BMD T-Score and Z-Score Calculator

Compute a T-score or a Z-score from a measured bone mineral density and its reference values — and get the right one, because a T-score in a premenopausal woman or a man under 50 is a diagnosis the ISCD says cannot be made that way.

Last updated 17 Sep 2026

T-score and Z-score from measured BMD

BMD + reference mean and SD → score

PopulationThis decides which score is being read, not merely how it is labelled. The ISCD position is that T-scores are preferred in postmenopausal women and men aged 50 and over, and that Z-scores, not T-scores, are preferred in premenopausal women and men under 50; it also states that osteoporosis cannot be diagnosed in men under 50 on the basis of bone density alone. Selecting the second option asks you for an age-matched reference mean and standard deviation rather than a young-adult one, and reads the answer against the Z-score categories.

Measured BMD (g/cm²)The areal bone mineral density from the DXA report, in grams per square centimetre, for one site and one region of interest. The ISCD asks for both the PA spine (L1–L4) and the hip in all patients, and for diagnosis uses the femoral neck or total hip, whichever is lowest — so enter the site you intend to diagnose from rather than the average of several. Do not enter a lumbar spine BMD against a femoral neck reference or vice versa: the reference values are site-specific and mixing them produces a number that looks plausible and means nothing.

Reference mean BMD (g/cm²)For a T-score, the mean BMD of the young-adult reference population for that site and that manufacturer. For a Z-score, the mean for people of the same age and sex. The default, 0.85 g/cm², is the Hologic female young-normal femoral neck mean recalculated from NHANES data in the 1996 study that identified the discrepancy between manufacturers; the equivalent Lunar figure in the same paper was 1.00 g/cm². Take the values from your own report or scanner documentation wherever possible — that difference alone is worth about 1.4 T-score units.

Reference SD (g/cm²)The standard deviation of the same reference population. It is the denominator of both scores, so an error here scales the whole answer: using 0.10 instead of 0.12 inflates a −2.0 to −2.4. The default of 0.11 g/cm² is the femoral neck figure from the same recalculation, and it happened to be identical for both manufacturers in that study — the discrepancy between them was in the mean, not the spread.

-2.1SDExample

Femoral neck BMD 0.62 g/cm² in a postmenopausal woman, against a young-adult reference mean of 0.85 and SD of 0.11 g/cm²

Formulae

T-score = (measured BMD − young-adult reference mean) ÷ young-adult reference SD
Z-score = (measured BMD − age-matched reference mean) ÷ age-matched reference SD

measured BMD: the areal bone mineral density in g/cm² for one site and one region of interest, from the DXA report. Not an average of several sites
the T-score’s reference: a young-adult population of the same sex. The ISCD asks for NHANES III for femoral neck and total hip T-scores, and for a uniform White, non-race-adjusted female database for women and for men of all ethnic groups
the Z-score’s reference: people of the same age and sex. The ISCD adds that Z-scores should be population-specific where adequate reference data exist
which score: T-scores in postmenopausal women and men aged 50 and over; Z-scores, not T-scores, in premenopausal women and men under 50. Osteoporosis cannot be diagnosed in men under 50 on bone density alone
which site: PA L1–L4 for the spine, and the femoral neck or total hip, whichever is lowest, for the hip. Forearm only when the hip and spine cannot be measured or interpreted
what the score is not: a fracture risk. It is one input to one. A fragility fracture permits a clinical diagnosis of osteoporosis whatever the density shows

