A DEXA scan is the reference test for bone density and one of the most accurate ways an ordinary person can measure body composition: the fat and muscle a scale cannot tell apart, and where the fat sits. It defines osteoporosis from the T-score and reads fracture risk from bone density, and it measures body fat and lean mass closely enough to track a meaningful change over months. Four limits qualify the numbers. Compare scans on the same machine, because different scanners do not read alike.
The muscle-mass cutoffs that define sarcopenia are still loosely drawn. It estimates the deep belly fat around the organs instead of measuring it directly the way a CT scan does. And the radiation is very low, far below a CT scan.
What It Is
DEXA, written more precisely as DXA, stands for dual-energy X-ray absorptiometry. It passes two low-dose X-ray beams of different energies through the body, and from how much each beam is absorbed it works out how much of what it passed through is bone, how much is fat, and how much is lean soft tissue. That single capability is why one scan measures two separate things: how dense your bones are, and how your weight divides into fat and muscle. Bone density and body composition are distinct outputs from the same machine. It is the reference standard for diagnosing osteoporosis, and one of the most accurate ways to measure body composition that an ordinary person can get to.
It measures what the everyday tools cannot. A bathroom scale gives one number for the whole body, and BMI counts a kilogram of muscle and a kilogram of fat the same, so a lean, muscular person can read as overweight on paper while someone at a normal weight can carry the central fat that raises risk. DEXA measures the parts instead of the total.
How a scan works. You lie still on a padded, open table for a few minutes while an arm passes over you. Nothing touches you and there is no tunnel. What the scan reports depends on the study ordered. For bone, the beam measures bone mineral density, usually at the hip and the lumbar spine, the two sites where a fracture matters most. For body composition, a whole-body scan splits your weight into fat, lean soft tissue and bone mineral, both for the whole body and region by region, so the arms, legs and trunk show separately.
What the numbers mean. Bone comes back as two comparisons. The T-score compares your bone density to a healthy young adult at peak bone mass, and diagnosis is built on it.
At or below -2.5 is osteoporosis. Between -1 and -2.5 is low bone mass. Above -1 is normal.
The Z-score compares you to other people your own age and sex, the more useful reading in younger adults and children, where a score well below the age-matched average points to an underlying cause worth investigating. The body-composition numbers have no single diagnostic cutoff the way the T-score does. Their value is in the parts and in the trend: total body fat, the split between fat and lean, the amount of muscle in the limbs, and the estimate of the visceral fat packed around the organs, each most informative once you have a baseline to compare a later scan against. This scan belongs to the wider family of home and clinical self-checks on the self-measures hub, and it is the deep-dive on the one that needs a clinic and a machine.
What It Measures Well
DEXA is graded here on study design as much as on how strong a result is. The bone results rest on large pooled cohorts and the diagnostic framework the field is built on. Several of the body-composition results are validation studies and cross-sectional snapshots, so they show that DEXA measures a thing accurately more than they show that changing the thing changes your future.
On bone, it is the reference. The T-score that defines osteoporosis comes from DXA, and the lower the density, the higher the fracture risk, in a steady gradient across the whole range. Modern DXA also reads a second bone number from the same spine image, the trabecular bone score, which grades bone texture and adds fracture-risk information beyond density alone. On body composition, it measures fat and lean tissue directly, tracks the best available laboratory model within about two percentage points, and its estimate of visceral fat lines up closely with a CT scan at a fraction of the radiation. It is also one of the standard tools for confirming the low muscle quantity, measured as appendicular lean mass, that defines sarcopenia.
The Research & Studies
Everything here is based on the research we have collected and checked, sorted into groups and ordered with the strongest evidence first. Click any claim to open the studies behind it.
Measurement And Diagnosis
Osteoporosis is defined by a DEXA T-score at or below -2.5
The number used to diagnose osteoporosis, the T-score, comes from a DXA scan. A T-score of -2.5 or lower means osteoporosis, and DXA of the hip and spine is the test that standard is based on.
