REMS Z-Score (Spine)

A spine Z-score compares your lower-back bone density with an age-matched reference group. A low result can reflect lower bone mass built earlier in life, previous bone loss, or a condition causing ongoing loss. One reading can't distinguish those possibilities or show how quickly your bone is changing.

This score comes from radiofrequency echographic multi-spectrometry, or REMS. The method estimates bone density using ultrasound. It has promising clinical studies, including studies of future fractures, but most examined bone density, T-scores, or the separate Fragility Score. Evidence supporting decisions based specifically on the REMS lumbar Z-score is more limited.

What Your Z-Score Compares You Against

A Z-score expresses how far your estimated density lies from the reference group's average, using the spread of values in that group. It has no physical unit. A lower score means lower density relative to that comparison group.

The comparison group separates a Z-score from a T-score. A T-score compares bone density with a young-adult reference. A Z-score uses an age- and sex-matched reference. For DXA, clinical guidance favors Z-scores before menopause and in men younger than 50, while T-scores support osteoporosis diagnosis after menopause and in men 50 or older. REMS reports both scores, but the diagnostic guidance and major treatment trials are built mainly on DXA.

An older adult can have an average Z-score and still have osteoporosis by DXA T-score. Being typical for your age doesn't establish that your bones are strong enough to withstand a fall.

How REMS Produces the Number

REMS analyzes ultrasound echoes from the first four lumbar vertebrae, L1 to L4. Its software rejects signals identified as artifacts and compares the remaining patterns with reference data. Age, sex and body mass index help select the patterns used in this process. Body mass index relates weight to height.

The software derives an estimate of bone mineral density and converts it into T- and Z-scores. Personal details therefore influence the density estimate itself, as well as the comparison used to interpret it. That distinction matters when your weight changes or a detail is entered incorrectly.

The scan can also produce a Fragility Score. This separate calculation compares ultrasound patterns with patterns associated with fractures. It aims to add information about fracture susceptibility, but it doesn't directly image the microscopic structure of bone.

How Well It Agrees With DXA

A European study enrolled 4,307 women and compared REMS with DXA after reviewing scan quality. For the lumbar spine, REMS identified about 91% of DXA-defined osteoporosis cases and correctly classified about 95% of those without it. These results concerned T-score-based osteoporosis classification, not the accuracy or clinical usefulness of the Z-score.

Results have been less consistent in routine care and older populations. In a study of 148 geriatric outpatients, lumbar REMS identified about 62% of the osteoporosis cases found by DXA. A separate Australian study of 69 people concluded that the two methods weren't interchangeable. A reassuring REMS result therefore cannot reliably exclude osteoporosis when your history suggests substantial fracture risk.

The early validation studies also excluded scans with acquisition or analysis errors. Their results show what the method can achieve under the studied conditions; they don't guarantee the same performance in every clinic or population.

Why the Two Scans Can Disagree

DXA uses X-rays to estimate bone density. Arthritis-related bone spurs, vertebral compression fractures and calcium in the abdominal aorta can inflate a lumbar DXA reading. REMS uses a different signal and is designed to reject some of these artifacts.

In a study of 70 people with osteoporosis, REMS gave lower spine values than DXA, with larger discrepancies associated with severe vertebral fractures, aortic calcium and diabetes. That study supports investigating disagreement between the scans. Without an independent measure of true bone density, it doesn't establish that every lower REMS value is more accurate.

Diabetes needs a separate explanation. People with type 2 diabetes can have normal or higher DXA density while still facing greater fracture risk. This isn't necessarily a falsely high scan: bone strength depends on more than mineral density. Studies have found lower REMS measurements in women with diabetes, but that finding hasn't established a treatment rule based on their REMS Z-score.

The software's use of personal details creates another source of disagreement, and it has been studied at the spine. In the Australian study of 69 people, age and weight alone explained 88% of the variation in lumbar REMS density, compared with 33 to 35% for lumbar DXA. In a study of 209 people, demographic and body-size information explained more than 90% of the variation in REMS density at both the spine and hip, and more than 80% for the lumbar Fragility Score. A hip-replacement study found a similar pattern at the femoral neck, where age, sex and body mass index explained about 90% of the variation. Changing the personal details entered for 7 patients in that study, and for 5 healthy volunteers in the 209-person study, changed the REMS estimates.