Worked example

Femoral neck BMD 0.62 g/cm² in a postmenopausal woman, against a young-adult reference mean of 0.85 and SD of 0.11 g/cm²
The population is a postmenopausal woman, so the T-score is the right score and the WHO categories apply
Difference from the young-adult mean: 0.62 − 0.85 = −0.23 g/cm²
Divide by the reference SD: −0.23 ÷ 0.11 = −2.1
−2.1 lies between −1.0 and −2.5, so this is low bone mass — osteopenia — and not osteoporosis. It is 0.4 SD from the osteoporosis threshold
Now the mistake this page exists to prevent. The Lunar female young-normal femoral neck mean is 1.00 g/cm² against Hologic's 0.85, and it is tempting to read the same 0.62 against it: (0.62 − 1.00) ÷ 0.11 = −3.5, osteoporosis instead of osteopenia. That calculation is invalid. The two means differ because the two scanners report different absolute values for the same bone — Lunar reads roughly 6% higher at the hip — so a Hologic reading may only ever be divided by a Hologic reference
Convert it properly and the disagreement mostly disappears. Through the standardised-BMD equations for the femoral neck, a Hologic 0.62 is sBMD 693 mg/cm², which is a Lunar reading of about 0.76 g/cm²; against the Lunar reference that is (0.76 − 1.00) ÷ 0.11 = −2.2, next to Hologic's −2.1. The −3.5 was an artefact of mixing one machine's number with another machine's normal range, not a real difference between scanners
That is why the ISCD asks manufacturers to use NHANES III for femoral neck and total hip T-scores: the systematic difference between those two manufacturers' own femoral neck T-scores was measured at 0.9 SD, and recalculating both against NHANES data made them agree
And change only the population. Had this been a 34-year-old woman, the T-score would be the wrong score: the ISCD prefers a Z-score, computed against an age-matched mean, and osteoporosis would not be diagnosable from bone density alone
Finally, what −2.1 does not settle. Most fragility fractures in the population happen to people in this band rather than below it, because far more people are in it. Absolute fracture risk — age, previous fracture, glucocorticoids, family history — is the question that decides treatment, and none of it is in this arithmetic

WHO densitometric categories, and who they apply to

The first four rows are the WHO classification and the last two are the ISCD’s wording for the groups the WHO categories do not cover. The distinction is not cosmetic: the ISCD states that Z-scores, not T-scores, are preferred in premenopausal women and men under 50, and that osteoporosis cannot be diagnosed in men under 50 on the basis of bone density alone. Applying ‘osteoporosis’ to a T-score of −2.6 in a 30-year-old is a common error and is a diagnosis the classification does not support.
Category	T-score	Applies to
Normal	−1.0 or greater	Postmenopausal women; men aged 50 and over
Low bone mass (osteopenia)	Between −1.0 and −2.5	Postmenopausal women; men aged 50 and over
Osteoporosis	−2.5 or below	Postmenopausal women; men aged 50 and over
Severe (established) osteoporosis	−2.5 or below with a fragility fracture	Postmenopausal women; men aged 50 and over
Below the expected range for age	Z-score −2.0 or lower	Premenopausal women; men under 50; children
Within the expected range for age	Z-score above −2.0	Premenopausal women; men under 50; children

The reference database is part of the answer

The last-but-one row is this page’s own arithmetic on the published reference values in the row above, and it is the argument for the ISCD position. One bone, one measurement, two reference populations, two different diagnoses. The measured 0.9 SD systematic difference between the two manufacturers’ femoral neck T-scores is more than a third of the distance from normal to osteoporosis, and the cause was not the scanners: it was that they used different normal reference populations and different statistical models for the young-normal mean and SD. Recalculating both from NHANES data made the results equivalent.
Quantity	Hologic	Lunar
Female young-normal femoral neck mean, recalculated from NHANES	0.85 g/cm²	1.00 g/cm²
Reference SD used at that site	0.11 g/cm²	0.11 g/cm²
Systematic difference in femoral neck T-score with each manufacturer’s own normative data	0.9 SD	0.9 SD
Difference at the lumbar spine (L1–L4)	None of clinical significance	None of clinical significance
T-score for a measured femoral neck BMD of 0.62 g/cm² against each reference above	−2.1 (osteopenia)	−3.5 (osteoporosis)
ISCD position	Use NHANES III as the reference standard for femoral neck and total hip T-scores	Same

Which site, which region of interest

Two practical consequences hide in this table. Because the hip figure that counts is the lower of the femoral neck and the total hip, and because the spine and the hip routinely disagree, a single number handed over the phone is rarely enough to classify a patient. And because manufacturers may still use their own lumbar spine reference databases while NHANES III is asked for at the hip, a spine T-score and a hip T-score from the same report are not referenced to the same thing.
Question	ISCD position
Which sites to measure	Measure BMD at both the PA spine and the hip in all patients
Which spine region	PA L1–L4
Which hip region	Femoral neck or total hip, whichever is lowest
Which site defines osteoporosis in a postmenopausal woman or a man 50+	Lumbar spine, total hip or femoral neck with a T-score of −2.5 or less
The WHO reference standard site	A T-score of −2.5 or less at the femoral neck
When to use the forearm	Only when the hip and/or spine cannot be measured or interpreted
Which reference for T-scores	A uniform White, non-race/ethnicity-adjusted female database for women of all ethnic groups, and the same female reference for men of all ethnic groups; NHANES III for femoral neck and total hip
Which reference for the lumbar spine	Manufacturers may use their own databases
Men under 50	Osteoporosis cannot be diagnosed on the basis of BMD alone
Premenopausal women	The WHO criteria may apply during the menopausal transition

