Osteoporosis Risk Screening Tools Fall Short in Younger Postmenopausal Women, New Research Reveals

Three osteoporosis risk assessment tools recommended by the US Preventive Services Task Force (USPSTF) demonstrated poor to fair ability to identify osteoporosis in postmenopausal women aged 50 to 64 years, according to a study published online March 18, in JAMA Network Open.

Carolyn J Crandall, MD, MA

The tools, the Osteoporosis Self-Assessment Tool (OST), the Osteoporosis Risk Assessment Instrument (OSIRIS), and the Osteoporosis Index of Risk (OSIRIS), all exhibited an area under the receiver operating curve (AUC) below 0.7, the threshold for an assessment tool to be considered an “acceptable” discriminator, reported study authors. These findings suggest that current guidelines recommending these tools for younger postmenopausal women may need reassessment.

The cross-sectional study, led by Carolyn J Crandall, MD, MA, a professor in the department of medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, analyzed data from 6,067 postmenopausal women participating in the Women’s Health Initiative Bone Density Substudy. The researchers measured bone mineral density (BMD) using dual-energy x-ray absorptiometry (DXA), and defined osteoporosis as a T-score of –2.5 or lower at the femoral neck, total hip, or lumbar spine.

Crandall et al found that 14.1% of the participants had osteoporosis at any 1 of the 3 sites, and 4.9% had osteoporosis at the femoral neck. All 3 risk prediction tools showed higher AUCs for predicting osteoporosis at the femoral neck compared to any of the 3 sites. The OSIRIS had the highest AUC for predicting femoral neck osteoporosis (AUC = 0.83), followed by the OST (AUC = 0.818) and the OSIRIS (AUC = 0.805).

Sensitivity, Specificity Fall

However, when guideline-recommended cutoff scores for predicting osteoporosis for each tool were applied, all demonstrated low sensitivity and moderate specificity. Specifically, the OSIRIS had a sensitivity of 37.8%, a specificity of 88.8%, and an AUC of 0.633. The OSIRIS had a sensitivity of 53.3%, a specificity of 79.4%, and an AUC of 0.663. The OST had a sensitivity of 62.4%, a specificity of 68.5%, and an AUC of 0.654. All 3 tools were categorized as having poor to fair discrimination, with AUCs between 0.5 and 0.7.

Prior research on ORAI predominantly assessed fracture risk in older women, typically aged 60 years or older, across various national and international cohorts, the authors said. Studies evaluating the OSIRIS in international populations reported AUC values ranging from 0.71 to 0.73. However, until now, data have been limited on the performance of both tools in identifying osteoporosis BMD specifically in postmenopausal women aged 50 to 64 years. This is the target population, the authors emphasized, for whom the USPSTF recommends formal risk assessment tools. This was the impetus for their research.

The final analytic sample for the study had a mean age of 57.7 years. The majority identified as White (80.2%); 16.7% self-identified as Black, and 8.8% as Hispanic. More than one-third of the participants (35.3%) had a BMI of 30 or higher, according to the results. Data were collected from October 1993 to December 1998 and analyzed between September 23, 2023, and April 10, 2024.

The study authors concluded that these findings “cast some doubt” on the USPSTF’s recommendations for using these risk assessment tools in clinical practice for younger postmenopausal women. The study exposes a gap in identifying these younger women with methods that use common clinical risk factors, Crandall et al wrote. They suggest that future studies should explore new tools, potentially leveraging machine learning, to improve osteoporosis screening in this population.

References

Identifying Younger Postmenopausal Women With Osteoporosis Using USPSTF-Recommended Osteoporosis Risk Assessment Tools Henry W. Zheng, BS; Alex A. T. Bui, PhD; Kristine E. Ensrud,

JAMA Network Open. 2025;8(3):e250626. doi:10.1001/jamanetworkopen.2025.0626