A DEXA scan classifies a T-score of -2.5 or lower as osteoporosis, while -1.0 or higher is considered normal. The range between them, from -1.0 to -2.5, is osteopenia. At Longevity Benchmark, we cover the tests sold by longevity platforms, and DEXA is unusual because the report comes with a diagnosis already printed on it.
DEXA stands for dual-energy X-ray absorptiometry, though many reports shorten it to DXA. Before menopause, and for men under 50, the Z-score is the meaningful measure; the T-score does not apply. And if you're comparing two scans, they belong on the same graph only if they came from the same scanner. That's why the report includes the scanner's make and model alongside the result.
The Verdict
- If you are a premenopausal woman, a man under 50, or a child, read the Z-score and ignore the T-score. The International Society for Clinical Densitometry (ISCD) states that the T-score does not apply to you.
- Otherwise read the lowest T-score among lumbar spine, total hip and femoral neck. That single number carries the diagnosis, whatever the other sites say.
- Then find the bone mineral density in g/cm² and the scanner make and model. Without both, a change since your last scan cannot be judged.
What a DEXA Scan T-Score Measures
A T-score counts standard deviations between your bone mineral density and the average density of a healthy young adult at peak bone mass. Every T-score is therefore a comparison against a reference group rather than a reading of your skeleton on its own. For the femoral neck, ISCD sets that reference as white women aged 20 to 29 from the third National Health and Nutrition Examination Survey (NHANES III). The same reference is applied to men and to every ethnic group.
The thresholds themselves come from a 1994 World Health Organization (WHO) study group, and they have not moved since.
| T-score | What the report calls it | What it means |
|---|---|---|
| -1.0 or higher | Normal | Bone density sits within one standard deviation of the young adult reference mean |
| Between -1.0 and -2.5 | Osteopenia, or low bone mass | Below the young adult mean but above the diagnostic line. Most people in this band never break a bone |
| -2.5 or lower | Osteoporosis | The World Health Organization diagnostic threshold. Treatment is usually discussed in postmenopausal women and in men aged 50 and over |
| -2.5 or lower plus a fragility fracture | Severe or established osteoporosis | A bone already broken from a low fall outweighs the number in deciding what happens next |
Osteopenia is the band that causes the most alarm and deserves the least. It covers a wide stretch of ordinary ageing, and being told you have low bone mass at 58 opens a conversation about fracture risk instead of settling one. Our page on normal versus optimal ranges works through the same problem, where a result inside a reference band still leaves a decision to make.
When the DEXA Scan Z-Score Replaces the T-Score
The Z-score compares your bone density to people of your own age, sex and ethnicity. ISCD says it is the number to read for premenopausal women, men under 50, and children. A Z-score of -2.0 or lower is reported as below the expected range for age. Anything above -2.0 is within the expected range for age. Neither phrasing is a diagnosis, and ISCD chose that wording deliberately.
| Question | T-score | Z-score |
|---|---|---|
| Compared against whom | A healthy young adult at peak bone mass | People of your own age, sex and ethnicity |
| Who it is reported for | Postmenopausal women and men aged 50 and over | Premenopausal women, men under 50, and children |
| The number that decides | -2.5 or lower is osteoporosis | -2.0 or lower is "below the expected range for age" |
| What it should trigger | A fracture risk calculation and a treatment conversation | A search for a cause, such as coeliac disease, glucocorticoid use or low sex hormones |
| Its hard limit | It cannot diagnose osteoporosis in a man under 50 by itself | It never carries an osteoporosis label on its own |
ISCD is explicit that osteoporosis cannot be diagnosed in a man under 50 from bone density alone. A low Z-score in a younger adult sends the workup toward a cause rather than a bone drug. Coeliac disease, hyperparathyroidism, long-term glucocorticoids, low testosterone or oestrogen, and a history of restrictive eating all show up this way.
The Lowest of the Measured Sites Sets the Diagnosis
ISCD asks for bone mineral density at both the posteroanterior (PA) spine and the hip in every patient, and the diagnosis follows whichever of the reported sites reads lowest. The spine is measured as L1 to L4. At the hip, the femoral neck and the total hip are both reported, and the lower of the two is used. Ward’s area and the greater trochanter are printed on many reports but are not valid for diagnosis.
The forearm is the backup site. ISCD calls for the 33% radius when the hip or spine cannot be measured or interpreted, in hyperparathyroidism, and in patients over the weight limit of the scanning table. Other forearm regions are not recommended.
Vertebrae get excluded when they are structurally abnormal, or when one vertebra differs from its neighbours by more than 1.0 T-score. Three vertebrae are used if four cannot be, and two if three cannot be. A diagnosis is never made from a single vertebra, so if only one survives the exclusions, the hip carries the result instead.