An analysis of manufacturer-held scans, whose authors included company founders, reached a different conclusion. Age and weight explained 77% of lumbar REMS density variation in the data used to build its models but only 58% when tested on separate participants, leaving roughly 30 to 43% unexplained across test sets. The studies disagree about the size of the demographic contribution, and they leave uncertainty about how accurately REMS captures small changes within one person over time.

What the Fracture Studies Show

REMS fracture research has mainly tested T-scores and the Fragility Score. A Z-score uses an age-adjusted comparison, so its relationship with future fractures cannot be assumed to match either of those measures.

One study enrolled 1,516 women, with 1,370 completing follow-up lasting up to five years. In an age-matched analysis, the REMS spine T-score ranked a woman who fractured as higher risk than a woman who didn't about 65 to 66% of the time, compared with 61% for the DXA spine T-score. This describes moderate ability to separate the groups; it isn't the percentage of individual fractures correctly predicted.

A later five-year multicenter study recruited 4,443 women, 4,017 of whom had REMS scans. Combining REMS T-score and Fragility Score sorted women into seven risk classes with progressively higher fracture rates. It did not validate a combination of Z-score and Fragility Score, and it did not directly compare that combined model with DXA.

A concerning Fragility Score may add context to your assessment. Its usefulness depends on your fracture history, age, other risk factors and the rest of the scan, and it isn't an established stand-alone trigger for medication.

Conditions Linked to Lower Bone Density

Several studies have examined REMS in conditions that affect bone health. Most were cross-sectional: they compared people at one point in time, so they couldn't establish the rate of bone loss or whether repeated Z-scores improve care.

• Anorexia nervosa: a study of young women found lower lumbar REMS density and Z-scores than in women without the condition.
• Osteogenesis imperfecta: adults with this inherited disorder had lower REMS bone-density measurements than controls. The disorder makes bones fragile, but the study didn't establish how well REMS monitors treatment.
• Systemic lupus erythematosus: in 106 adults, REMS and DXA differed in their measurements and osteoporosis classification. REMS measurements were associated with existing vertebral fractures, but the study couldn't separate the effects of lupus, steroid exposure and other risk factors.

Studies of inflammatory spinal arthritis have also found differences in the Fragility Score even when density measurements were similar. Those findings concern a separate REMS output and don't prove that the lumbar Z-score detects every form of bone weakness.

When Results Can Be Misleading

• Incorrect personal details: errors in age, sex, height or weight can change the model's density estimate and the resulting scores.
• Scan acquisition: probe position, depth, focus and operator training affect the ultrasound signals available for analysis. Bowel gas and a large abdominal tissue depth can make lumbar imaging harder.
• Different methods or reference data: REMS and DXA values aren't interchangeable. A difference between them cannot be treated as bone gained or lost, and results from different devices or software versions need care when compared.
• Demographic changes over time: changes in weight or the model's age-related inputs can contribute to a change in the reported estimate. A small shift doesn't necessarily represent a matching change in bone tissue.
• Peritoneal dialysis: in a study of 20 people, changes in lumbar REMS and DXA density did not track each other reliably. The significant bone loss detected by DXA was at the femoral neck, so that finding should not be described as proven lumbar loss missed by REMS.

What a Trend Can Show

Longitudinal assessment needs the underlying bone-density measurements as well as the Z-score. Because the Z-score comparison changes with age, a stable score can accompany a decline in absolute density.

Short-term REMS precision has been encouraging. In one study with 42 lumbar examinations, repeated measurements by the same operator had a precision error of 0.47%. That figure describes measurement repeatability under the study's conditions. It doesn't establish a universal percentage change that proves your bones have changed.

Observational treatment studies have reported changes in lumbar REMS density, but evidence for monitoring an individual's treatment response remains less established than for DXA. Scan quality, changes in personal details and the size of measurement error all matter when interpreting a series.