Two scores, two reference populations, and one error worth avoiding

Both scores are the same arithmetic — a measured bone mineral density expressed as a number of standard deviations from a reference mean — and they differ only in which reference population is used. A T-score compares the patient with a young adult of the same sex at peak bone mass. A Z-score compares them with people of their own age and sex. That single difference is what makes them answer different questions: a T-score asks how much bone has been lost from the young-adult peak, which is what matters after the menopause or in later life, and a Z-score asks whether the bone density is where it should be for someone of this age, which is what matters in anyone who has not yet lost bone with age.

The error worth avoiding follows directly. A 32-year-old woman has a T-score, and it is meaningless as a diagnosis: she is being compared with young adults, which is what she is, and if she sits at the lower end of a normal distribution her T-score will be low without anything being wrong. The ISCD’s position is explicit — Z-scores, not T-scores, are preferred in premenopausal women and men under 50 — and it goes further for men, stating that osteoporosis cannot be diagnosed in men under 50 on the basis of bone density alone. Reporting ‘osteoporosis’ from a T-score in a young patient is common, generates a great deal of avoidable anxiety, and can start a treatment that was never indicated. The right reading of a Z-score of −2.0 or lower in that group is the ISCD’s own phrase, ‘below the expected range for age’, which is a prompt to look for a secondary cause rather than a diagnosis. It usually finds one: glucocorticoids, amenorrhoea or hypogonadism, low body weight, coeliac disease or another malabsorption, primary hyperparathyroidism, thyrotoxicosis, chronic kidney disease, or an anticonvulsant.

The second thing this page exists to show is that the reference population is part of the answer. In the mid-1990s the two dominant scanner manufacturers produced femoral neck T-scores that differed systematically by 0.9 standard deviations, with no clinically significant difference at the lumbar spine. The scanners were not the problem: the manufacturers used different normal reference populations and different statistical models for the young-normal mean and standard deviation. Recalculating both from NHANES data — which gave female young-normal femoral neck values of 0.85 ± 0.11 g/cm² for one and 1.00 ± 0.11 g/cm² for the other — made the results equivalent. That is why the ISCD asks manufacturers to continue using NHANES III as the reference standard for femoral neck and total hip T-scores, and why it asks for a uniform, non-race-adjusted female reference for women and for men of all ethnic groups. Nought point nine of a standard deviation is more than a third of the distance from normal to osteoporosis; a reader who does not know which database produced a T-score does not fully know what the T-score means.

Site and region of interest matter in the same way. The ISCD asks for the PA spine and the hip in all patients, uses PA L1–L4 for the spine and the femoral neck or total hip — whichever is lower — for the hip, and reserves the forearm for when the hip and spine cannot be measured or interpreted. Spine and hip routinely disagree, and the disagreement is usually informative rather than noise: degenerative change, a vertebral fracture, scoliosis and aortic calcification all raise apparent spine density and can hide a lower hip value. Diagnosing from whichever number happens to be quoted, or from an average of several sites, is how a patient gets classified by an artefact.

Finally, neither score is a fracture risk, and treating them as one is the most consequential misreading of all. Most fragility fractures in the population occur in people whose T-score is in the osteopenic range rather than below −2.5, simply because far more people are in that range. Absolute risk depends on age, previous fracture, parental hip fracture, glucocorticoid use, smoking, alcohol and rheumatoid arthritis as well as on the density — and a clinical diagnosis of osteoporosis may be made independently of the bone density in someone who has already had a fragility fracture. Use the score as one input. For the biochemistry that sits beside it, the albumin-corrected calcium, the PTH converter, the vitamin D status interpreter and the bone turnover markers — CTX and P1NP — are on the site.

Frequently asked questions

How is a T-score calculated?

Subtract the young-adult reference mean BMD from the measured BMD and divide by the standard deviation of that young-adult reference population, all in g/cm². A femoral neck BMD of 0.62 g/cm² against a reference mean of 0.85 and an SD of 0.11 gives (0.62 − 0.85) ÷ 0.11 = −2.1. The result is a number of standard deviations, not a percentage and not a density.

What is the difference between a T-score and a Z-score?

