REMS Fragility Score (Spine)

Bone density captures only part of fracture risk. DXA is the standard X-ray bone-density scan. The lumbar spine Fragility Score from a REMS scan adds an ultrasound-based estimate of skeletal fragility. Higher scores have been associated with subsequent fractures in prospective studies, including fractures outside the spine. It is an emerging measure with outcome data, but its role in choosing treatment and monitoring your response is still being established.

How the Score Is Made

REMS stands for Radiofrequency Echographic Multi-Spectrometry. It analyzes the frequency patterns in raw ultrasound echoes from the lumbar vertebrae, usually L1 through L4. The software compares portions of these signals with reference patterns from people who have and haven't sustained fragility fractures. Those reference groups are matched by age, sex, body mass index, and skeletal site.

The score runs from 0 to 100. A higher value means more of the analyzed signal resembles the fracture-associated reference pattern. It is an indirect estimate of fragility and doesn't directly image the tiny connections inside bone. The 0 to 100 number is not itself a probability. The REMS software can convert it into an estimated 5-year fracture risk using models built from the manufacturer's datasets, so your report may show a risk percentage or a risk class. Those models were developed mainly in White Italian cohorts. The same scan also produces a bone-density estimate and T-score, which are separate outputs.

What the Fracture Studies Show

Prospective studies have linked the spine score with later fragility fractures. The comparison is usually with a density T-score alone, so better performance doesn't establish that it improves on a full assessment combining density, age, fracture history, and other risk factors.

Researchers use AUC to describe how well a test separates people with and without a later fracture. A value of 0.5 represents chance discrimination and 1.0 represents perfect discrimination. Among the 1,289 women, the score reached 0.811 before any adjustment, compared with 0.709 for the REMS spine T-score and 0.678 for the DXA spine T-score. Accounting for age alone pulled the score down to 0.713, because the women who fractured were much older. After adjusting for age and body mass index, the score still led: 0.715 versus 0.636 and 0.603 in women, and 0.758 versus 0.592 and 0.579 in men.

A high score is still not a forecast. At the cutoff that performed best in that study, about 35% of women scoring above it had a fracture during follow-up, while about 95% of women scoring below it did not. The score was better at identifying who was unlikely to fracture than at singling out who would.

The larger study's participants were White adults recruited at one Italian hospital between 2016 and 2019. A 2024 expert review that formally graded the REMS evidence rated the claim that the Fragility Score predicts fractures as only weakly supported and called for studies in non-White populations. These findings support fracture prediction in the studied populations. They don't establish that using the score to guide care prevents more fractures.

When Density and Fragility Disagree

Spine DXA measures mineral along the X-ray beam's path. Bone spurs, degenerative changes, and compressed vertebrae can make the result appear higher; aortic calcification has a minimal effect in most cases. REMS software is designed to reject signals from some of these structures. Studies comparing REMS density with DXA suggest that it may be less affected by certain artifacts, though the same 2024 expert review rated that claim as only weakly supported. That is also evidence about the density output, not direct proof of the Fragility Score's accuracy.

In one cross-sectional study, adults with radiographic axial spondyloarthritis had higher REMS Fragility Scores than controls, despite similar REMS bone-density readings. Axial spondyloarthritis is an inflammatory disease affecting the spine and nearby joints. The study's density comparison used REMS, not DXA, and the groups differed in age, body size, smoking, and fracture history. It supports further study of the score in this condition without proving that it detects fractures missed by DXA.

Type 2 diabetes raises a different issue: fracture risk can be higher even when measured bone density is preserved. That isn't necessarily an imaging artifact. A disagreement between density and fragility measures needs interpretation in the context of your medical history; it doesn't establish that one device is correct.

How Age Enters the Result

In a clinical cohort, the age, weight, and height entered into the device together statistically explained more than 80% of the variation in lumbar spine Fragility Scores, in both women and men. The authors concluded that REMS outputs reflect an assessment strongly based on demographic details rather than a direct ultrasound measurement of the skeleton. They advised caution with outlier results, with entering REMS values into FRAX, and with using the score to monitor change over time.

A population-level finding like this doesn't mean 80% of any one person's score is mechanically calculated from those entries. It does mean the score carries less independent information about your bones than its 0 to 100 scale suggests, which fits the drop in performance after age adjustment in the larger fracture study.

The same researchers changed the entered age and weight during repeat scans in five volunteers. That experiment measured hip REMS bone density, a different output. It demonstrated sensitivity of that density estimate to demographic entries, but did not directly measure how changing those entries alters the spine Fragility Score. Accurate entries matter, and changes in this score should not automatically be attributed to changes in bone structure.

When Results Can Be Misleading

• Data entry: age, sex, height, and weight strongly shape the result and determine reference comparisons. Incorrect entries can undermine interpretation.
• Acquisition quality: probe position, scan depth, and focus affect the ultrasound signals. Bowel gas can obscure the lumbar vertebrae.
• Different measurements: spine and hip Fragility Scores, REMS density, DXA T-scores, and trabecular bone score describe different measurements. Their numbers cannot be substituted for one another.
• Overreading a reassuring score: a lower value doesn't cancel an established fragility fracture, osteoporosis diagnosis, or medication-related fracture risk.
• Overreading a concerning score: the result alone doesn't identify a broken vertebra, establish a cause of bone weakness, or determine which treatment you need. In the main validation study, most women with a high score did not fracture during follow-up.

What a Trend Can Establish

Small studies have found good short-term repeatability of the Fragility Score; in one study of 15 patients, repeat spine scans differed by less than 1%. Repeatability shows that repeated measurements can agree; it doesn't establish how much change represents a meaningful improvement in fracture risk. Because the score leans heavily on age and body measurements, a change over time may partly reflect updated entries. Evidence for using this specific score to monitor treatment remains limited, and studies tracking REMS bone density cannot be treated as studies of the Fragility Score.

DXA, fracture history, and clinical risk assessment remain the main basis for osteoporosis decisions in widely used treatment guidelines. A prior hip or vertebral fragility fracture can justify treatment even when density isn't in the osteoporosis range. A favorable Fragility Score has not been shown to make withholding otherwise indicated treatment safe.