Fragility Score: predicting fractures 5 years out
A REMS-based indicator for incident fragility fractures
Pisani, Conversano, Muratore, Adami, Brandi, Caffarelli, E. Casciaro, Di Paola, Franchini, Gatti, Gonnelli, Guglielmi, Lombardi, Natale, Testini & S. Casciaro. Aging Clinical and Experimental Research (2023). doi.org/10.1007/s40520-023-02358-2
1,989
Patients enrolled, followed for up to 5 years
AUC 0.78–0.81
Fragility Score's fracture-prediction accuracy
9x higher risk
Fracture risk identified above the FS cut-off in women
Best of 3
FS outpredicted both DXA and REMS T-scores
Why this study matters for your bone health
Bone density alone has never told the whole story. DXA's standard T-score threshold — the −2.5 cut-off used to diagnose osteoporosis — misses a large share of people who go on to fracture with less severe bone loss, and it flags some people who never fracture at all. The gap between what a density scan measures and what actually predicts a broken bone is the problem this study set out to close.
The Fragility Score (FS) is a second measurement REMS produces during the same scan, alongside the density T-score. Instead of measuring how much bone is there, FS analyzes the raw ultrasound signal for the textural signature of fragile bone — comparing a patient's scan against a database of spectral patterns from people who have and have not previously fractured. It is, in effect, a direct read on bone quality, not just bone quantity.
Researchers tracked 1,989 patients, both women and men, for up to 5 years, and tested whether FS could identify who would go on to suffer a fragility fracture — comparing its accuracy head-to-head against both the DXA T-score and the REMS T-score at the same anatomical sites.
What it means in practice: a REMS scan doesn't just tell your physician how dense your bone is — it can flag fragility that density alone doesn't capture, in a single non-ionizing scan.
The study's authors also note that FS's precision (how consistent repeated measurements are) matched or beat published figures for DXA — meaning FS is stable and repeatable enough to track a patient's bone quality over time, not just as a single snapshot.
What the study found
Four findings, in plain language.
- 1
FS beat both T-score methods at every comparison
For every group tested — women and men, spine and hip — FS's prediction accuracy (AUC 0.780–0.811) was significantly higher than both the REMS T-score and the DXA T-score, whose accuracy ranged from 0.472 to 0.709 across the same comparisons.
- 2
Women above the FS cut-off had 9 times the fracture risk
At the lumbar spine, women scoring above the FS threshold of 37.2 had roughly a 9-fold higher risk of fragility fracture — compared to a 2.5-fold higher risk identified by the DXA T-score and 3.6-fold by the REMS T-score in the same women.
- 3
The advantage held for men, and for hip fracture specifically
Men above the FS cut-off had about a 9.5-fold higher fracture risk at the spine and 8.3-fold at the hip. At the femoral neck, women above the FS threshold had 6 times the hip-fracture risk — in every case, a stronger signal than either T-score method produced.
- 4
The result held even after accounting for age and weight
Age and BMI can make risk look higher than it is. After statistically adjusting for both, FS still predicted fracture risk more accurately than either T-score method at every site and in both sexes — meaning FS is measuring something about bone quality itself, not simply standing in for age.
Study details
- Full title
- Fragility Score: a REMS-based indicator for the prediction of incident fragility fractures at 5 years
- Authors
- Paola Pisani, Francesco Conversano, Maurizio Muratore, Giovanni Adami, Maria Luisa Brandi, Carla Caffarelli, Ernesto Casciaro, Marco Di Paola, Roberto Franchini, Davide Gatti, Stefano Gonnelli, Giuseppe Guglielmi, Fiorella Anna Lombardi, Alessandra Natale, Valentina Testini, Sergio Casciaro
- Journal
- Aging Clinical and Experimental Research (2023)
- Published
- 8 February 2023
- DOI
- 10.1007/s40520-023-02358-2
- Population
- 1,989 Caucasian patients (women and men), enrolled in Lecce, Italy, followed up to 5 years
- Comparison
- REMS-derived Fragility Score vs. DXA BMD T-score and REMS BMD T-score for predicting incident fragility fracture
- Key result
- FS achieved AUC 0.780–0.811 across sexes and sites, versus 0.472–0.709 for either T-score method; advantage held after age/BMI adjustment
A reading that goes beyond bone density
A REMS bone density scan takes about 15 minutes, uses no radiation, and includes the Fragility Score — a measurement this study found to be the strongest available predictor of who will fracture over the next 5 years. Serving patients across New York and New Jersey.
- No radiation at any point during the scan
- Includes the Fragility Score alongside your density T-score
- Safe to repeat for monitoring as often as your physician recommends
More published research on REMS technology is collected on the research library page.