracture risk; studies assessing screening, bone measurement tests or clinical risk assessments, pharmacologic treatment.

Data Extraction and Synthesis: Dual, independent review of titles/abstracts and full-text articles; study quality rating; random-effects meta-analysis.

Main Outcomes and Measures: Incident fractures and related morbidity and mortality, diagnostic and predictive accuracy, harms of screening or treatment.

Results: One hundred sixty-eight fair- or good-quality articles were included. One randomized clinical trial (RCT) (n = 12,483) comparing screening with no screening reported fewer hip fractures (2.6% vs 3.5%; hazard ratio [HR], 0.72 [95% CI, 0.59-0.89]) but no other statistically significant benefits or harms. The accuracy of bone measurement tests to identify osteoporosis varied (area under the curve [AUC], 0.32-0.89). The pooled accuracy of clinical risk assessments for identifying osteoporosis ranged from AUC of 0.65 to 0.76 in women and from 0.76 to 0.80 in men; the accuracy for predicting fractures was similar. For women, bisphosphonates, parathyroid hormone, raloxifene, and denosumab were associated with a lower risk of vertebral fractures (9 trials [n = 23,690]; relative risks [RRs] from 0.32-0.64). Bisphosphonates (8 RCTs [n = 16,438]; pooled RR, 0.84 [95% CI, 0.76-0.92]) and denosumab (1 RCT [n = 7868]; RR, 0.80 [95% CI, 0.67-0.95]) were associated with a lower risk of nonvertebral fractures. Denosumab reduced the risk of hip fracture (1 RCT [n = 7868]; RR, 0.60 [95% CI, 0.37-0.97]), but bisphosphonates did not have a statistically significant association (3 RCTs [n = 8988]; pooled RR, 0.70 [95% CI, 0.44-1.11]). Evidence was limited for men: zoledronic acid reduced the risk of radiographic vertebral fractures (1 RCT [n = 1199]; RR, 0.33 [95% CI, 0.16-0.70]); no studies demonstrated reductions in clinical or hip fractures. Bisphosphonates were not consistently associated with reported harms other than deep vein thrombosis (raloxifene vs placebo; 3 RCTs [n = 5839]; RR, 2.14 [95% CI, 0.99-4.66]).

Conclusions and Relevance: In women, screening to prevent osteoporotic fractures may reduce hip fractures, and treatment reduced the risk of vertebral and nonvertebral fractures; there was not consistent evidence of treatment harms. The accuracy of bone measurement tests or clinical risk assessments for identifying osteoporosis or predicting fractures varied from very poor to good.

Introduction

Screening to prevent osteoporotic fractures may reduce fracture-related morbidity and mortality.1-4 Screening involves clinical fracture risk assessment, bone measurement testing (eg, dual-energy x-ray absorptiometry [DXA]), or both. Pharmacologic treatments for osteoporosis inhibit osteoclastic bone resorption (antiresorptive agents) or stimulate osteoblastic new bone formation (anabolic agents).5

In 2011, the US Preventive Services Task Force (USPSTF) recommended screening for osteoporosis in women 65 years and older and in younger women whose fracture risk is equal to or greater than that of a 65-year-old white woman who has no additional risk factors (B recommendation).6 The USPSTF concluded that the evidence was insufficient to assess the balance of benefits and harms of screening in men.6 To inform an updated recommendation, the evidence about the benefits and harms of screening and treatment to prevent osteoporotic fractures in community-dwelling adults relevant to US primary care was reviewed.

Methods

Scope of Review

Detailed methods, calibration and reclassification outcomes, evidence tables, sensitivity analyses, and contextual information are available in the full evidence report at https://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/osteoporosis-screening1. The analytic framework and key questions (KQs) that guided the review are shown in Figure 1.

Data Sources and Searches

PubMed, the Cochrane Library, and Embase were searched for English-language articles published from November 1, 2009, through October 1, 2016, with active surveillance through March 23, 2018. ClinicalTrials.gov, Drugs@FDA.gov, HSRProj, Cochrane Clinical Trials Registry, and the World Health Organization International Clinical Trials Registry Platform were also searched. To supplement systematic electronic searches, studies included in relevant existing systematic reviews1, 8, 9 and reference lists of pertinent articles, and studies suggested by reviewers, were reviewed.

Study Selection

Two investigators independently reviewed titles, abstracts, and full-text articles using prespecified inclusion criteria for each KQ, with disagreements about inclusion resolved by discussion. For KQ1, KQ2, and KQ3 (benefits and harms of screening), studies for which the majority of participants were community-dwelling adults with no known low-trauma fractures or metabolic bone disease were included. For KQ4 and KQ5 (benefits and harms of treatment), studies were included if the majority of participants had an increased fracture risk.

Eligible screening tests included bone tests (eg, DXA, quantitative ultrasound) and clinical risk assessments for osteoporosis or fracture risk if externally validated and publicly available. Eligible treatments included US Food and Drug Administration (FDA)–approved pharmacotherapy (specifically, bisphosphonates, estrogen agonists/antagonists, estrogen- and/or progestin-based hormone therapy, parathyroid hormone, and RANK ligand inhibitors [eg, denosumab]). Eligible outcomes included diagnostic or predictive accuracy (as measured by area under the curve [AUC]), incident fractures, fracture-related morbidity or mortality, all-cause mortality, and harms.

Randomized clinical trials (RCTs) and systematic reviews were eligible for all KQs; observational study designs were also eligible for accuracy of screening (KQ2) and harms of screening and treatment (KQ3 and KQ5). Only studies published in English and conducted in countries categorized as “very high” by the 2015 Human Development Index were included.10

Data Extraction and Quality Assessment

For each included study, 1 investigator extracted information about design, population, intervention, and outcomes, and a second investigator reviewed for completeness and accuracy. Two independent investigators assessed the quality of each study as good, fair, or poor, using predefined criteria developed by the USPSTF7 and others for assessing the risk of bias of diagnostic tests,11 prognostic tests,12 trials,13 observational studies,14 and systematic reviews.11, 15

Data Synthesis and Analysis

Findings were qualitatively synthesized for each KQ in tabular and narrative formats. Studies were included if they met all study selection criteria and were fair or good quality; this included studies from the prior review that informed the USPSTF 2011 recommendation that continued to meet the study selection criteria for this update. When at least 3 independent and similar RCTs were available,16 random-effects models using the inverse-variance weighted method of DerSimonian and Laird was used to estimate pooled effects for pooled AUCs or relative risks.17 Statistical heterogeneity was assessed using the I2 statistic.13 All quantitative analyses were conducted using OpenMetaAnalyst or Comprehensive Meta Analysis.18, 19 The strength of evidence for each outcome was assessed based on the Agency for Healthcare Quality and Research Methods Guide for Effectiveness and Comparative Effectiveness Reviews,20 which specifies the assessment of study limitations, directness, consistency, precision, and reporting bias for each intervention comparison and major outcome of interest.

Results

A total of 168 articles of good or fair quality were included (Figure 2). Several cohorts of study participants contributed to multiple publications; as a result, the total number of participants cannot be calculated accurately.

Benefits of Screening

Key Question 1. Does screening (clinical risk assessment, bone density measurement, or both) for osteoporotic fracture risk reduce fractures and fracture-related morbidity and mortality in adults?

The Screening for Osteoporosis in Older Women for the Prevention of Fracture (SCOOP) trial randomized 12,483 women aged 70 to 85 years in the United Kingdom to screening with the Fracture Risk Assessment Tool (FRAX) or usual care (details not reported).21 In this fair-quality trial, participants in the intervention group who were identified as high risk based on FRAX-generated 10-year hip fracture risk were invited to undergo DXA testing. The investigators recalculated the FRAX risk for those who undertook DXA screening and communicated the results to the participant’s general practitioner, who then offered treatment as appropriate.21

At 5 years’ follow-up, comparing the intervention group with usual care, no difference was reported for the primary outcome of any osteoporotic fracture (12.9% vs 13.6%; hazard ratio [HR], 0.94 [95% CI, 0.85-1.03]), for all clinical fractures (15.3% vs 16.0%; HR, 0.94 [95% CI, 0.86-1.03]), or for mortality (8.8% vs 8.4%; HR, 1.05 [95% CI, 0.93-1.19]). However, a statistically significant difference in hip fracture incidence was observed (2.6% vs 3.5%; HR, 0.72 [95% CI, 0.59-0.89]).

Diagnostic and Predictive Accuracy of Screening

Key Question 2a. What is the accuracy and reliability of screening approaches to identify adults who are at increased risk for osteoporotic fracture?

Studies of tests to identify osteoporosis (as defined by bone mineral density [BMD] T-score ≤ −2.5), predict osteoporotic fracture, or both, were included. The results below focus primarily on pooled results; nonpooled results are available in the full evidence report.

Identifying Osteoporosis

Clinical Risk Assessments

Thirty-eight studies reported on the diagnostic accuracy of 16 clinical risk assessment instruments for identifying osteoporosis. In women, pooled AUC estimates ranged from 0.65 (95% CI, 0.60-0.71; I2 = 97.8%; 10 studies [16,780 participants]) for the Osteoporosis Risk Assessment Instrument [ORAI] to 0.76 (95% CI, 0.63-0.90; I2 = 98.5%; 4 studies [2692 participants]) for the Osteoporosis Self-Assessment Tool for Asians. AUCs from individual studies have a wider range in women (0.3222-0.8723) than in men (0.6224-0.8925). In men, the pooled AUC for the Osteoporosis Self-assessment Tool (OST) was 0.76 (95% CI, 0.71-0.80; I2 = 93.2%; 7 studies [7798 participants]); for the Male Osteoporosis Risk Estimation Score, the pooled AUC was 0.80 (95% CI, 0.71-0.88; I2 = 97.6%; 3 studies [4828 participants]). AUCs for FRAX could not be pooled but ranged from 0.5826 to 0.82.26

AUCs in younger women (<65 years) varied from 0.5826 to 0.85.27 One study found the accuracy of using the FRAX threshold associated with the 2011 USPSTF recommendation (10-year risk of major osteoporotic fracture ≤9.3%) was modestly better than chance (AUC, 0.60) and inferior to accuracy using the OST (AUC, 0.72) and Simple Calculated Osteoporosis Risk Estimation (AUC, 0.75) instruments in identifying women aged 50 to 64 years with osteoporosis (femoral neck T-score, −2.5).28 Instruments that assess more clinical risks did not report higher AUCs than instruments measuring fewer risks.

Thirty-five studies reported other measures of diagnostic accuracy (ie, sensitivity, specificity), but the instrument score threshold used to assess diagnostic accuracy varied considerably across studies. Even with a common threshold, results for the same instrument varied widely; as an example, the sensitivity of the ORAI instrument ranged from 50%29 to 100%30 and specificity from 10%30 to 75%.29

Bone Measurement Tests

Seven studies in women and 3 studies in men compared calcaneal quantitative ultrasound to centrally measured DXA for identifying osteoporosis. Reported AUCs varied from 0.69 to 0.90. For women, the pooled AUC estimate was 0.77 (95% CI, 0.72-0.81; I2 = 82.3%; 7 studies [1969 participants]). For men, the pooled AUC estimate was 0.80 (95% CI, 0.67-0.94; I2 = 98.2%; 3 studies [5142 participants]). Similar findings were observed for digital x-ray radiogrammetry, peripheral DXA, and radiographic absorptiometry.

Predicting Osteoporotic Fractures

Clinical Risk Assessments

One good-quality systematic review of 45 studies supplemented by 13 additional fair- or good-quality studies reported on the accuracy of 12 different clinical risk assessments for predicting incident fracture (Table 1). Pooled results are reported herein.