Why a Repeat DEXA Scan T-Score on a New Machine Is Not Comparable
A change between two scans only counts if it exceeds the least significant change (LSC) of the machine that produced them. Every facility has to measure its own precision error, either 15 patients scanned three times or 30 patients scanned twice, with repositioning between each scan. The LSC is derived from that error at 95% confidence. ISCD sets the minimum acceptable precision for a technologist at 1.9% for the lumbar spine, 1.8% for the total hip and 2.5% for the femoral neck. Those figures produce an LSC of 5.3%, 5.0% and 6.9% at the three sites.
Run the arithmetic on a real report. Say your total hip read 0.780 g/cm² two years ago and reads 0.762 g/cm² today, on the same scanner, with the same technologist. That is a loss of 0.018 g/cm², or 2.3%. At an LSC of 5.0%, the drop would have to clear 0.039 g/cm² before anyone could call it bone loss. Your 0.018 sits inside the measurement noise, and the T-score printed beside it moved for the same reason.
Now change the clinic. ISCD holds that if a cross-calibration assessment has not been performed, no quantitative comparison to the prior machine can be made, and a new baseline density has to be established. Cross-calibration is not a setting a technologist can flip, either. The ISCD method is to scan 30 patients once on the old system and twice on the new one within 60 days, then calculate the relationship between the two.
Hardware differences are the reason. Hologic and GE HealthCare build most of the DXA scanners in clinical use. Their systems differ in how they find the edge of a bone and in the reference data behind the spine T-score. ISCD standardised the femoral neck reference across manufacturers. The lumbar spine has no equivalent single standard, so the same spine can return a different T-score on each system with no bone having changed.
What Makes DEXA Scan Results Wrong or Misleading
Degenerative change in an older spine is the most common way a DEXA scan result misleads. Osteoarthritis, osteophytes, a calcified aorta, scoliosis and a collapsed vertebra all add mineral the scanner counts as bone. That lifts the lumbar spine number and can mask real bone loss. That is why a 74-year-old can show a normal spine and an osteoporotic hip on the same report. Past about 65 the hip is the more trustworthy site.
Four other things break a scan.
- Hardware at the measured level. Surgical rods, a laminectomy or a vertebroplasty make that vertebra unusable, so it has to be excluded and the remaining levels carry the result.
- A previous fracture at the site. A healed compression fracture reads as extra density, which inflates the number for that vertebra.
- Positioning error. A rotated hip or an abducted leg changes the measured area, which moves the density even when the bone has not changed. This is the reason repeat scans are positioned to match the first.
- Recent contrast or radioisotope. Barium, iodinated contrast for a computed tomography (CT) scan, and nuclear medicine tracers all leave material in the field. RadiologyInfo advises waiting 10 to 14 days, so book the DEXA first or push it back.
None of these are reasons to distrust the report wholesale. They are reasons to read the site-by-site breakdown rather than the headline number, and to ask which vertebrae were excluded and why. A well-written report states both.
What the DEXA Scan T-Score Does Not Tell You on Its Own
Bone density explains only part of who breaks a bone, which is why fracture risk is calculated separately. FRAX, the fracture risk assessment tool developed at the University of Sheffield, returns a 10-year probability of hip fracture and of major osteoporotic fracture. Femoral neck bone density is one input among twelve. The others are age, sex, weight, height, a previous fracture, a parent with a hip fracture, smoking, glucocorticoids, rheumatoid arthritis, secondary osteoporosis and heavy drinking.
FRAX runs without a bone density value at all, which shows how much of fracture risk sits outside the scan. The Bone Health and Osteoporosis Foundation (BHOF) uses the combination in its clinician guidance. In the osteopenia band, treatment enters the conversation at a 10-year hip fracture probability of 3% or more, or a major osteoporotic fracture probability of 20% or more. Below those figures the same T-score usually does not.
Muscle and balance carry the other half of the risk, since most fractures happen in a fall. Our page on grip strength by age covers the cheapest proxy for that. Vitamin D timing and creatine for women cover two inputs that act on bone and muscle rather than on the scan.
Trabecular Bone Score on a DEXA Report
Trabecular bone score (TBS) is a texture measurement taken from the same lumbar spine image the scanner already acquired. Bone mineral density says how much mineral is present. TBS estimates how well organised the internal structure is, which two spines of identical density can differ on. The Medimaps Group software that calculates it needs no extra appointment, radiation or scanning time, so a facility that offers it is reading data it already holds.
ISCD considers TBS appropriate in adults aged 40 and over. Its position is to use the TBS value for adjusting reported fracture risk rather than as a standalone diagnosis. One limitation follows from its source. TBS is computed from the lumbar spine image, so the same osteophytes and hardware that distort a spine T-score can distort TBS. If your spine was excluded from the diagnosis, treat the TBS on that report with the same caution.