The World Health Organization diagnostic thresholds, set from bone mineral density measured by DXA, define osteoporosis as a bone density 2.5 or more standard deviations below the young-adult mean, a T-score at or below -2.5, with the range from -1 to -2.5 termed low bone mass. DXA of the hip and spine is the measurement these thresholds are built on, which is why it is treated as the reference standard for diagnosis. The T-score is a statistical comparison to a young-adult reference rather than a direct measure of bone strength, and diagnosis and treatment weigh it alongside age, previous fractures and other risk factors.
The study · 1
Kanis et al., The diagnosis of osteoporosis · J Bone Miner Res 1994;9(8):1137-41
DEXA body fat tracks the lab criterion model within about two percentage points
DXA measures body fat about as well as the best laboratory method available, within roughly two percentage points, though it reads a little low in very lean people.
In 152 healthy adults aged 18 to 59, DXA percent body fat tracked the four-compartment criterion model closely (r-squared 0.952, standard error of estimate 1.6 percent fat), though DXA read about 1.8 percentage points lower on average (18.9 versus 20.7 percent) and progressively underestimated fat in leaner people. What could explain it instead: The four-compartment reference model carries its own measurement error, and beam-hardening from differences in body thickness can bias DXA, so part of the disagreement reflects the limits of both methods rather than DXA alone.. Agreement is close but not exact, and the small bias means DXA is best for tracking your own change over time rather than for pinning down a single true percentage.
The study · 1
Van Der Ploeg et al., Percent body fat via DEXA: comparison with a four-compartment model · J Appl Physiol 2003;94(2):499-506
DEXA appendicular lean mass is a standard confirmation of sarcopenia
When clinicians diagnose age-related muscle loss, called sarcopenia, DXA is one of the standard tools they use to confirm how much muscle a person actually has.
The 2019 European Working Group on Sarcopenia in Older People consensus lists DXA-measured appendicular lean mass, adjusted for height, as one of the recommended ways to confirm the low muscle quantity that defines sarcopenia, used alongside a measure of muscle strength. DXA quantifies how much lean tissue is present, not its quality or function, so it sits within a diagnosis that also requires a strength or performance test.
The study · 1
Cruz-Jentoft et al., Sarcopenia: revised European consensus on definition and diagnosis (EWGSOP2) · Age Ageing 2019;48(1):16-31
DEXA visceral fat matches a CT scan closely, r-squared 0.957
DXA can estimate the fat packed around your organs, the visceral fat that matters most for metabolic health, and it lines up closely with a CT scan while using far less radiation.
In 124 adults aged 18 to 90 scanned by both methods within an hour, automated DXA visceral fat agreed closely with abdominal CT, the reference standard (r-squared 0.957 combined), and the authors concluded DXA can measure visceral fat precisely in both men and women at very low radiation. What could explain it instead: CT, the reference method, carries its own measurement error, and DXA estimates a volume from a two-dimensional scan, so some of the scatter between them reflects the limits of both techniques.. Agreement was close on average but the individual limits of agreement were fairly wide, so a single DXA visceral-fat number is better read as a trend over time than as an exact volume.
The study · 1
Kaul et al., Dual-energy X-ray absorptiometry for quantification of visceral fat · Obesity (Silver Spring) 2012;20(6):1313-8
Progress Markers
Each SD drop in bone density carries about 1.5 times the fracture risk
The lower your DXA bone-density score, the higher your fracture risk, in a steady gradient. Each one-step drop in density raised fracture risk by roughly half, and measuring at the hip predicted hip fractures best.
In a meta-analysis of 11 prospective cohorts with about 90,000 person-years of follow-up and more than 2,000 fractures, each one standard deviation fall in bone mineral density carried a relative risk of about 1.5 (95% CI 1.4 to 1.6) for fracture, rising to 2.3 (1.9 to 2.8) for spine measurement predicting vertebral fracture and 2.6 (2.0 to 3.5) for hip measurement predicting hip fracture. Bone density predicts fracture risk across a population but cannot identify which individual will fracture, and the authors did not recommend mass screening on density alone.