FRAX. The discriminative ability of FRAX for predicting future fracture varied by sex, site of fracture prediction, and whether BMD was used in the risk prediction. For women, pooled estimates based on 10 to 17 studies with 62,054 and 190,795 participants ranged somewhat higher (0.66-0.79). In men, pooled estimates of AUC from 3 to 44 studies (13,970-15,842 participants) ranged from 0.62 to 0.76 (depending on inclusion of BMD in the prediction model). Within that range, pooled estimates were higher for predicting hip fracture than for major osteoporotic fracture and higher when BMD was included in the prediction model. For cohorts of men and women combined, pooled estimates for the prediction of major osteoporotic fracture based on 3 studies (66,777 participants) were similar (AUC without BMD, 0.67 [95% CI, 0.66-0.67; I2 = 47.1%]; AUC with BMD, 0.69 [95% CI, 0.69-070; I2 = 70.3%]). Two studies predicting hip fracture in combined cohorts of men and women reported similar AUC estimates as women-only cohorts.54, 55

Garvan Fracture Risk Calculator. In women, the pooled AUC for risk assessment with BMD was 0.68 (95% CI, 0.64-0.71; I2 = 84.8%; 3 studies [6534 participants]) for predicting major osteoporotic fracture and 0.73 (95% CI, 0.66-0.79; I2 = 97.3%; 4 studies [7809 participants]) for predicting hip fracture.

Other Fracture Risk Assessment Instruments. Across 9 fracture risk assessment instruments (the Women’s Health Initiative algorithm,63 OST,65 Simple Calculated Osteoporosis Risk Estimation,67 Fracture and Immobilization Score,68 Fracture Risk Score,70 Fracture Risk Calculator,71 ORAI,74 QFracture,60 and Osteoporosis Index of Risk75), AUC estimates ranged from 0.53 to 0.82 for major osteoporotic fracture8, 36, 46, 47, 66 and from 0.80 to 0.89 for hip fracture.8, 63, 64, 73 A tenth instrument, the Canadian Association of Radiologists and Osteoporosis Canada, did not provide AUC estimates76 but reported a sensitivity for predicting fracture of 0.54 (95% CI, 0.52-0.56) among women and 0.31 (95% CI, 0.24-0.38) among men.77 The reported specificities were 0.75 (95% CI, 0.74-0.75) for women and 0.86 (0.85-0.87) for men.

Bone Measurement Tests

Twenty-three studies evaluated the accuracy of various bone measurement tests for predicting fracture (Table 2). In general, no meaningful differences in accuracy by type of bone test or by sex were observed. AUC estimates were generally higher for prediction of hip fracture than for prediction of fractures at other sites.

Key Question 2b. What is the evidence to determine screening intervals for osteoporosis and low bone density?

Two studies included participants with widely varying baseline BMD. Both suggest no advantage to repeated bone measurement testing (at 8 years86 and 3.7 years87 apart).87 However, 3 studies that developed prognostic models suggested that the optimal screening interval varies by baseline BMD.88-90 Age and use of hormone replacement therapy also influence optimal screening intervals.88, 89

Harms of Screening

Key Question 3. What are the harms of screening for osteoporotic fracture risk?

One trial, SCOOP (previously described in KQ1),21 assessed the effect of screening on anxiety (State-Trait Anxiety Inventory) and quality of life (EuroQol 5-Dimension tool and the Short-Form Health Survey 12 [physical and mental health]) and found no differences between participants allocated to screening vs usual care (variance not reported, P >0.10 for all outcomes).

Benefits of Treatment

Key Question 4a. What is the effectiveness of pharmacotherapy for the reduction of fractures and related morbidity and mortality?

Bisphosphonates

Eleven RCTs reported outcomes related to the effect of various bisphosphonates on fracture incidence.91-101

Vertebral Fracture. Among women, bisphosphonates (as a class) were associated with fewer vertebral fractures compared with placebo (2.1% vs 3.8% relative risk [RR], 0.57 [95% CI, 0.41-0.78]; I2 = 0.0%; 5 RCTs [5433 participants]).91-95

One RCT in 1199 men reported fewer radiographic vertebral fractures for zoledronic acid compared with placebo (1.5% vs 4.6%; RR, 0.33 [95% CI, 0.16-0.70]).101

Nonvertebral Fracture. Among women, a pooled analysis of RCTs reporting nonvertebral fractures observed an association with fewer fractures in the treatment group compared with placebo (8.9% vs 10.6%; RR, 0.84 [95% CI, 0.76-0.92]; I2 = 0.0%; 8 RCTs [16,438 participants]).91, 93-95, 99, 100, 102

The same trial of zoledronic acid in men previously described for vertebral fractures also reported on nonvertebral fractures; no between-group differences in incidence were observed (0.9% vs 1.3%; RR, 0.65 [95% CI, 0.21-1.97]).101

Hip Fractures. Among women, the pooled estimate suggested no statistically significant association between treatment with bisphosphonates and incidence of hip fracture (0.70% vs 0.96%; RR, 0.70 [95% CI, 0.44-1.11]; I2 = 0.0%; 3 RCTs [n = 8988]). Only 191 of the 3 studies91, 102, 103 was powered to detect differences in hip fractures.

No studies reported on hip fractures in men.

Raloxifene

Raloxifene (60 mg/d) reduced radiographic vertebral fracture (7.5% vs 12.5%; RR, 0.64 [95% CI, 0.53-0.76]) compared with placebo in 1 RCT of 7705 women.104, 105 Treatment with raloxifene (60 mg/d or 120 mg/d) did not have an effect on incidence of nonvertebral or hip fracture.

Estrogen

A recently completed systematic review on the benefits and harms of estrogen therapy, with and without progestin, in primary care populations incorporated information from the Women’s Health Initiative and other similar trials.106 Women taking only estrogen had lower risks for total osteoporotic fractures (HR, 0.72 [95% CI, 0.64-0.80]) compared with women taking placebo. Women taking estrogen plus progestin therapy also had lower risks for fractures (RR, 0.80 [95% CI, 0.68-0.94]) compared with women taking placebo.

Denosumab

One large study107 (7868 women) demonstrated a statistically significant difference between denosumab and placebo in incident vertebral fractures (2.3% vs 7.2%; RR, 0.32 [95% CI, 0.26-0.41]), nonvertebral fractures (6.1% vs 7.5%; RR, 0.80 [95% CI, 0.67-0.95]), and hip fractures (0.7% vs 1.1%; RR, 0.60 [95% CI, 0.37-0.97]). Three smaller RCTs of denosumab reported no effect of treatment on incident clinical, osteoporotic, or vertebral fractures.

Parathyroid Hormone

Vertebral Fractures. Among 2061 women without a prevalent fracture at baseline, parathyroid hormone produced a significant (0.7% vs 2.1%; RR, 0.32 [95% CI, 0.14-0.75]) reduction in new radiographic vertebral fractures compared with placebo.108 No studies met the inclusion criteria to assess the effects of teriparatide on vertebral fractures in men.

Nonvertebral Fractures. In an RCT of 2532 women with and without prevalent fractures at baseline,108 no significant difference in new nonvertebral fractures was observed between treatment and placebo (5.6% vs 5.8%; RR, 0.97 [95% CI, 0.71-1.33]).

One trial of men reported a reduction in nonvertebral fractures in both treatment groups of teriparatide (doses of 20 μg [the FDA-approved dose] [n = 151 men] or 40 μg [n = 139 men] compared with placebo [n = 147 men]),109 although the results did not reach statistical significance because of a small number of fractures and early termination of the study for safety concerns (20 μg vs placebo: 1.3% vs 2.0%; RR, 0.65 [95% CI, 0.11-3.83]; 40 μg vs placebo: 0.7% vs 2.0%; RR, 0.35 [95% CI, 0.04-3.35]).

Key Question 4b. How does the effectiveness of pharmacotherapy for the reduction of fractures and related morbidity and mortality vary by subgroup?

One trial each offered further analyses on subgroups for alendronate,91 risedronate,103 raloxifene,110 and denosumab.111, 112 None reported differences in effectiveness by age, baseline BMD, prior fractures, or a combination of risk factors.

Harms of Treatment

Key Question 5. What are the harms associated with pharmacotherapy?

Bisphosphonates

When comparing medication with placebo, there was no significant association between use of bisphosphonates and discontinuation (RR, 0.99 [95% CI, 0.91-1.07]; I2 = 0.0%; 20 RCTs [17,369 participants]), serious adverse events (RR, 0.98 [95% CI, 0.92-1.04]; I2 = 0.0%; 17 RCTs [11,745 participants]), or upper gastrointestinal events (RR, 1.01 [95% CI, 0.98-1.05]; I2 = 0.0%; 13 RCTs [20,485 participants]) for any individual bisphosphonate drug or overall as a class.

Two studies did not report a statistically significant risk of atrial fibrillation with bisphosphonates compared with placebo. One study was in women (alendronate: 2.5% vs 2.2%; RR, 1.14 [95% CI, 0.83-1.56]),113 and 1 study was in men (zoledronic acid: 1.2% vs 0.8%; RR, 1.45 [95% CI, 0.46-4.56]).101 Two studies of women reported no cases of atrial fibrillation.114, 115 A case-control study using a Danish registry reported a relative risk of atrial fibrillation of 0.75 (95% CI, 0.49-1.16; 3.2% vs 2.9%) for new users of bisphosphonates.116

Rare outcomes were not generally observed in the included evidence. Specifically, 3 studies (1 in men and 2 in women) reported that they found no cases of osteonecrosis of the jaw.101, 114, 115 No studies included in the review reported atypical femur fracture outcomes or kidney failure.

Raloxifene

Pooled estimates of women followed up from 1 to 4 years and randomized to raloxifene or placebo found no significant association between raloxifene use and discontinuation of treatment because of adverse events (12.6% vs 11.2%; RR, 1.12 [95% CI, 0.98-1.28]; I2 = 0.0%; 6 RCTs [6438 participants]). The pooled analysis suggested a possible association between raloxifene use and deep vein thromboses (0.7% vs 0.3%; RR, 2.14 [95% CI, 0.99-4.66]; I2 = 0.0%; 3 RCTs [5839 participants]), an association between use and hot flashes (11.2% vs 7.6%; RR, 1.42 [95% CI, 1.22-1.66]; I2 = 0.0%; 5 RCTs [n = 6249 participants]), but no association between use and leg cramps (8.0% vs 4.8%; RR, 1.41 [95% CI, 0.92-2.14]; I2 = 67.1%; 3 RCTs [n = 6000]). No significant association between raloxifene and coronary heart disease (1.0% vs 1.1%; HR, 0.88 [95% CI, 0.56-1.40]),117 stroke (0.9% vs 1.2%; RR, 0.69 [95% CI, 0.40-1.18]),118 or endometrial cancer (0.2% vs 0.2%; RR, 1.01 [95% CI, 0.29-3.48])119 was observed.

Estrogen

A recently completed review on the benefits and harms of estrogen therapy, with and without progestin, in primary care populations found that compared with women receiving placebo, women receiving estrogen with or without progesterone experienced a higher rate of gallbladder events, stroke, and venous thromboembolism over 5-year follow-up106 and an increased risk of urinary incontinence during follow-up of 1 year. In addition, women receiving estrogen plus progestin, compared with women receiving placebo, were found to have a higher risk of invasive breast cancer, coronary heart disease, and probable dementia over 5-year follow-up.

Denosumab

Pooled estimates suggested no significant association between denosumab use and discontinuation because of adverse events (2.4% vs 2.1%; RR, 1.14 [95% CI, 0.85-1.52]; I2 = 0.0%; 3 RCTs [8451 participants]) or serious adverse events (23.8% vs 23.9%; RR, 1.12 [95% CI, 0.88-1.44]; I2 = 14.1%; 4 RCTs [8663 participants]). Although treatment groups had higher rates of serious infections than placebo groups, confidence intervals for the pooled estimate spanned the null effect (4.0% vs 3.3%; RR, 1.89 [95% CI, 0.61-5.91]; I2 = 40.1%).

Parathyroid Hormone

Among 2532 postmenopausal women in 1 study,108 the treatment group had higher rates of discontinuation because of adverse events when compared with the placebo group (30.2% vs 24.6%; RR, 1.22 [95% CI, 1.08-1.40]). Hypercalcemia, hypercalciuria, nausea, and vomiting were more common in the treatment group compared with placebo.

In 1 RCT among 437 men,109 both the 20-μg and 40-μg treatment groups had a higher proportion of withdrawals than the placebo group (9.2% and 12.9%, respectively, vs 4.8%), which was statistically significantly higher in the 40-μg treatment group than in the placebo group (RR, 2.72 [95% CI, 1.17-6.31]) but not in the group receiving the FDA-approved dose of 20 μg (RR, 1.94 [95% CI, 0.81-4.69]). Cancers were reported in 2 groups (3/147 in the placebo group and 3/151 in the 20-μg treatment group), but none were reported as osteosarcomas.