Only the reference population. A T-score compares the measured BMD with a young adult of the same sex at peak bone mass; a Z-score compares it with people of the same age and sex. So a T-score measures loss from the young-adult peak and a Z-score asks whether the density is as expected for the patient’s age. T-scores are preferred in postmenopausal women and men aged 50 and over, and Z-scores in premenopausal women, men under 50 and children.

What T-score means osteoporosis?

−2.5 or below. The WHO international reference standard for the diagnosis of osteoporosis is a T-score of −2.5 or less at the femoral neck, and the ISCD extends that to the lumbar spine, total hip or femoral neck in postmenopausal women and men aged 50 and over. Between −1.0 and −2.5 is low bone mass, or osteopenia; −1.0 or greater is normal; and −2.5 or below with a fragility fracture is severe or established osteoporosis.

Can you diagnose osteoporosis from a T-score in a 30-year-old?

No. The ISCD’s position is that Z-scores, not T-scores, are preferred in premenopausal women and men under 50, and specifically that osteoporosis cannot be diagnosed in men under age 50 on the basis of BMD alone. In that group a Z-score of −2.0 or lower is reported as ‘below the expected range for age’, which is a prompt to look for a secondary cause — glucocorticoids, hypogonadism or amenorrhoea, low body weight, malabsorption, primary hyperparathyroidism, thyrotoxicosis, chronic kidney disease, anticonvulsants — rather than a diagnosis.

Does the reference database change the T-score?

Substantially. Two scanner manufacturers’ femoral neck T-scores were found to differ systematically by 0.9 standard deviations when each used its own normative data, with no clinically significant difference at the lumbar spine; the cause was different normal reference populations and different statistical models for the young-normal mean and SD. Recalculating both against NHANES data made them equivalent, which is why the ISCD asks manufacturers to continue using NHANES III for femoral neck and total hip T-scores.

Which site should be used for diagnosis?

The ISCD asks for BMD at both the PA spine and the hip in all patients, using PA L1–L4 for the spine and the femoral neck or total hip — whichever is lowest — for the hip, with the forearm reserved for when the hip and spine cannot be measured or interpreted. The WHO reference standard is specifically the femoral neck. Do not diagnose from an average of sites, and treat a spine result with caution where there is degenerative change, a vertebral fracture, scoliosis or aortic calcification, all of which raise apparent spine density.

Is a normal T-score reassuring about fracture risk?

Only partly. Most fragility fractures in the population occur in people whose T-score is in the osteopenic range rather than below −2.5, because far more people are in that range. Absolute fracture risk depends on age, previous fracture, parental hip fracture, glucocorticoid use, smoking, alcohol and rheumatoid arthritis as well as on bone density, and a clinical diagnosis of osteoporosis may be made independently of bone density in someone who has had a fragility fracture.

Related calculators

References

International Society for Clinical Densitometry. 2023 ISCD Official Positions — Adult. iscd.org — the T-score versus Z-score populations, the WHO reference standard of a femoral neck T-score of −2.5 or less, the NHANES III requirement for femoral neck and total hip T-scores, the uniform non-race-adjusted female reference database, PA L1–L4 and the lower of femoral neck and total hip, the forearm exception, the Z-score wording ‘below the expected range for age’ at −2.0 or lower, and the position that osteoporosis cannot be diagnosed in men under 50 on BMD alone.
Osteoporosis: clinical evaluation. In: Endotext. South Dartmouth (MA): MDText.com (NBK279049), Table 1 — the WHO densitometric categories with their exact T-score boundaries, the definition of the T-score as the measured BMD minus the young-adult reference mean divided by the young-adult reference SD, the Z-score’s meaning below −2.0, and the clinical diagnosis of osteoporosis independent of BMD after a fragility fracture.
Faulkner KG, Roberts LA, McClung MR. Discrepancies in normative data between Lunar and Hologic DXA systems. Osteoporos Int. 1996;6(6):432–436 — the 0.9 SD systematic difference in femoral neck T-scores between manufacturers, its cause in differing reference populations and statistical models, and the NHANES-recalculated female young-normal femoral neck values of 0.85 ± 0.11 g/cm² (Hologic) and 1.00 ± 0.11 g/cm² (Lunar).
Kanis JA, Melton LJ 3rd, Christiansen C, Johnston CC, Khaltaev N. The diagnosis of osteoporosis. J Bone Miner Res. 1994;9(8):1137–1141 — the original WHO working-group definition on which the T-score categories rest.

Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.