Who this may not transfer to:The prospective cohorts pooled here were of women. The same density-to-fracture gradient has since been documented in men, though this meta-analysis itself did not measure it, so the exact figures are from women.
The study · 1
Marshall et al., Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures · BMJ 1996;312(7041):1254-9
Each SD drop in trabecular bone score carries about 1.4 times the fracture risk, independent of bone density
Modern DXA reads a second bone number from the spine scan, the trabecular bone score, which grades bone texture. Each one-step fall in it carried about 1.4 times the fracture risk, and it still predicted fractures after accounting for bone density, so it tells you something density alone does not.
In an individual-level meta-analysis of 17,809 men and women across 14 prospective cohorts followed a mean of 6.7 years, the trabecular bone score, a texture measure read from the same lumbar-spine DXA image, carried a gradient of risk of 1.44 (95% CI 1.35 to 1.53) per standard deviation for major osteoporotic fracture, and 1.32 (95% CI 1.24 to 1.41) after adjustment for the FRAX 10-year fracture probability, so it added fracture-risk information beyond bone density itself. The gradient was similar in men and women. The trabecular bone score is an addition to bone density and a fracture-risk calculator rather than a replacement, and how much it should shift a clinical decision is still being worked out.
Who this may not transfer to:The pooled cohorts included both men and women and the gradient of risk was similar in each, so the finding reads across the sexes.
The study · 1
McCloskey et al., A meta-analysis of trabecular bone score in fracture risk prediction and its relationship to FRAX · J Bone Miner Res 2016;31(5):940-8
Low muscle mass tracks with a higher death rate in older men
In large groups of older men followed over years, those with the muscle loss that DXA helps identify tended to die at higher rates than those without it.
Across pooled population-based cohorts of older men, sarcopenia defined by recent criteria, which include low DXA-measured muscle mass, was associated with higher all-cause mortality over follow-up, with the size of the association varying by which definition was applied. What could explain it instead: Reverse causation and underlying illness. Serious disease both wastes muscle and raises the risk of death, so part of the association reflects existing illness rather than low muscle mass causing death.. This analysis was in men only, the strength of the association depended on the definition used, and it is observational rather than a trial.
Who this may not transfer to:Measured in men only. Age-related muscle loss tracks with mortality in women in other cohorts, but this analysis did not include women, so the size of the association here should not be read across to them.
The study · 1
Westbury et al., Recent sarcopenia definitions: prevalence, agreement and mortality associations among men · J Cachexia Sarcopenia Muscle 2023;14(1):565-575
Visceral fat tracks cardiometabolic risk more tightly than overall body fat
The visceral fat DXA measures lines up with heart and metabolic risk factors more tightly than overall body fat does, and a little better than a tape-measure waist.
In 4,831 Australian adults aged 45 to 69, DXA-measured central fat, including abdominal visceral adipose tissue, showed stronger associations with metabolic syndrome and cardiometabolic risk factors than overall-adiposity measures, and DXA visceral fat was more strongly associated than waist circumference. What could explain it instead: Cross-sectional design and reverse causation. The scan and the risk factors were measured at the same time, so the data cannot show that visceral fat came first, and metabolic disease can itself change where fat is stored.. This was a cross-sectional snapshot, so it shows association at one point in time rather than predicting future events, and the added value over a simple waist measurement was modest.
The study · 1
Zhu et al., DXA-derived versus standard anthropometric measures for predicting cardiometabolic risk in middle-aged Australian men and women · J Clin Densitom 2022;25(3):299-307
How it works
DEXA is precise, but scans compare only on the same machine
DXA is precise enough to track small real changes, but different machines do not read exactly alike, so repeat scans should be done on the same machine to compare fairly.