Discussion

Table 3 and Table 4 summarize the strength of evidence and findings from this review. This updated review for the UPSPTF incorporates new evidence on the direct link between screening for osteoporosis and health outcomes. One trial (SCOOP) addressed the morbidity, mortality, and harms associated with screening to prevent osteoporotic fractures (KQ1, KQ3). The trial found evidence of benefit for a secondary outcome only, the incidence of hip fractures (low strength of evidence of benefit). For all other outcomes (osteoporosis-related fractures, clinical fractures, mortality, anxiety, quality of life; insufficient strength of evidence), the trial did not report statistically significant differences in benefits or harms. The release of guidelines during trial recruitment122 and observation123 may have changed standards for usual care; differences between study groups may have been attenuated as a result. The use of the 10-year risk of hip fracture rather than the risk of major osteoporotic fracture as the threshold for DXA testing may have increased the likelihood of effectiveness for preventing hip rather than other fractures (given that risks of hip and other fractures are correlated but not identical21). The discrepancy in results between the hip fracture outcomes and other fractures points to the need for caution in interpreting the results.

Although results from studies of accuracy of bone measurement tests or clinical risk assessments for identifying osteoporosis or predicting fractures vary, in general they report no more than moderate accuracy (KQ2), and this evidence was graded as low to moderate. On average, clinical risk assessment tools to identify osteoporosis performed better in men than in women. FRAX performed only better than chance in younger women. Predictions of hip fractures were more accurate than prediction of fractures at other sites or composite fracture outcomes. Sixteen clinical risk assessment tools for the identification of osteoporosis were found, and although these instruments had many risk variables in common (eg, age, weight, hormone therapy use), there was considerable heterogeneity in the patient populations studied and the anatomical sites used to measure bone density. The evidence for clinical risk assessments varies in the incorporation of BMD and number of risks. In general, tools incorporating BMD had higher accuracy than tools without BMD. The accuracy of tools with more clinical variables was similar to the accuracy of tools with fewer risk factors, suggesting that future research could focus on simpler instruments that can be easily incorporated into clinical practice. Future study into the optimal thresholds to trigger further diagnostic evaluation (eg, DXA testing) or to begin treatment is also critical, because valid and reliable cutoffs for high and low-risk categories are necessary for clinical decision making.

Pharmacotherapy treatment studies in women show that multiple classes of medications (bisphosphonates, parathyroid hormone, raloxifene, and denosumab) reduce the risk of vertebral and nonvertebral fractures (KQ4); this evidence was graded as low to moderate for reducing fractures. Two of 3 studies of bisphosphonates that reported hip fractures were not powered to detect effects on hip fractures; the pooled evidence did not demonstrate a statistically significant benefit. Evidence for benefit in men is limited to 1 trial of a bisphosphonate, which demonstrated a large reduction in radiographic vertebral fractures. No studies demonstrated reductions in risk of clinical vertebral fractures or nonvertebral fractures for men. No studies reporting on hip fractures, fracture-related morbidity, or mortality were identified.

Although several trials reported on harms (KQ5), they varied substantially in definitions used. No consistent evidence of harms with bisphosphonates (strength of evidence graded as moderate) was identified and no bisphosphonate trials reported rare harms, such as osteonecrosis of the jaw, atypical femur fractures, or kidney failure. The evidence on harms in men was very limited but was consistent with harms for women when available.

Limitations

This review has several limitations. First, it focused on treatment with prescription medications only; it does not address other interventions that might reduce the risk of osteoporotic fractures, such as functional assessment, safety evaluations, vision examinations, exercise or physical therapy, vitamin supplementation, and diet interventions. Further, this review did not consider comparative effectiveness of pharmacologic treatment.

Second, treatment studies included in this review relied on BMD T-scores to enroll participants into trials. Risk factors beyond bone density, such as microarchitectural deterioration of bone tissue and decline in bone quality, contribute to osteoporotic fractures; therefore, approaches that rely on BMD measurement wholly or in part may not be the most accurate approaches for identifying patients at highest risk for osteoporotic fractures.

Third, included studies on diagnosing osteoporosis or predicting fractures are heterogeneous with respect to prevalence of baseline fractures, baseline BMD, prior treatment, and length of study follow-up (which was sometimes shorter than the time horizon of the risk prediction instrument); most meta-analyses demonstrated high statistical heterogeneity (I2 >80%), suggesting that the variance can be explained by heterogeneity rather than chance. Fourth, the evidence base is sparse on screening interval, screening in men and premenopausal women, and long-term studies on the harms of screening and treatment.

Conclusions

In women, screening to prevent osteoporotic fractures may reduce hip fractures, and treatment reduced the risk of vertebral and nonvertebral fractures; there was not consistent evidence of treatment harms. The accuracy of bone measurement tests or clinical risk assessments for identifying osteoporosis or predicting fractures varied from very poor to good.

Article Information

Source: This article was first published in the Journal of the American Medical Association on June 26, 2018 (JAMA. 2018;319(24):2532-2551). doi:10.1001/jama.2018.6537

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: This research was funded under contract HHSA-290-2012-00015-I, Task Order No. 6, from the Agency for Healthcare Research and Quality (AHRQ), US Department of Health and Human Services, under a contract to support the USPSTF.

Role of the Funder/Sponsor: Investigators worked with USPSTF members and AHRQ staff to develop the scope, analytic framework, and key questions for this review. AHRQ had no role in study selection, quality assessment, or synthesis. AHRQ staff provided project oversight, reviewed the report to ensure that the analysis met methodological standards; and distributed the draft for peer review. Otherwise, AHRQ had no role in the conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript findings. The opinions expressed in this document are those of the authors and do not reflect the official position of AHRQ or the US Department of Health and Human Services.

Additional Information: A draft version of the full evidence report underwent external peer review from 5 content experts (Rosanne Leipzig, MD, Mount Sinai Medical Center; Diana Pettiti, MD, Arizona State University; Margaret Gourlay, MD, University of North Carolina at Chapel Hill; Carolyn Crandall, MD, University of California, Los Angeles; Mary Roary, PhD, National Institutes of Health) and 4 anonymous reviewers. Comments from reviewers were presented to the USPSTF during its deliberation of the evidence and were considered in preparing the final evidence review.

References:

1. Nelson HD, Haney EM, Dana T, Bougatsos C, Chou R. Screening for osteoporosis: an update for the U.S. Preventive Services Task Force. Ann Intern Med. 2010;153(2):99-111.
2. Nelson HD, Haney EM, Chou R, Dana T, Fu R, Bougatsos C. Screening for Osteoporosis: Systematic Review to Update the 2002 U.S. Preventive Services Task Force Recommendation: Evidence Synthesis No. 77. Rockville, MD: Agency for Healthcare Research and Quality; 2010. AHRQ publication 10-05145-EF-1.
3. Committee on Practice Bulletins—Gynecology, the American College of Obstetricians and Gynecologists. ACOG practice bulletin No. 129: osteoporosis. Obstet Gynecol. 2012;120(3):718-734.
4. Ensrud KE. Epidemiology of fracture risk with advancing age. J Gerontol A Biol Sci Med Sci. 2013;68(10):1236-1242.
5. Langdahl B, Ferrari S, Dempster DW. Bone modeling and remodeling: potential as therapeutic targets for the treatment of osteoporosis. Ther Adv Musculoskelet Dis. 2016;8(6):225-235.
6. U.S. Preventive Services Task Force. Screening for osteoporosis: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2011;154(5):356-364.
7. US Preventive Services Task Force (USPSTF). Procedure Manual. USPSTF website. https://www.uspreventiveservicestaskforce.org/Page/Name/methods-and-processes. Published 2015. Accessed August 24, 2016.
8. Marques A, Ferreira RJ, Santos E, Loza E, Carmona L, da Silva JA. The accuracy of osteoporotic fracture risk prediction tools: a systematic review and meta-analysis. Ann Rheum Dis. 2015;74(11):1958-1967.
9. Crandall CJ, Newberry SJ, Diamant A, et al. Treatment to Prevent Fractures in Men and Women With Low Bone Density or Osteoporosis: Update of a 2007 Report. Comparative Effectiveness Review No. 53. Rockville, MD: Agency for Healthcare Research and Quality; 2012. AHRQ publication 12-EHC023-EF.
10. United Nations Development Programme (UNDP). Human Development Report 2015: work for human development: Table 1: Human Development Index and its components. UNDP website. http://hdr.undp.org/en/composite/HDI. Published 2015. Accessed August 24, 2016.
11. Whiting PF, Rutjes AW, Westwood ME, et al; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536.
12. Wolf R. PROBAST project summary. Kleijnen Systematic Reviews website. http://www.systematic-reviews.com/probast. Published 2014. Accessed January 7, 2015.
13. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane website. http://handbook.cochrane.org. Published 2011. Accessed August 24, 2016.
14. Viswanathan M, Berkman ND. Development of the RTI item bank on risk of bias and precision of observational studies. J Clin Epidemiol. 2012;65(2):163-178.
15. International Society for Clinical Densitometry (ISCD). 2015 ISCD Official Positions—Adult. ISCD website. http://www.iscd.org/official-positions /2015-iscd-official-positions-adult/. Published 2015. Accessed August 10, 2016.
16. West SL, Gartlehner G, Mansfield AJ, et al. Comparative Effectiveness Review Methods: Clinical Heterogeneity. Rockville MD: Agency for Healthcare Research and Quality; 2010. AHRQ publication 10-EHC070-EF.
17. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177-188.
18. Wallace BC, Dahabreh IJ, Trikalinos TA, Lau J, Trow P, Schmid CH. Closing the gap between methodologists and end-users: R as a computational back-end. J Stat Softw. 2012;49(5):1-15.
19. Comprehensive Meta Analysis version 3.3.070. [software program] Englewood, NJ: Biostat; 2014.
20. Agency for Healthcare Research and Quality. Methods Guide for Effectiveness and Comparative Effectiveness Reviews. Rockville, MD: Agency for Healthcare Research and Quality; 2014. AHRQ publication 10(14)-EHC063-EF.
21. Shepstone L, Lenaghan E, Cooper C, et al; SCOOP Study Team. Screening in the community to reduce fractures in older women (SCOOP): a randomised controlled trial. Lancet. 2018;391(10122):741-747.
22. D’Amelio P, Tamone C, Pluviano F, Di Stefano M, Isaia G. Effects of lifestyle and risk factors on bone mineral density in a cohort of Italian women: suggestion for a new decision rule. Calcif Tissue Int. 2005;77(2):72-78.
23. Park HM, Sedrine WB, Reginster JY, Ross PD; OSTA. Korean experience with the OSTA risk index for osteoporosis: a validation study. J Clin Densitom. 2003;6(3):247-250.
24. Machado P, Coutinho M, da Silva JA. Selecting men for bone densitometry: performance of osteoporosis risk assessment tools in Portuguese men. Osteoporos Int. 2010;21(6):977-983.
25. Sinnott B, Kukreja S, Barengolts E. Utility of screening tools for the prediction of low bone mass in African American men. Osteoporos Int. 2006;17(5):684-692.
26. Bansal S, Pecina JL, Merry SP, et al. US Preventative Services Task Force FRAX threshold has a low sensitivity to detect osteoporosis in women ages 50-64 years. Osteoporos Int. 2015;26(4):1429-1433.
27. Mauck KF, Cuddihy MT, Atkinson EJ, Melton LJ III. Use of clinical prediction rules in detecting osteoporosis in a population-based sample of postmenopausal women. Arch Intern Med. 2005;165(5):530-536.
28. Crandall CJ, Larson J, Gourlay ML, et al. Osteoporosis screening in postmenopausal women 50 to 64 years old: comparison of US Preventive Services Task Force strategy and two traditional strategies in the Women’s Health Initiative. J Bone Miner Res. 2014;29(7):1661-1666.
29. Rud B, Jensen JE, Mosekilde L, Nielsen SP, Hilden J, Abrahamsen B. Performance of four clinical screening tools to select peri- and early postmenopausal women for dual X-ray absorptiometry. Osteoporos Int. 2005;16(7):764-772.
30. Chan SP, Teo CC, Ng SA, Goh N, Tan C, Deurenberg-Yap M. Validation of various osteoporosis risk indices in elderly Chinese females in Singapore. Osteoporos Int. 2006;17(8):1182-1188.
31. Kanis JA, Oden A, Johnell O, et al. The use of clinical risk factors enhances the performance of BMD in the prediction of hip and osteoporotic fractures in men and women. Osteoporos Int. 2007;18(8):1033-1046.
32. Friis-Holmberg T, Rubin KH, Brixen K, Tolstrup JS, Bech M. Fracture risk prediction using phalangeal bone mineral density or FRAX(®)?—a Danish cohort study on men and women. J Clin Densitom. 2014;17(1):7-15.
33. Ettinger B, Ensrud KE, Blackwell T, Curtis JR, Lapidus JA, Orwoll ES; Osteoporotic Fracture in Men (MrOS) Study Research Group. Performance of FRAX in a cohort of community-dwelling, ambulatory older men: the Osteoporotic Fractures in Men (MrOS) study. Osteoporos Int. 2013;24(4):1185-1193.
34. Leslie WD, Morin S, Lix LM, et al; Manitoba Bone Density Program. Fracture risk assessment without bone density measurement in routine clinical practice. Osteoporos Int. 2012;23(1):75-85.
35. Iki M, Fujita Y, Tamaki J, et al. Trabecular bone score may improve FRAX® prediction accuracy for major osteoporotic fractures in elderly Japanese men: the Fujiwara-kyo Osteoporosis Risk in Men (FORMEN) Cohort Study. Osteoporos Int. 2015;26(6):1841-1848.
36. Henry MJ, Pasco JA, Merriman EN, et al. Fracture risk score and absolute risk of fracture. Radiology. 2011;259(2):495-501.
37. van Geel TA, Eisman JA, Geusens PP, van den Bergh JP, Center JR, Dinant GJ. The utility of absolute risk prediction using FRAX® and Garvan Fracture Risk Calculator in daily practice. Maturitas. 2014;77(2):174-179.
38. Tebé Cordomí C, Del Río LM, Di Gregorio S, et al. Validation of the FRAX predictive model for major osteoporotic fracture in a historical cohort of Spanish women. J Clin Densitom. 2013;16(2):231-237.
39. Cheung EY, Bow CH, Cheung CL, et al. Discriminative value of FRAX for fracture prediction in a cohort of Chinese postmenopausal women. Osteoporos Int. 2012;23(3):871-878.
40. Ensrud KE, Lui LY, Taylor BC, et al; Study of Osteoporotic Fractures Research Group. A comparison of prediction models for fractures in older women: is more better? Arch Intern Med. 2009;169(22):2087-2094.
41. Donaldson MG, Palermo L, Schousboe JT, Ensrud KE, Hochberg MC, Cummings SR. FRAX and risk of vertebral fractures: the fracture intervention trial. J Bone Miner Res. 2009;24(11):1793-1799.
42. Azagra R, Roca G, Encabo G, et al. FRAX® tool, the WHO algorithm to predict osteoporotic fractures: the first analysis of its discriminative and predictive ability in the Spanish FRIDEX cohort. BMC Musculoskelet Disord. 2012;13:204.
43. Bolland MJ, Siu AT, Mason BH, et al. Evaluation of the FRAX and Garvan fracture risk calculators in older women. J Bone Miner Res. 2011;26(2):420-427.
44. Sornay-Rendu E, Munoz F, Delmas PD, Chapurlat RD. The FRAX tool in French women: how well does it describe the real incidence of fracture in the OFELY cohort? J Bone Miner Res. 2010;25(10):2101-2107.
45. Tamaki J, Iki M, Kadowaki E, et al. Fracture risk prediction using FRAX®: a 10-year follow-up survey of the Japanese Population-Based Osteoporosis (JPOS) Cohort Study. Osteoporos Int. 2011;22(12):3037-3045.
46. Rubin KH, Abrahamsen B, Friis-Holmberg T, et al. Comparison of different screening tools (FRAX®, OST, ORAI, OSIRIS, SCORE and age alone) to identify women with increased risk of fracture: a population-based prospective study. Bone. 2013;56(1):16-22.
47. Crandall CJ, Larson JC, Watts NB, et al. Comparison of fracture risk prediction by the US Preventive Services Task Force strategy and two alternative strategies in women 50-64 years old in the Women’s Health Initiative. J Clin Endocrinol Metab. 2014;99(12):4514-4522.
48. González-Macías J, Marin F, Vila J, Díez-Pérez A. Probability of fractures predicted by FRAX® and observed incidence in the Spanish ECOSAP Study cohort. Bone. 2012;50(1):373-377.
49. Trémollieres FA, Pouillès JM, Drewniak N, Laparra J, Ribot CA, Dargent-Molina P. Fracture risk prediction using BMD and clinical risk factors in early postmenopausal women: sensitivity of the WHO FRAX tool. J Bone Miner Res. 2010;25(5):1002-1009.
50. Sambrook PN, Flahive J, Hooven FH, et al. Predicting fractures in an international cohort using risk factor algorithms without BMD. J Bone Miner Res. 2011;26(11):2770-2777.
51. Kälvesten J, Lui LY, Brismar T, Cummings S. Digital X-ray radiogrammetry in the study of osteoporotic fractures: comparison to dual energy X-ray absorptiometry and FRAX. Bone. 2016;86:30-35.
52. Pressman AR, Lo JC, Chandra M, Ettinger B. Methods for assessing fracture risk prediction models: experience with FRAX in a large integrated health care delivery system. J Clin Densitom. 2011;14(4):407-415.
53. Sund R, Honkanen R, Johansson H, et al. Evaluation of the FRAX model for hip fracture predictions in the population-based Kuopio Osteoporosis Risk Factor and Prevention Study (OSTPRE). Calcif Tissue Int. 2014;95(1):39-45.
54. Fraser LA, Langsetmo L, Berger C, et al; CaMos Research Group. Fracture prediction and calibration of a Canadian FRAX® tool: a population-based report from CaMos. Osteoporos Int. 2011;22(3):829-837.
55. Leslie WD, Lix LM, Johansson H, Oden A, McCloskey E, Kanis JA; Manitoba Bone Density Program. Independent clinical validation of a Canadian FRAX tool: fracture prediction and model calibration. J Bone Miner Res. 2010;25(11):2350-2358.
56. Leslie WD, Lix LM, Johansson H, Oden A, McCloskey E, Kanis JA; Manitoba Bone Density Program. A comparative study of using non-hip bone density inputs with FRAX®. Osteoporos Int. 2012;23(3):853-860.
57. Nguyen ND, Frost SA, Center JR, Eisman JA, Nguyen TV. Development of prognostic nomograms for individualizing 5-year and 10-year fracture risks. Osteoporos Int. 2008;19(10):1431-1444.
58. Langsetmo L, Nguyen TV, Nguyen ND, et al; Canadian Multicentre Osteoporosis Study Research Group. Independent external validation of nomograms for predicting risk of low-trauma fracture and hip fracture. CMAJ. 2011;183(2):E107-E114.
59. Ahmed LA, Nguyen ND, Bjørnerem Å, et al. External validation of the Garvan nomograms for predicting absolute fracture risk: the Tromsø study. PLoS One. 2014;9(9):e107695.
60. Hippisley-Cox J, Coupland C. Predicting risk of osteoporotic fracture in men and women in England and Wales: prospective derivation and validation of QFracture Scores. BMJ. 2009;339:b4229.
61. Collins GS, Mallett S, Altman DG. Predicting risk of osteoporotic and hip fracture in the United Kingdom: prospective independent and external validation of QFracture Scores. BMJ. 2011;342:d3651.
62. Hippisley-Cox J, Coupland C. Derivation and validation of updated QFracture algorithm to predict risk of osteoporotic fracture in primary care in the United Kingdom: prospective open cohort study. BMJ. 2012;344:e3427.
63. Robbins J, Aragaki AK, Kooperberg C, et al. Factors associated with 5-year risk of hip fracture in postmenopausal women. JAMA. 2007;298(20):2389-2398.
64. Hundrup YA, Jacobsen RK, Andreasen AH, Davidsen M, Obel EB, Abrahamsen B. Validation of a 5-year risk score of hip fracture in postmenopausal women: the Danish Nurse Cohort Study. Osteoporos Int. 2010;21(12):2135-2142.