Who Should Not Be Reading a T-Score
A healthy 35-year-old reading the bone line on a body composition scan is looking at the wrong number. Under 50 for men, and before menopause for women, the WHO T-score criteria do not apply, and ISCD asks for the Z-score instead. A T-score of -1.3 in that reader is not osteopenia, and no clinician should label it that way. The number to check is the Z-score. If it has fallen to -2.0 or lower, the next step is finding the cause rather than starting a bone drug.
What would change our answer is a fracture. If you have already broken a bone from a fall at standing height or less, or from no trauma at all, the T-score band stops being the gate. BHOF puts treatment on the table in that case regardless of where the number lands, because the fracture has already answered the question the scan was asked. Several months of glucocorticoids works the same way. Fracture risk on steroids climbs faster than bone density falls, so the T-score understates it.
A bone reading is a different kind of result from the methylation and biological age scores sold alongside it. It measures one tissue against a fixed clinical threshold. Our guide to reading a biological age report covers that distinction, and what biological age measures covers where a bone scan fits. If you have not booked a scan yet, when to get a DEXA scan and what one costs answer the timing and the price. Pull up your report and find the g/cm² column and the scanner name printed beside the DEXA scan T-score. Take both to whoever ordered the scan before you agree to a repeat elsewhere.
Frequently Asked Questions
What does a DEXA scan T-score of -1.5 mean?
A T-score of -1.5 is osteopenia, which most reports print as low bone mass. It sits between the normal cut-off of -1.0 and the osteoporosis cut-off of -2.5 set by the World Health Organization. Osteopenia is not a disease and most people who have it never fracture. It does earn you a fracture risk calculation. The Bone Health and Osteoporosis Foundation puts the treatment discussion for this band behind a FRAX result rather than behind the T-score.
What is the difference between a DEXA scan Z-score and T-score?
The T-score compares your bone mineral density to a healthy young adult at peak bone mass. The Z-score compares it to people of your own age, sex and ethnicity. A 70-year-old woman can hold a T-score of -2.6, which is osteoporosis, while her Z-score is -0.4, meaning she has about as much bone as her peers. Both statements are true at once. The T-score answers whether she is at risk of fracture, and the Z-score answers whether anything unusual is going on for her age.
What is a normal DEXA scan Z-score range?
Anything above -2.0 is reported as within the expected range for age, and -2.0 or lower is reported as below the expected range for age. That wording comes from the International Society for Clinical Densitometry, and it is deliberately not a diagnosis. A Z-score below -2.0 in a younger adult is a prompt to look for a cause. Coeliac disease, hyperparathyroidism, long-term steroid use, low testosterone or oestrogen, and restrictive eating all show up this way.
What DEXA scan score indicates osteoporosis?
A T-score of -2.5 or lower at the lumbar spine, total hip or femoral neck indicates osteoporosis in a postmenopausal woman or a man aged 50 or over. The 33% radius in the forearm counts too when the hip and spine cannot be used. The diagnosis follows the lowest of those sites, so a spine of -1.8 with a femoral neck of -2.6 is still osteoporosis. Ward’s area and the greater trochanter are not used for diagnosis, even though some reports still print them.
Can DEXA scan results be wrong?
Yes, and the lumbar spine is where it happens most. Osteoarthritis, osteophytes, a collapsed vertebra, spinal fusion hardware and a calcified aorta all add density the scanner counts as bone. That pushes the spine T-score up and hides real bone loss. Positioning is the other common source, since a rotated hip or an abducted leg changes the measured area. Recent barium, iodinated contrast or a nuclear medicine study also invalidates the scan, and RadiologyInfo advises waiting 10 to 14 days.
Why did my DEXA scan T-score get worse in one year?
Check the scanner before you believe the change. Every facility has to measure its own precision error, and the least significant change derived from it is the smallest difference that counts as real. At ISCD minimum precision the least significant change is 5.3% at the lumbar spine, 5.0% at the total hip and 6.9% at the femoral neck. A 2% drop is inside the measurement noise on all three. If the second scan happened on a different machine, there is no comparison to make at all until the two have been cross-calibrated.
What is TBS on a DEXA scan report?
TBS stands for trabecular bone score, a texture measure calculated from the same lumbar spine image the scanner already captured. It estimates how well organised the internal bone structure is, which bone mineral density alone cannot show. The Medimaps Group software that produces it needs no extra appointment or radiation. ISCD considers TBS appropriate in adults aged 40 and over and says to use it for adjusting reported fracture risk rather than as a diagnosis on its own.
Does a body composition DEXA scan give a real T-score?
It gives a bone reading from the same hardware, and for a healthy adult under 50 that reading should be read as a Z-score. Body composition scans are aimed at fat mass, lean mass and visceral fat, and they often skip the dedicated hip and spine acquisitions a diagnostic bone scan runs. Our guide to DEXA body composition results covers what those scans are built to answer.