The International Society for Clinical Densitometry sets minimum precision standards for DXA and requires a cross-calibration study whenever a scanner is changed or replaced, and especially between manufacturers, because bone-density readings do not transfer directly from one machine to another. A change that looks meaningful can be machine drift instead, which is why it must exceed the scanner and technologist precision error before it counts.
For tracking bone or body composition, return to the same DXA machine each time. If you must switch, a result from a different scanner or manufacturer is not directly comparable without a cross-calibration.
The study · 1
Shepherd et al., Cross-calibration and minimum precision standards for dual-energy X-ray absorptiometry: the 2005 ISCD Official Positions · J Clin Densitom 2006;9(1):31-6
Its Limits
DEXA is accurate within four limits, and each one changes how a number should be read.
Compare scans on the same machine, not across machines. Different DXA scanners, especially from different manufacturers, do not read exactly alike, and the professional standards require a cross-calibration study whenever the hardware changes. A difference between two clinics can be the machines, not you, so return to the same machine each time.
A change must clear the measurement error before it counts. DEXA is precise enough to track small changes, but a single reading carries some scanner and technologist error. A change has to exceed that least-significant-change margin, so read the trend across a baseline and a repeat, not a single reading.
The sarcopenia cutoffs are still loosely drawn. DEXA measures how much lean tissue is present, not its quality or its strength, and the appendicular-lean-mass thresholds that define low muscle quantity vary between the consensus definitions. That is why a muscle-mass number is paired with a strength or performance test.
It estimates visceral fat instead of measuring it directly. The visceral-fat figure is drawn from the trunk region of a two-dimensional scan, not a direct volume the way a CT slice measures one. Agreement with CT is close on average, but the individual limits are wide enough that a single visceral-fat number is better read as a trend over time than as an exact volume.
How to Use It
Treat a scan as a baseline you return to. Book the repeat at the same clinic on the same machine, and judge a change by whether it clears the measurement error.
For bone, your clinician sets the retest interval from your starting score, your risk and any treatment. Many people are rescanned every one to two years, sometimes longer when bone is stable, because bone changes slowly and scanning too often mostly measures the machine. For body composition, a scan every few months to a year is enough to see whether a change in training or eating is moving fat and muscle in the direction you intended, and reading it more often mostly reads noise.
Who gets the most from it: postmenopausal women and older adults, in whom bone loss and muscle loss are common and the bone result can change management; anyone with a condition or medication that thins bone; and people tracking muscle and fat while training or losing weight, for whom the scan shows whether what they are losing is fat and not the muscle they want to keep. If a low reading points toward thin bone, the osteoporosis guide covers what rebuilds it and where the medications fit. If you are lifting to hold muscle as you lean out, DEXA shows whether it is working.
Go Deeper
- Osteoporosis: what a low bone-density reading means and the training, nutrition and medication that lower fracture risk, in order.
- Self-measures hub: DEXA alongside grip strength, walking speed and blood pressure, each with the protocol its numbers came from.
- Grip strength: a no-equipment marker of whole-body strength and aging that complements the muscle-mass picture.
- Protein and muscle: the intake that builds and holds the lean mass a scan measures.
- Muscle as an organ: why the lean mass on a DEXA report matters for metabolism and aging, not just strength.
The Chinese Medicine View
Reading A DEXA Scan Well
Everything to be aware of is here, in one place. This practice suits most healthy people; a few situations call for real care.
A DEXA scan uses very little radiation, far less than a CT scan
A review of X-ray imaging used to assess the skeleton reports that radiation doses from DXA are very low, well below the 1 to 3 millisieverts delivered by CT-based bone methods, while noting that, as with any X-ray, each scan should be clinically justified. Very low is not none. DXA is generally avoided in pregnancy, and children are more sensitive to radiation than adults, so their scan protocols are adjusted to keep the dose down.Damilakis et al., Radiation exposure in X-ray-based imaging techniques used in osteoporosis
The radiation is very small
A DEXA scan uses a very low dose of radiation, far below a CT scan, though it is not zero. That is why each scan should have a clinical reason behind it and is not run out of curiosity.