65. Koh LKH, Sedrine WB, Torralba TP, et al; Osteoporosis Self-Assessment Tool for Asians (OSTA) Research Group. A simple tool to identify Asian women at increased risk of osteoporosis. Osteoporos Int. 2001;12(8):699-705.
66. Morin S, Tsang JF, Leslie WD. Weight and body mass index predict bone mineral density and fractures in women aged 40 to 59 years. Osteoporos Int. 2009;20(3):363-370.
67. Lydick E, Cook K, Turpin J, Melton M, Stine R, Byrnes C. Development and validation of a simple questionnaire to facilitate identification of women likely to have low bone density. Am J Manag Care. 1998;4(1):37-48.
68. Tanaka S, Yoshimura N, Kuroda T, Hosoi T, Saito M, Shiraki M. The Fracture and Immobilization Score (FRISC) for risk assessment of osteoporotic fracture and immobilization in postmenopausal women—a joint analysis of the Nagano, Miyama, and Taiji Cohorts. Bone. 2010;47(6):1064-1070.
69. Tanaka S, Kuroda T, Saito M, Shiraki M. Urinary pentosidine improves risk classification using fracture risk assessment tools for postmenopausal women. J Bone Miner Res. 2011;26(11):2778-2784.
70. Henry MJ, Pasco JA, Sanders KM, Nicholson GC, Kotowicz MA. Fracture Risk (FRISK) score: Geelong Osteoporosis Study. Radiology. 2006;241(1):190-196.
71. Ettinger B, Hillier TA, Pressman A, Che M, Hanley DA. Simple computer model for calculating and reporting 5-year osteoporotic fracture risk in postmenopausal women. J Womens Health (Larchmt). 2005;14(2):159-171.
72. Ettinger B, Liu H, Blackwell T, Hoffman AR, Ensrud KE, Orwoll ES; Osteoporotic Fracture in Men (MrOS) Research Group. Validation of FRC, a fracture risk assessment tool, in a cohort of older men: the Osteoporotic Fractures in Men (MrOS) Study. J Clin Densitom. 2012;15(3):334-342.
73. Lo JC, Pressman AR, Chandra M, Ettinger B. Fracture risk tool validation in an integrated healthcare delivery system. Am J Manag Care. 2011;17(3):188-194.
74. Cadarette SM, Jaglal SB, Kreiger N, McIsaac WJ, Darlington GA, Tu JV. Development and validation of the Osteoporosis Risk Assessment Instrument to facilitate selection of women for bone densitometry. CMAJ. 2000;162(9):1289-1294.
75. Sedrine WB, Chevallier T, Zegels B, et al. Development and assessment of the Osteoporosis Index of Risk (OSIRIS) to facilitate selection of women for bone densitometry. Gynecol Endocrinol. 2002;16(3):245-250.
76. Siminoski K, Leslie WD, Frame H, et al. Recommendations for bone mineral density reporting in Canada: a shift to absolute fracture risk assessment. J Clin Densitom. 2007;10(2):120-123.
77. Leslie WD, Lix LM; Manitoba Bone Density Program. Simplified 10-year absolute fracture risk assessment: a comparison of men and women. J Clin Densitom. 2010;13(2):141-146.
78. Stewart A, Kumar V, Reid DM. Long-term fracture prediction by DXA and QUS: a 10-year prospective study. J Bone Miner Res. 2006;21(3):413-418.
79. Hans D, Goertzen AL, Krieg MA, Leslie WD. Bone microarchitecture assessed by TBS predicts osteoporotic fractures independent of bone density: the Manitoba study. J Bone Miner Res. 2011; 26(11):2762-2769.
80. Leslie WD, Aubry-Rozier B, Lamy O, Hans D; Manitoba Bone Density Program. TBS (trabecular bone score) and diabetes-related fracture risk. J Clin Endocrinol Metab. 2013;98(2):602-609.
81. Nguyen TV, Center JR, Eisman JA. Bone mineral density-independent association of quantitative ultrasound measurements and fracture risk in women. Osteoporos Int. 2004;15(12):942-947.
82. Kwok T, Khoo CC, Leung J, et al. Predictive values of calcaneal quantitative ultrasound and dual energy X ray absorptiometry for non-vertebral fracture in older men: results from the MrOS study (Hong Kong). Osteoporos Int. 2012;23(3):1001-1006.
83. Chan MY, Nguyen ND, Center JR, Eisman JA, Nguyen TV. Absolute fracture risk prediction by a combination of calcaneal quantitative ultrasound and bone mineral density. Calcif Tissue Int. 2012;90(2):128-136.
84. Bauer DC, Ewing SK, Cauley JA, Ensrud KE, Cummings SR, Orwoll ES; Osteoporotic Fractures in Men (MrOS) Research Group. Quantitative ultrasound predicts hip and non-spine fracture in men: the MrOS study. Osteoporos Int. 2007;18(6):771-777.
85. Iki M, Tamaki J, Kadowaki E, et al. Trabecular bone score (TBS) predicts vertebral fractures in Japanese women over 10 years independently of bone density and prevalent vertebral deformity: the Japanese Population-Based Osteoporosis (JPOS) cohort study. J Bone Miner Res. 2014;29(2):399-407.
86. Hillier TA, Stone KL, Bauer DC, et al. Evaluating the value of repeat bone mineral density measurement and prediction of fractures in older women: the study of osteoporotic fractures. Arch Intern Med. 2007;167(2):155-160.
87. Berry SD, Samelson EJ, Pencina MJ, et al. Repeat bone mineral density screening and prediction of hip and major osteoporotic fracture. JAMA. 2013;310(12):1256-1262.
88. Gourlay ML, Fine JP, Preisser JS, et al; Study of Osteoporotic Fractures Research Group. Bone-density testing interval and transition to osteoporosis in older women. N Engl J Med. 2012;366(3):225-233.
89. Frost SA, Nguyen ND, Center JR, Eisman JA, Nguyen TV. Timing of repeat BMD measurements: development of an absolute risk-based prognostic model. J Bone Miner Res. 2009;24(11):1800-1807.
90. Gourlay ML, Overman RA, Fine JP, et al; Women’s Health Initiative Investigators. Baseline age and time to major fracture in younger postmenopausal women. Menopause. 2015;22(6):589-597.
91. Cummings SR, Black DM, Thompson DE, et al. Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. JAMA. 1998;280(24):2077-2082.
92. Liberman UA, Weiss SR, Bröll J, et al; Alendronate Phase III Osteoporosis Treatment Study Group. Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N Engl J Med. 1995;333(22):1437-1443.
93. Meunier PJ, Confavreux E, Tupinon I, Hardouin C, Delmas PD, Balena R. Prevention of early postmenopausal bone loss with cyclical etidronate therapy (a double-blind, placebo-controlled study and 1-year follow-up). J Clin Endocrinol Metab. 1997;82(9):2784-2791.
94. Mortensen L, Charles P, Bekker PJ, Digennaro J, Johnston CC Jr. Risedronate increases bone mass in an early postmenopausal population: two years of treatment plus one year of follow-up. J Clin Endocrinol Metab. 1998;83(2):396-402.
95. Fogelman I, Ribot C, Smith R, Ethgen D, Sod E, Reginster JY; BMD-MN Study Group. Risedronate reverses bone loss in postmenopausal women with low bone mass: results from a multinational, double-blind, placebo-controlled trial. J Clin Endocrinol Metab. 2000;85(5):1895-1900.
96. Ascott-Evans BH, Guanabens N, Kivinen S, et al. Alendronate prevents loss of bone density associated with discontinuation of hormone replacement therapy: a randomized controlled trial. Arch Intern Med. 2003;163(7):789-794.
97. Chesnut CH III, McClung MR, Ensrud KE, et al. Alendronate treatment of the postmenopausal osteoporotic woman: effect of multiple dosages on bone mass and bone remodeling. Am J Med. 1995;99(2):144-152.
98. Herd RJ, Balena R, Blake GM, Ryan PJ, Fogelman I. The prevention of early postmenopausal bone loss by cyclical etidronate therapy: a 2-year, double-blind, placebo-controlled study. Am J Med. 1997;103(2):92-99.
99. Välimäki MJ, Farrerons-Minguella J, Halse J, et al. Effects of risedronate 5 mg/d on bone mineral density and bone turnover markers in late-postmenopausal women with osteopenia: a multinational, 24-month, randomized, double-blind, placebo-controlled, parallel-group, phase III trial. Clin Ther. 2007;29(9):1937-1949.
100. Reid IR, Brown JP, Burckhardt P, et al. Intravenous zoledronic acid in postmenopausal women with low bone mineral density. N Engl J Med. 2002;346(9):653-661.
101. Boonen S, Reginster JY, Kaufman JM, et al. Fracture risk and zoledronic acid therapy in men with osteoporosis. N Engl J Med. 2012;367(18):1714-1723.
102. Pols HA, Felsenberg D, Hanley DA, et al; Fosamax International Trial Study Group. Multinational, placebo-controlled, randomized trial of the effects of alendronate on bone density and fracture risk in postmenopausal women with low bone mass: results of the FOSIT study. Osteoporos Int. 1999;9(5):461-468.
103. McClung MR, Geusens P, Miller PD, et al; Hip Intervention Program Study Group. Effect of risedronate on the risk of hip fracture in elderly women. N Engl J Med. 2001;344(5):333-340.
104. Ettinger B, Black DM, Mitlak BH, et al; Multiple Outcomes of Raloxifene Evaluation (MORE) Investigators. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: results from a 3-year randomized clinical trial. JAMA. 1999;282(7):637-645.
105. Delmas PD, Ensrud KE, Adachi JD, et al; Mulitple Outcomes of Raloxifene Evaluation Investigators. Efficacy of raloxifene on vertebral fracture risk reduction in postmenopausal women with osteoporosis: four-year results from a randomized clinical trial. J Clin Endocrinol Metab. 2002;87(8):3609-3617.
106. Gartlehner G, Patel SV, Feltner C, et al. Hormone therapy for the primary prevention of chronic conditions in postmenopausal women: evidence report and systematic review for the US Preventive Services Task Force. JAMA. 2017;318(22):2234-2249.
107. Cummings SR, San Martin J, McClung MR, et al; FREEDOM Trial. Denosumab for prevention of fractures in postmenopausal women with osteoporosis. N Engl J Med. 2009;361(8):756-765.
108. Greenspan SL, Bone HG, Ettinger MP, et al; Treatment of Osteoporosis With Parathyroid Hormone Study Group. Effect of recombinant human parathyroid hormone (1-84) on vertebral fracture and bone mineral density in postmenopausal women with osteoporosis: a randomized trial. Ann Intern Med. 2007;146(5):326-339.
109. Orwoll ES, Scheele WH, Paul S, et al. The effect of teriparatide [human parathyroid hormone (1-34)] therapy on bone density in men with osteoporosis. J Bone Miner Res. 2003;18(1):9-17.
110. Sontag A, Wan X, Krege JH. Benefits and risks of raloxifene by vertebral fracture status. Curr Med Res Opin. 2010;26(1):71-76.
111. McClung MR, Boonen S, Törring O, et al. Effect of denosumab treatment on the risk of fractures in subgroups of women with postmenopausal osteoporosis. J Bone Miner Res. 2012;27(1):211-218.
112. Boonen S, Adachi JD, Man Z, et al. Treatment with denosumab reduces the incidence of new vertebral and hip fractures in postmenopausal women at high risk. J Clin Endocrinol Metab. 2011;96(6):1727-1736.
113. Cummings SR, Schwartz AV, Black DM. Alendronate and atrial fibrillation. N Engl J Med. 2007;356(18):1895-1896.
114. Grey A, Bolland M, Wattie D, Horne A, Gamble G, Reid IR. Prolonged antiresorptive activity of zoledronate: a randomized, controlled trial. J Bone Miner Res. 2010;25(10):2251-2255.
115. McClung M, Miller P, Recknor C, Mesenbrink P, Bucci-Rechtweg C, Benhamou CL. Zoledronic acid for the prevention of bone loss in postmenopausal women with low bone mass: a randomized controlled trial. Obstet Gynecol. 2009;114(5):999-1007.
116. Sørensen HT, Christensen S, Mehnert F, et al. Use of bisphosphonates among women and risk of atrial fibrillation and flutter: population based case-control study. BMJ. 2008;336(7648):813-816.
117. Barrett-Connor E, Cauley JA, Kulkarni PM, Sashegyi A, Cox DA, Geiger MJ. Risk-benefit profile for raloxifene: 4-year data from the Multiple Outcomes of Raloxifene Evaluation (MORE) randomized trial. J Bone Miner Res. 2004;19(8):1270-1275.
118. Barrett-Connor E, Grady D, Sashegyi A, et al; MORE Investigators (Multiple Outcomes of Raloxifene Evaluation). Raloxifene and cardiovascular events in osteoporotic postmenopausal women: four-year results from the MORE (Multiple Outcomes of Raloxifene Evaluation) randomized trial. JAMA. 2002;287(7):847-857.
119. Cauley JA, Norton L, Lippman ME, et al. Continued breast cancer risk reduction in postmenopausal women treated with raloxifene: 4-year results from the MORE trial. Breast Cancer Res Treat. 2001;65(2):125-134.
120. Greenspan S, Field-Munves E, Tonino R, et al. Tolerability of once-weekly alendronate in patients with osteoporosis: a randomized, double-blind, placebo-controlled study. Mayo Clin Proc. 2002;77(10):1044-1052.
121. Hosking D, Adami S, Felsenberg D, et al. Comparison of change in bone resorption and bone mineral density with once-weekly alendronate and daily risedronate: a randomised, placebo-controlled study. Curr Med Res Opin. 2003;19(5):383-394.
122. Compston J, Cooper A, Cooper C, et al; National Osteoporosis Guideline Group (NOGG). Guidelines for the diagnosis and management of osteoporosis in postmenopausal women and men from the age of 50 years in the UK. Maturitas. 2009;62(2):105-108.
123. National Institute for Health and Care Excellence (NICE). Osteoporosis: Assessing the Risk of Fragility Fracture. London, United Kingdom: NICE; 2012. NICE Clinical Guideline 146.

Figure 1. Analytic Framework and Key Questions

This figure is an analytic framework depicting the key questions (KQs) within the context of the populations, interventions, comparisons, outcomes, time frames, and settings (PICOTS) relative to the effectiveness and harms of screening and treatment for osteoporosis. The figure illustrates the relationship between osteoporosis for asymptomatic adults age 40 years and older without known reasons for secondary osteoporosis. KQ1 concerns the relationship between screening and risk assessment and reduced fracture-related morbidity and mortality. The population is shown as being at an increased risk or not at an increased risk (KQ2a and KQ2b respectively). If the population is not at increased risk, they are shown as circling back to the beginning of the pathway (KQ2c). KQ2d examines bone measurement testing in screening. For those at an increased risk, the harms of screening are examined (KQ3). The pathway indicates that populations at an increased risk are determined to have normal bone mass or osteoporosis or low bone mass. If the population has osteoporosis or low bone mass, they continue to treatment (KQ4). A pathway from osteoporosis or low bone mass following treatment leads to reduced fractures and harms of treatment are also examined (KQ5). Finally, the outcome of reduced fracture-related morbidity and mortality is depicted.

Evidence reviews for the US Preventive Services Task Force (USPSTF) use an analytic framework to visually display the key questions that the review will address to allow the USPSTF to evaluate the effectiveness and safety of a preventive service. The questions are depicted by linkages that relate to interventions and outcomes. A dashed line is used to reflect the natural progression of disease between an intermediate outcome and a health outcome. Further details are available from the USPSTF procedure manual.7

Figure 2. Literature Search Flow Diagram

This figure is a Preferred Reporting of Systematic Reviews and Meta-Analysis (PRISMA) tree showing the flow of articles through the systematic review process. In the identification section of the PRISMA, it shows that after duplicates were removed, 4,969 records were identified through database searching. Additionally, 238 records were identified through other sources. 5,207 records were screened and 4,365 records were excluded. 842 full-text articles were assessed for eligibility and 674 full-text articles were excluded. The reason for exclusion and number of articles excluded is as follows: not original research (23), ineligible population(s) (130), ineligible study design (40), ineligible geographic setting (22), ineligible clinical setting (3), ineligible or no screening/intervention(s) (86), ineligible or no comparator(s) (80), ineligible or no outcome(s) (186), could not obtain the full text article (3), bone measurement occurred after the outcome (19), the screening tool is not commercially available (16), the screening tool is not FDA approved (2), study is not in an externally validated cohort (25), study not conducted in a very high HDI country (11), the study is superseded by new evidence (4), the study was used for hand search only (4), the article was retracted (1), there was insufficient information for abstraction (1), or the study was poor quality (18). There were 168 articles included in the systematic review. 2 articles were publications of the previous review.

FDA indicates US Food and Drug Administration; HDI, Human Development Index; KQ, key question.
a Including hand search of Nelson et al,1 2010 Evidence Report,2 Crandall et al,9 and Marques et al.8
b Because of overlap in studies across populations and results sections, only article counts are reported. Citation counts by KQ are not unique; studies may contribute to multiple KQs.
c Not included in individual study counts at the bottom level of the diagram.
d KQ1 and KQ3: 1 study (1 article).
e KQ2a: Accuracy of clinical risk assessment tools for identifying osteoporosis, 38 studies (41 articles); accuracy of bone measurement tests used to identify low bone mass and osteoporosis, 11 studies (11 articles); accuracy of fracture risk prediction instruments, 5 systematic reviews supplemented by 13 studies; accuracy of bone measurement tests used to predict fracture, 23 studies (24 articles); calibration of fracture risk prediction instruments, 14 studies (14 articles); reclassification risk, 10 studies (10 articles).
f KQ2b: 2 studies (2 articles).
g KQ4a: Alendronate, 7 studies (7 articles); zolendronic acid, 2 studies (2 articles); risedronate, 4 studies (4 articles); etidronate, 2 studies (2 articles); ibandronate, 0 studies; raloxifine, 1 study (2 articles); estrogen, 0 studies; denosumab, 4 studies (5 articles); parathyroid hormone, 2 studies (2 articles). KQ4b: 4 studies (5 articles).
h Alendronate, 16 studies (16 articles); zolendronic acid, 4 studies (4 articles); risedronate, 6 studies (6 articles); etidronate, 2 studies (2 articles); ibandronate, 7 studies (7 articles); raloxifine, 6 studies (12 articles); estrogen, 0 studies; denosumab, 4 studies (5 articles); parathyroid hormone, 2 studies (2 articles).

Table 1. Characteristics and Accuracy of Fracture Risk Prediction Models in Predicting Fracture (KQ2a)a

Risk Prediction Tool, Tool Components, Age Range, Prediction Time Sex Type of Incident Fracture Bone Test Included No. of Studies (No. of Participants) AUC (95% CI)b
FRAXc31
Age, sex, weight, height, previous fracture, parental hip fracture, current smoking, glucocorticoid steroid use, rheumatoid arthritis, secondary osteoporosis, alcohol use; hip BMDd optional
Ages 40-90 y
Prediction time 10 y
Men MOF Hip BMD 432-35 (n = 15,842) 0.67 (0.66-0.68)
I2 = 0.0%
Men MOF None 332-34 (n = 13,970) 0.62 (0.61-0.63)
I2 = 40.5%
Men Hip Hip BMD 332-34 (n = 13,970) 0.76 (0.72-0.80)
I2 = 96.7%
Men Hip None 332-34 (n = 13,970) 0.73 (0.68-0.77)
I2 = 96.7%
Women MOF Hip BMD 1232, 34, 36-45 (n = 62,054) 0.70 (0.68-0.71)
I2 = 92.1%
Women MOF None 1732, 34, 36, 37, 39-51 (n = 158,897) 0.66 (0.63-0.69)
I2 = 99.2%
Women Hip Hip BMD 1032, 34, 37, 39, 40, 42, 43, 45, 52, 53
(n = 161,984)
0.79 (0.76-0.81)
I2 = 99.1%
Women Hip None 1232, 34, 39, 40, 42, 43, 45, 48, 50-53 (n = 190,795) 0.76 (0.72-0.81)
I2 = 99.8%
Both sexes MOF Hip BMD 354-56 (n = 66,777) 0.69 (0.69-0.70)
I2 = 70.3%
Both sexes MOF None 354-56 (n = 66,777) 0.66 (0.66-0.67)
I2 = 47.1%
Both sexes Hip Hip BMD 254, 55 (n = 669754 and 39,60355) 0.80 (0.77-0.83)
0.83 (0.82-0.85)
Both sexes Hip None 254, 55 (n = 669754 and 39,60355) 0.77 (0.73-0.79)
0.79 (0.78-0.82)
Garvan nomogram/FRC57
Age, sex, weight, previous nontraumatic fracture since age 50 y, fall within past 12 mo; hip BMDe optionalf
Ages 60-96 y
Prediction time 10 yg
Men MOFh Hip BMD 158 (n = 1606) 0.70 (NR)
Men Hiph Hip BMD 258, 59 (n = 134659 and 160658) 0.79 (NR)
0.85 (NR)
Men Hip None 159 (n = 1285) 0.65 (NR)
Men Nonvertebral Hip BMD 159 (n = 1346) 0.67 (NR)
Men Nonvertebral None 159 (n = 1355) 0.61 (NR)
Women MOFh Hip BMD 336, 43, 58 (n = 6174) 0.68 (0.64-0.71)
I2 = 84.8%
Women MOF None 136 (n = 600) 0.66 (0.61-0.72)
Women Any OF Hip BMD 137 (n = 506) 0.69 (NR)
Women Any OF None 137 (n = 506) 0.65 (NR)
Women Hiph Hip BMD 437, 43, 58, 59 (n = 7449) 0.72 (0.66-0.79)
I2 = 97.3%
Women Hip None 159 (n = 1369) 0.68 (NR)
Women Nonvertebral Hip BMD 159 (n = 1646) 0.62 (NR)
Women Nonvertebral None 159 (n = 1637) 0.58 (NR)
QFracture60
Age, sex, weight, height, smoking, parental fracture or osteoporosis, previous fall, glucocorticoid steroid use, rheumatoid arthritis, alcohol use, hormone replacement therapy,i asthma, endrocrine disease, cardiovascular disease, menopausal symptoms,i malapsorptive gastrointestinal disease, liver disease, type 2 diabetes, tricyclic antidepressant use (or other antidepressant usej), ethnicity,i previous fracture,i dementia,i kidney disease,i epilepsy,i Parkinson disease,i living in a nursing home,i COPD,i cancer,i lupus,i anticonvulsant use,i type 1 diabetesi
Ages 30-85 yk
Prediction time 1-10 y
 
   2009 version of instrument Men MOFl None 260, 61
(n = 633,76460 and 1,108,21961)
0.69 (0.68-0.69)
0.74 (NR)
Men Hip None 28, 60, 61
(n = 633,7648, 60 and 1,108,21961)
0.86 (0.85-0.86)
0.86 (NR)
Women MOFl None 28, 60, 61
(n = 642,1538, 60 and 1,136,41761)
0.79 (0.79-0.79)
0.82 (NR)
Women Hip None 28, 61
(n = 642,15360 and 1,136,4178, 61)
0.89 (0.89-0.89)
0.89 (NR)
   2012 version of instrument Men MOFl None 162 (n = 778,810) 0.71 (0.70-0.72)
Men Hip None 162 (n = 778,810) 0.88 (0.87-0.88)
Women MOFl None 162 (n = 804,563) 0.79 (0.79-0.79)
Women Hip None 162 (n = 804,563) 0.89 (0.89-0.90)
WHI63
Age, weight, height, self-reported health, previous fracture after age 55 y, race/ethnicity, physical activity, smoking, parental hip fracture after age 40 y, diabetes treated with medications, glucocorticoid steroid use; hip BMDm optional
Ages 50-79 y
Prediction time 5 y
Women Hip Hip BMD 163 (n = 10,750) 0.80 (0.75-0.85)
Women Hip None 263, 64
(n = 10,75063 and 13,35364)
0.80 (0.77-0.82)
0.82 (NR)
OST65
Age, weight (score calculated as 0.2 × [weight in kg − age in y])
Ages 45-88 y
Prediction time NAn
Women MOF (3-y risk) None 246, 66
(n = 825466 and 361446)
0.56 (0.52-0.60)
0.71 (0.68-0.75)
Women MOF (10-y risk) None 147 (n = 62,492) 0.52 (0.52-0.53)
SCORE67
Age, weight, race, rheumatoid arthritis, prior nontraumatic fracture,
prior estrogen use
Ages ≥45 y
Prediction time NAn
Women MOF (3-y risk) None 146 (n = 3614) 0.70 (0.66-0.74)
Women MOF (10-y risk) None 147 (n = 62,492) 0.53 (0.53-0.54)
FRISC68
Age, weight, menopausal status, secondary osteoporosis, prior fracture, back pain, dementia, lumbar BMD
Ages 40-79 y
Prediction time 1, 3, 5, or 10 y
Women MOF Lumbar BMD 168 (n = 400) 0.73 (NR)
Women Long bone and vertebral fractureo Lumbar BMD 169 (n = 765) 0.69 (0.64-0.73)
FRISK70
Age, weight, height, prior fracture, prior falls, lumbar and hip BMD optional
Ages ≥60 y
Prediction time 5 or 10 y
Women MOF Lumbar and Hip BMD 136 (n = 600) 0.66 (0.60-0.71)
Women MOF None 136, 70 (n = 600) 0.62 (0.56-0.67)
FRC71
Age, sex, BMI, prior fracture, parental fracture, smoking, alcohol use, glucocorticosteroid use, rheumatoid arthritis, secondary osteoporosis, race/ethnicity, BMDp optional
Ages 45-75 y
Prediction time 10 yq
Men MOF Hip BMD 172 (n = 893) 0.70 (NR)
Men MOF None 172 (n = 893) 0.66 (NR)
Men Hip Hip BMD 172 (n = 893) 0.79 (NR)
Men Hip None 172 (n = 893) 0.71 (NR)
Women Hip Hip BMD 173 (n = 94,489) 0.85 (0.84-0.86)
Women Hip None 173 (n = 94,489) 0.83 (0.82-0.84)
ORAI74
Age, weight, current estrogen use
Ages ≥45 y
Prediction time NAn
Women MOF (3-y risk) None 146 (n = 3614) 0.71 (0.68-0.75)
Women Any OF (3-y risk) None 146 (n = 3614) 0.69 (0.66-0.72)
OSIRIS75
Age, weight, current hormone therapy use, prior fracture
Ages 60-80 y
Prediction time NAn
Women MOF (3-y risk) None 146 (n = 3614) 0.70 (0.66-0.74)
Women Any OF (3-y risk) None 146 (n = 3614) 0.68 (0.65-0.72)

Abbreviations: AUC, area under the curve; BMD, bone mineral density; BMI, body mass index; COPD, chronic obstructive pulmonary disease; FRAX, Fracture Risk Assessment Tool; FRC, Fracture Risk Calculator; FRISC, Fracture and Immobilization Score; FRISK, Fracture Risk Score; KQ, key question; MOF, major osteoporotic fractures (fractures of the proximal femur, distal radius, and proximal humerus, and clinical vertebral fractures); NA, not applicable; NR, not reported; OF, osteoporotic fracture; ORAI, Osteoporosis Risk Assessment Instrument; OSIRIS, Osteoporosis Index of Risk; OST, Osteoporosis Self-assessment Tool; SCORE, Simple Calculated Osteoporosis Risk Estimation Tool; WHI, Women’s Health Initiative.
a Studies summarized in this table include instruments predicting fracture risk over a specified time horizon (eg, 5 or 10 years). Additional studies predicting fracture by a certain age are summarized in the narrative.
b Updated pooled estimates are provided where possible; otherwise, range of AUC estimates from relevant studies is provided.
c FRAX has been updated several times since its initial release. Studies included in this review do not consistently report which version was used; thus, findings reflect various versions of FRAX released from the initial version through the current version. Further, although FRAX predicts 10-year fracture risk, the range of actual follow-up used by studies reporting accuracy of fracture risk prediction varied from 2 years to 10 years.
d Based on dual-energy x-ray absorptiometry (DXA) at the femoral neck with T-scores based on National Health and Nutrition Examination Survey (NHANES) reference values for women aged 20 to 29 years.
e Based on DXA, site unspecified, reference values for T-scores unspecified.
f Either BMD or body weight is used in the nomogram.
g This instrument can be used for prediction of either 5- or 10-year fracture risk.
h One study58 reported discrimination using Harrell C statistic.
i Risk factor only used in prediction of fracture for women.
j Risk factor not included in the original QFracture but present in the 2012 update to QFracture.
k Original instrument was validated for up to age 85 years; 2012 updated version included up to age 100 years.
l Two studies60, 62 did not include fractures of the proximal humerus in their definition of MOF.
m Based on DXA of the proximal femur, reference values for T-scores unspecified.
n These instruments were initially developed to predict osteoporosis, not incident fracture. Studies have evaluated their use for fracture prediction with length of follow-up over 3 years or over 10 years as indicated.
o Only 5 risk factors from the original FRISC model were used for this estimate: age, weight, prior fracture, lumbar BMD, back pain.
p Based on DXA of the total hip and hip subregions, T-scores based on NHANES reference values for men.
q Originally developed on a cohort of only women for 5-year risk prediction, with a smaller set of clinical risk factors. Subsequent validation studies included added risk factors, included 10-year risk predictions, and applied the model to a cohort of only men.