Avoided in pregnancy, and adjusted for children
Low is not none. DEXA is generally avoided in pregnancy, and children are more sensitive to radiation than adults, so their scan protocols are set to keep the dose down. Tell the clinic if you are or might be pregnant.
Compare scans on the same machine
Different DXA scanners, especially from different manufacturers, do not read exactly alike. To track a change, return to the same machine each time. A result from a different scanner is not directly comparable without a cross-calibration, so a difference between two clinics can be the machines, not you.
A number to trend, not a verdict
One scan is a snapshot, and a single reading carries some machine and technologist error, so a change has to exceed that margin before it means something. A T-score or a body-composition number is one input that a clinician weighs alongside age, fracture history and medications. Nothing here is a reason to start, stop or change a medication.
Start slow, be smart, read the research, and consult a professional if you have any concerns. This is here to inform your choice, not make it for you.
Common Questions
What is a DEXA scan, and what does it measure?
DEXA, written more precisely as DXA, is dual-energy X-ray absorptiometry. It sends two low-dose X-ray beams of different energies through the body and, from how much each is absorbed, separates bone, fat and lean soft tissue. That lets one scan report bone mineral density, usually at the hip and spine, and whole-body composition, the split of your weight into fat and muscle, including the visceral fat around the organs. You lie still on an open table for a few minutes and nothing touches you.
What does my T-score mean?
The T-score compares your bone density to a healthy young adult at peak bone mass. Above -1 is normal, -1 to -2.5 is low bone mass, and -2.5 or lower is the diagnostic line for osteoporosis, the standard set from DXA measurement. Risk rises steadily as the score falls: each one standard deviation drop carried about 1.5 times the fracture risk across pooled cohorts. In younger adults and children the Z-score, which compares you to others your own age, is the more useful reading.
Is DEXA accurate for body fat, compared with a smart scale or BMI?
For where your weight actually sits, yes. BMI cannot tell muscle from fat, and home bioimpedance scales estimate body fat indirectly from a small electrical current, which drifts with hydration. DEXA measures fat and lean tissue directly and, in a study of 152 adults, tracked the best available laboratory model within about two percentage points of body fat. It also reports the parts, arms, legs, trunk and the visceral fat around the organs, which a single scale number cannot. It reads a little low in very lean people, so it is best for tracking your own change over time.
How much radiation does a DEXA scan use?
A very low dose, far below a CT scan, which is one reason the same scan can be repeated to track a change. It is not zero, so each scan should have a clinical reason, DEXA is generally avoided in pregnancy, and children are scanned on dose-reduced protocols because they are more sensitive to radiation than adults.
How often is it worth getting a DEXA scan?
For bone, the interval is a clinical decision, often every one to two years and sometimes longer when bone is stable, because bone changes slowly and scanning too often mostly measures the machine. For body composition, every few months to a year is enough to see whether a change in training or eating is moving fat and muscle as intended. Reading either more often than that mostly reads noise.
Can I compare scans from different machines?
Not directly. Different DXA scanners, particularly from different manufacturers, do not read exactly alike, and the professional standards require a cross-calibration whenever the hardware changes. If your second scan is on a different machine, a shift in the number can be the scanner, not your body. Returning to the same clinic and machine is the simplest way to make sure a change you see is a change that happened.
Explore Related
Other pages this one connects to, by the evidence they share, the outcomes they touch, and the ground they cover.
All 10 sources on this page independently checked and cross-referenced.
Thomas Dehli, Founder & Editor, Sacred Lotus
Sacred Lotus has published Chinese medicine reference material since 2001. Integrative pages are held to the same standard as the herb and formula library: cite the source, grade the claim at its real strength, and say where the research has not looked. This page is educational and it is not medical advice. Last reviewed and updated August 10, 2026.