Table 2. Summary of Bone Measurement Tests Predicting Fracture (KQ2a)

Type of Incident Fracture Site of Test Sex No. of Studies
(No. of Participants)
Age Range
at Baseline, y
Summary of Accuracy (AUC)
DXA/DXA aBMD
Any osteoporotic or nonspine fracture Lumbar spine Women 378-81 (n = 33,839) 44-95 0.64-0.77 (unadjusted)
0.66 (adjusted)a
Men 182 (n = 1921) 65-≥75 0.71 (adjusted)b
Total hip Women 249, 79, 80 (n = 29,963) 46-95 0.66-0.68 (unadjusted)
Men 182 (n = 1921) 65-≥75 0.72 (adjusted)b
Femoral neck Women 1038, 39, 43-45, 68, 78-81, 83 (n = 41,294) 40-95 0.59-0.76 (unadjusted)
0.54 (unadjusted by baseline T-score −1)
0.57 (unadjusted by baseline T-score ≤−1 to >−2.5)
0.63 (unadjusted by baseline T-score ≤−2.5)
0.64a-0.71 (adjusted)c
Men 382-84 (n = 7972) 60-≥75 0.68 (unadjusted)
0.71c-0.72 (adjusted)b
Combined 254, 55 (n = 46,300) ≥50 0.66-0.68 (unadjusted)
Middle phalanges Women 232, 51 (n = 12,830) 40-90 0.71 (unadjusted)
0.68 (adjusted)
Men 132 (n = 5206) 40-90 0.64 (unadjusted)
Vertebral spine fracture Thoracolumbar vertebra, spine Women 369, 79, 80, 85 (n = 30,837) 50-95 0.61-0.69 (unadjusted)
Total hip Women 279, 80, 83 (n = 29,861) 50-95 0.71 (unadjusted)
0.77 (adjusted)c
Femoral neck Women 279, 80, 83 (n = 29,861) 50-95 0.71 (unadjusted)
0.70 (adjusted)c
Men 183 (n = 445) ≥60 0.75 (adjusted)c
Hip fracture Thoracolumbar vertebra, spine Women 179, 80 (n = 29,407) 50-95 0.65 (unadjusted)
Total hip Women 179, 80 (n = 29,407) 50-95 0.81 (unadjusted)
Men 183 (n = 445) ≥60 0.77 (adjusted)c
Femoral neck Women 739, 43, 45, 53, 79, 80, 83 (n = 38,322) 40-95 0.64-0.86 (unadjusted)
0.75 (adjusted)d
Men 184 (n = 5606) ≥65 0.85 (unadjusted)
Combined 254, 55 (n = 46,300) ≥50 0.76-0.80 (unadjusted)
Middle phalanges Women 232 (n = 12,830) 40-90 0.83 (unadjusted)
Men 132 (n = 5206) 40-90 0.64 (unadjusted)
DXA TBS
Any osteoporotic fracture Spine Women 179, 80 (n = 29,407) 50-95 0.63 (unadjusted)
Vertebral, spine fracture Thoracolumbar vertebra, spine Women 279, 80, 85 (n = 30,072) 53-61; 50-95 0.66-0.68 (unadjusted)
Hip fracture Spine Women 179, 80 (n = 29,407) 50-95 0.68 (unadjusted)
DXA aBMD and TBS
Any osteoporotic fracture Spine Women 179, 80 (n = 29,407) 50-95 0.66 (unadjusted)
DXA BMD total hip + TBS spine Women 179, 80 (n = 29,407) 50-95 0.69 (unadjusted)
DXA BMD femoral neck + TBS spine Women 179, 80 (n = 29,407) 50-95 0.69 (unadjusted)
Vertebral, spine fracture Thoracolumbar vertebra, spine Women 279, 80, 85 (n = 30,072) 53-61; 50-95 0.70-0.71 (unadjusted)
0.72d-0.73 (adjusted)e
DXA BMD total hip + TBS spine Women 179, 80 (n = 29,407) 50-95 0.73 (unadjusted)
DXA BMD femoral neck + TBS spine Women 179, 80 (n = 29,407) 50-95 0.73 (unadjusted)
Hip fracture Spine Women 179, 80 (n = 29,407) 50-95 0.69 (unadjusted)
DXA BMD total hip + TBS spine Women 179, 80 (n = 29,407) 50-95 0.82 (unadjusted)
DXA BMD femoral neck + TBS spine Women 179, 80 (n = 29,407) 50-95 0.81 (unadjusted)
QUS (BUA)f
Any osteoporotic fracture Heel Women 178 (n = 775) 44-56 0.72 (adjusted)a
Men 282, 84 (n = 1921; n = 5606) 65-≥75; ≥65 0.68 (unadjusted)
0.65 (adjusted)b
Hip fracture Heel Men 184 (n = 5606) ≥65 0.84 (unadjusted)
QUS (SOS)f
Any osteoporotic fracture Heel Men 182 (n = 1921) 65-≥75 0.64 (adjusted)b
QUS (QUI)f
Any osteoporotic or nonspine fracture Heel Men 182 (n = 1921) 65-≥75 0.66 (adjusted)b
QUS (BUA) and DXA BMDf
Any osteoporotic or nonspine fracture QUS (heel), DXA (femoral neck) Women 183 (n = 454) ≥60 0.73 (adjusted)c
Men 283, 84 (n = 5606) ≥65; ≥60 0.69 (unadjusted)
0.71 (adjusted)c
Vertebral fracture QUS (heel), DXA (femoral neck) Women 183 (n = 454) ≥60 0.72 (adjusted)c
Men 183 (n = 454) ≥60 0.75 (adjusted)c
Hip fracture QUS (heel), DXA (femoral neck) Women 183 (n = 454) ≥60 0.81 (adjusted)c
Men 283, 84 (n = 5606; n = 445) ≥65; ≥60 0.85 (unadjusted)
0.78 (adjusted)c

Abbreviations: aBMD, areal bone mineral density; AUC, area under the curve; BUA, broadband ultrasound attenuation; DXA, dual-energy x-ray absorptiometry; QUI, quantitative ultrasound index (combines BUA and SOS); QUS, quantitative ultrasound; SOS, speed of sound; TBS, trabecular bone score.
a Adjusted for age, height, weight, menopausal status, and neck bone mineral density (QUS only).
b Adjusted for age and fracture history.
c Adjusted for age, falls, and fracture history.
d Adjusted for age.
e Adjusted for age and prevalent vertebral deformity.
f Quantitative ultrasound measured at the calcaneus in all studies.

Table 3. Summary of Evidence (KQ1, KQ2, KQ3)

Population No. of Studies (No. of Observations) Summary of Findings by Outcome Consistency and Precision Reporting Bias Body of Evidence Limitations EPC Assessment
of Strength of Evidence: For Outcome
Applicability Overall
Quality
of Studies
KQ1: Effectiveness of Screeninga
Women 1 (12,483) Osteoporotic fractures: 12.9% vs 13.6%; HR, 0.94 (95% CI, 0.85-1.03)
All clinical fractures: 15.3% vs 16.0%; HR, 0.94 (95% CI, 0.86-1.03)
Mortality: 8.8% vs 8.4%; HR, 1.05 (95% CI, 0.93-1.19)
Hip fractures: 2.6% vs 3.5%; HR, 0.72 (95% CI, 0.59-0.89)
Consistency unknown (single study); precise for hip fractures, imprecise for other outcomes No evidence of reporting bias Potential for contamination Low for benefit for hip fractures, insufficient for other outcomes Unclear whether findings apply to men or younger women Fair
KQ2a: Accuracy of Clinical Risk Assessment Instruments for Identifying Osteoporosis
Womenb 27 (55,898) AUCs range from 0.32 to 0.87 for all included instruments (pooled AUCs range from 0.65-0.76) Inconsistent; precise No evidence of reporting bias Heterogeneity in included studies Moderate Unclear whether findings apply to subgroups defined by age or race Fair
Men 11 (14,052) AUCs range from 0.62 to 0.89 for all included instruments (pooled AUCs range from 0.76-0.80) Inconsistent; precise No evidence of reporting bias Heterogeneity in included studies Low Unclear whether findings apply to subgroups defined by age or race Fair
KQ2a: Accuracy of Bone Measurement Tests for Identifying Osteoporosis
Women 7 (1969) BMD tests for identifying osteoporosis: AUCs range from 0.67 to 0.94 for all included bone measurement testsc (pooled AUC for calcaneal QUS, 0.77 [95% CI, 0.72-0.82]) Inconsistent; precise No evidence of reporting bias Heterogeneity in included studies Moderate Unclear whether findings apply to subgroups defined by age or race Fair
Men 3 (5142) BMD tests for identifying osteoporosis: AUC for calcaneal QUS, 0.80 (95% CI, 0.66-0.94) Inconsistent; precise No evidence of reporting bias Ultrasound imaging only; heterogeneity in size, estimate of effect, and applicability of included studies Low Unclear whether findings apply to subgroups defined by age or race Fair
KQ2a: Accuracy of Bone Measurement Tests for Fracture Prediction
Women Varies by type of imaging test and site of test Centrally measured DXA BMD, TBS, or a combination of both predicting fractures from 14 studies, n = 46,036: AUCs range from 0.59 to 0.86
For other bone measurement tests or combination of tests (2 studies, n = 712):
QUS alone predicting fractures: AUCs range from 0.66 to 0.72
QUS + DXA BMD predicting fractures: AUCs range from 0.72 to 0.81
Inconsistent; precise No evidence of reporting bias Inconsistent control for baseline variables Moderate Unclear whether findings apply to nonwhite subgroups Fair
Men 3 (7972) Centrally measured DXA BMD or TBS predicting fractures: AUCs range from 0.68 to 0.85
QUS alone predicting fractures: AUCs range from 0.64 to 0.84
QUS + DXA BMD predicting fractures: AUCs range from 0.69 to 0.85
Inconsistent; precise No evidence of reporting bias Inconsistent control for baseline variables Moderate Unclear whether findings apply to nonwhite, non–East Asian subgroups Fair to good
Women and men combined 2 (46,300) Centrally measured DXA BMD predicting fractures: AUCs range from 0.66 to 0.80 Inconsistent; precise No evidence of reporting bias None identified Moderate Findings limited to Canadian samples, unclear whether results are applicable to other populations Fair to good
KQ2a: Accuracy of Fracture Risk Prediction Instruments
Women Varies by instrument AUCs for fracture risk prediction instruments range from 0.53 to 0.89 and vary by instrument, type of fracture, and whether BMD is used in the prediction. Within this range, prediction of hip fractures and predictions that use BMD report higher AUCs
Pooled AUC for FRAX prediction of hip fractures without BMD, 0.76 (95% CI, 0.72-0.81); I2 = 99.8%; 12 studies, 190,795 women
Pooled AUC for FRAX prediction of hip fractures with BMD, 0.79 (95% CI, 0.76-0.81); I2 = 99.1%; 10 studies, 161,984 women
Pooled AUC for FRAX prediction of MOF without BMD, 0.66 (95% CI, 0.63-0.69); I2 = 99.2%; 17 studies, 158,897 women
Pooled AUC for FRAX prediction of MOF with BMD, 0.70 (95% CI, 0.68-0.71); I2 = 92.1%; 12 studies, 62,054 women
Inconsistent; precise No evidence of reporting bias Some studies did not follow participants for the entire duration of the prediction interval (ie, 10 y).
Heterogenous study populations, that may have included participants with osteoporosis, with prior fracture, or receiving treatment
Moderate Other than FRAX, most instruments have not been calibrated for use in US populations. Unclear whether findings apply to nonwhite subgroups Fair
Men Varies by instrument   Inconsistent; precise No evidence of reporting bias Some studies did not follow participants for the entire duration of the prediction interval (ie, 10 y).
Heterogenous study populations, that may have included participants with osteoporosis, with prior fracture, or receiving treatment
Moderate Other than FRAX, most instruments have not been calibrated for use in US populations. Unclear whether findings apply to nonwhite subgroups Fair
KQ2b: Screening Intervals for Osteoporosis and Low Bone Density
Women and men (1 study each) 2 (4926) Similar accuracy of predicting fracture with repeat BMD when compared with baseline BMD alone Consistent; precise No evidence of reporting bias Limited number of studies; follow-up period inadequate for women, small N for men, inconsistent screening intervals Insufficient Unclear whether all findings apply to subgroups by age, sex, or race Fair
KQ3: Harms of Screening
Women 1 (12,483) Anxiety: P = 0.15 for repeated measures (variance NR)
Quality of life: P >0.10 for repeated measures for Euroquol
5-Dimension and Short Form Health Survey 12 (variance NR)
Consistency and precision unknown (single study) No evidence of reporting bias Potential for contamination and reporting bias Insufficient Unclear whether findings apply to men or younger women Fair

Abbreviations: AUC, area under the curve; BMD, bone mineral density; DXA, dual-energy x-ray absorptiometry; EPC, Evidence-based Practice Center; FRAX, Fracture Risk Assessment Tool; HR, hazard ratio; KQ, key question; MOF, major osteoporotic fractures; QUS, quantitative ultrasound; TBS, trabecular bone score.
a The comparator in the 1 KQ1 study was no screening.
b One study (not included in strength of evidence ratings; n = 282) evaluated the accuracy of FRAX and Osteoporosis Self-assessment Tool in a mixed population with 45.1% women. AUCs ranged from 0.68 to 0.76 and are consistent with findings in men and women separately.
c Included studies evaluated calcaneal QUS, peripheral DXA, digital x-ray radiogrammetry, and radiographic absorptiometry.

Table 4. Summary of Evidence (KQ4 and KQ5)a

Population No. of Studies (No. of Observations) Summary of Findings by Outcome Consistency
and Precision
Reporting Bias Body of Evidence Limitations EPC Assessment of Strength of Evidence: For Outcome Applicability Overall
Quality
of Studies
KQ4a: Effectiveness of Pharmacotherapy
Women and men Varies by outcome Bisphosphonates (women):
Vertebral fractures: 2.1% vs 3.8%; RR, 0.57 (95% CI, 0.41-0.78); 5 trials, n = 5433
Nonvertebral fractures: 8.9% vs 10.6%; RR, 0.84 (95% CI, 0.76-0.92); 8 trials, n = 16,438
Hip fractures: 0.7% vs 0.96%; RR, 0.70 (95% CI, 0.44-1.11); 3 trials, n = 8988
Zoledronic acid (men):
Morphometric vertebral fractures: 1.5% vs 4.6%; RR, 0.33 (95% CI, 0.16-0.70); 1 trial, n = 1199
Nonvertebral fractures: 0.9% vs 1.3%; RR, 0.65 (95% CI, 0.21-1.97); 1 trial, n = 1199
Clinical fractures (vertebral or nonvertebral): 1.0% vs 1.8%; RR, 0.57 (95% CI, 0.21-1.52); 1 trial, n = 1199
Vertebral fractures: consistent and precise
Nonvertebral fractures: consistent and precise
Hip outcomes: consistent and imprecise
No evidence of reporting bias Evidence dominated by 1 big study for each drug Moderate for benefit for bisphosphonates for vertebral and nonvertebral fractures, low for hip fractures Unclear whether all findings apply to subgroups by age, sex, or race Fair
Women 1 (7705) Raloxifene:
Vertebral fractures: 7.5% vs 12.5%; RR, 0.64 (95% CI, 0.53-0.76)
Nonvertebral fractures: 12.1% vs 12.9%; RR, 0.93 (95% CI, 0.81-1.06)b
Consistency unknown (single trial); precise for vertebral fractures, imprecise for nonvertebral fractures No evidence of reporting bias Single large trial Moderate for benefit for vertebral fractures, low for nonvertebral fractures Unclear whether findings apply to other subpopulations defined by age, sex, or race Good
Women 2 (varies by outcome) Denosumab:
Vertebral fractures: 2.3% vs 7.2% in trial with events; RR, 0.32 (95% CI, 0.26-0.41); 2 trials, 8020 participants
Nonvertebral fractures: 6.1% vs 7.5%; RR, 0.80 (95% CI, 0.67-0.95); 1 trial, 7808 participants
Hip fractures: 0.7% vs 1.1%; RR, 0.60 (95% CI, 0.37-0.97); 1 trial, 7808 participants
Consistency unknown (single trial for most outcomes); precise No evidence of reporting bias Single large trial for most outcomes Low for benefit for vertebral, nonvertebral, and hip fractures Unclear whether findings apply to subgroups by age, sex, or race Fair
Women and men 2 (2830) Parathyroid hormone (women; 1 trial, n = 2532):
Vertebral fractures: 0.7% vs 2.1%; RR, 0.32 (95% CI, 0.14-0.75)
Nonvertebral fractures: 5.6% vs 5.8%; RR, 0.97 (95% CI, 0.71-1.33)
Parathyroid hormone (men; 1 trial, n = 298):
Nonvertebral fractures: 1.3% vs 2.0%; RR, 0.65 (95% CI, 0.11-3.83)
Women: consistency unknown (single trial); precise for vertebral fractures
Men: consistency unknown (single trial); imprecise for vertebral fractures
No evidence of reporting bias Single trial each for men and women; small trial for men Low for benefit for vertebral fractures for women, insufficient for men for vertebral fractures Unclear whether findings apply to subgroups by age, sex, or race Fair
KQ4b: Effectiveness of Pharmacotherapy by Subgroup
Women 4 (varies by drug) Similar results by subgroup for:
Alendronate for baseline BMD (1 trial, n = 3737)
Risedronate for age (1 trial, n = 2648)
Raloxifene (prior fractures, 1 trial, n = 5114)
Denosumab for age, baseline BMD, and a combination of risk factors (1 trial, n = 7868)
Consistency unknown (single trial); precise No evidence of reporting bias Single trial for each drug Low for no differences No information on variations by menopausal status Fair
KQ5: Harms of Pharmacotherapy
Women and men Varies by outcome Bisphosphonatesc:
Discontinuations: 11.5% vs 11.8%; RR, 0.99 (95% CI, 0.91-1.07); 20 trials, n = 17,369d
Serious adverse events: 21.0% vs 23.4%; RR, 0.98 (95% CI, 0.92-1.04); 17 trials, n = 11,745d
Upper GI events: 35.3% vs 35.6%; RR, 1.01 (95% CI, 0.98-1.05); 13 trials, n = 20,485d
No statistically significant differences for cardiovascular outcomes
No reports of osteonecrosis of the jaw, atypical femur fracture, or kidney failure
3 trials that combined results for men and women or included men only had results consistent with trials of women only for discontinuations, serious adverse events, and upper GI tract events
Consistent and precise for discontinuations, serious adverse events, and upper GI tract events; inconsistent and imprecise for cardiovascular outcomes, osteonecrosis, and atypical femur fractures No evidence of reporting bias Evidence dominated by 1 big study for each drug Moderate for no harms bisphosphonates for discontinuation, serious adverse events, and upper gastrointestinal events; insufficient for cardiovascular events, osteonecrosis, and atypical femur fractures Unclear whether findings for all drugs apply to subpopulations defined by age, sex, or race Fair
Women Varies by outcome Raloxifene:
Discontinuations: 12.6% vs 11.2%; RR, 1.12 (0.98-1.28); 6 trials, n = 6438
Deep vein thrombosis: 0.7% vs 0.3%; RR, 2.14 (95% CI, 0.99-4.66); 3 trials, n = 5839
Hot flashes: 11.2% vs 7.6%; RR, 1.42 (95% CI, 1.22-1.66); 5 trials, n = 6249
Leg cramps: 8.0% vs 4.8%; RR, 1.41 (95% CI, 0.92-2.14); 3 trials, n = 6000
Inconsistent and precise for deep vein thrombosis, leg cramps, and hot flashes; consistent and imprecise for discontinuation No evidence of reporting bias Single large trial dominating results Low for harm for deep vein thrombosis and hot flashes; low for no harm discontinuation and leg cramps Unclear whether findings apply to other subpopulations defined by age, sex, or race Good
Women Varies by outcome Denosumab:
Discontinuations: 2.4% vs 2.1%; RR, 1.14 (95% CI, 0.85-1.52)
Serious adverse events: 23.8% vs 23.9%; RR, 1.12 (95% CI, 0.88-1.44)
Serious infections: 4.0% vs. 3.3%; RR, 1.89 (95% CI, 0.61-5.91)
Inconsistent and imprecise for discontinuations; consistent and imprecise for serious adverse events and serious infections No evidence of reporting bias Single large trial dominating results Insufficient for discontinuation; low for no harm for serious adverse events and serious infections Unclear whether findings apply to subgroups by age, sex, or race Fair
Women and men 2 (2830) Parathyroid hormone (women; 1 trial, n = 2532):
Discontinuations: 30.2% vs 24.6%; RR, 1.23 (95% CI, 1.08-1.40)
Parathyroid hormone (men; 1 trial, n = 298 for 20 μg [FDA-approved dose] vs placebo):
Discontinuations: 9.2% vs 4.8%; RR, 1.94 (95% CI, 0.81-4.69)
Cancers: 2.0% vs 2.0%; RR, 0.97 (95% CI, 0.2-4.74)
Consistency unknown (single trial); precise for women
Consistency unknown (single trial); imprecise for men
No evidence of reporting bias Single trial each for men and women; small trial in men Low for harm for women for discontination
Insufficient for men for discontinuations and serious adverse events
Unclear whether findings apply to subgroups by age or race Fair

Abbreviations: BMD, bone mineral density; EPC, Evidence-based Practice Center; FDA, US Food and Drug Administration; GI, gastrointestinal; KQ, key question; RR, relative risk.
a The comparator for all KQ4 and KQ5 studies is placebo or no treatment.
Data available only for combined group of participants receiving dosages of 60 mg/d or 120 mg/d. Recommended dosage is 60 mg/d.
c Pooled estimates include men, women, and combined estimates (1 study did not provide adverse events by sex).120
d Sum of N in trials in meta-analysis, after accounting for the duplication in patients in the placebo group for a 3-group study.121

Current as of: June 2018

Internet Citation: Evidence Summary: Osteoporosis to Prevent Fractures: Screening. U.S. Preventive Services Task Force. July 2018.
https://www.uspreventiveservicestaskforce.org/Page/Document/evidence-summary1/osteoporosis-screening1

USPSTF Program Office   5600 Fishers Lane, Mail Stop 06E53A, Rockville, MD 20857