DEXA scans: are you getting the bone test or the body-composition test?

You can hold the same weight for a decade and still be losing muscle and bone the whole time. A bathroom scale reports none of that. A DEXA scan reports all of it, but only if you booked the right one: clinics sell two different scans off the same machine, and the one that takes a flattering picture of your body fat is not the one that can diagnose osteoporosis. Knowing which you are buying is the single most useful thing on this page.

What is the difference between the two DEXA scans?

The bone scan, central DXA, images your lower spine and one or both hips, and those are the sites the diagnosis is written around: the lumbar spine, the femoral neck or the total hip.

The whole-body scan is a different acquisition. It splits your weight into fat, lean soft tissue and bone mineral by region, including the fat sitting deep in the abdomen. It also produces a whole-body bone number, and that number is explicitly not what a diagnosis is made from.

DXA is the current abbreviation for dual-energy X-ray absorptiometry; DEXA is the older spelling that stuck.

How does the machine tell bone from fat?

DEXA sends two X-ray beams of different energies through you, like shining two colors of light through stained glass. Each material dims each color by a different amount, so the pair of shadows says what is in the way and how much of it there is.

The dose is small, under a tenth of a chest X-ray. It is still ionising radiation, which is why the scan is not done in pregnancy, but on the scale of medical imaging it sits at the bottom.

What does the bone scan actually tell you?

It reports how much mineral sits in a measured area of bone, as a score against a young adult reference population.

That score is the T-score: minus 1.0 or higher is normal, minus 1.0 to minus 2.4 is low bone density, often called osteopenia, and minus 2.5 or lower meets the definition of osteoporosis. Premenopausal women and men under 50 get a Z-score instead, which compares them against people their own age and sex.

The number is areal, not volumetric: a larger bone reads as denser than a smaller bone of identical true density, so body size is baked into the score.

And the spine is the fussier site: the scan is less accurate after spinal surgery, deformity, fracture or arthritis, all of which become more common with age and all of which push the number up rather than down. It is why a report gives more than one site.

The US Preventive Services Task Force recommends screening women 65 and older, and postmenopausal women under 65 at increased risk. For men the evidence is insufficient to weigh benefits against harms, which is not a recommendation against screening men but a statement that the studies have not been done.

What bone density does not predict

The most important thing on this page: a T-score sorts populations well and sorts individuals poorly.

The arithmetic is not mysterious. Women with osteoporosis had the highest fracture rate per person and were also the smallest group, and a high rate in a small group produces fewer total fractures than a moderate rate in a large one. The scan was right about risk and still missed most of the fractures: roughly three in four happened in women it had not labeled.

The same caution shows up on the treatment side. A meta-regression of placebo-controlled trials asked how much of each drug's fracture benefit was explained by how much it moved bone density. For vertebral fractures, most of it. For hip fractures, less. For non-vertebral fractures, no significant association at all. Even where a treatment works, density is a partial account of why, and at one site not an account at all.

The score is a risk factor, not a forecast.

The ACR and RSNA say it plainly: the test cannot predict who will experience a fracture. That is why clinicians combine bone density with clinical risk factors in a formal risk assessment rather than reading a T-score alone, and why a normal result is a reason for reassurance rather than a reason to stop paying attention to falls, strength and balance.

What does the body-composition scan tell you?

It splits your weight into fat, lean soft tissue and bone mineral, and says where on your body each one sits.

That is more than a scale or a BMI can do. Two people at the same weight and height can carry very different amounts of muscle, and fat around the organs behaves differently from fat under the skin. A scan separates those and gives you a repeatable number to train against.

The visceral fat estimate deserves a note, because it is the line clinics circle. Checked against MRI it agreed closely, but whether it tracks change in one person over time is unestablished. Good at telling one group from another; less good at watching your own number.

One asymmetry belongs here. Bone-density screening has a national recommendation for defined groups; whole-body composition scanning has none for healthy adults. That does not make it useless, but its case rests on the measurement being useful to you and to whoever reads it, rather than on evidence that scanning well people improves what happens to them.

Side by side

Why your number moves, and how much of that is you

Some of the change between two scans is your body. Some is the scan.

Start with water. Drinking water equal to 1.5% of body mass raised measured whole-body lean tissue by about 0.8 kg within the hour, with no muscle gained. If you are watching lean mass across scans, arriving in a consistent state matters more than most people are told.

Then the machine. Scanned on two manufacturers' systems, the same adults' lumbar spine density read up to 20% lower on one than the other. Two reports from two clinics cannot be laid side by side and read as change. The gap between machines is larger than the change you are hoping to detect.

This is what the least significant change is for. A facility is expected to know its own precision and state the smallest change that counts as real for that machine and operator. Anything smaller is not a small improvement, it is noise wearing a decimal point.

Where a DEXA scan stops

It cannot tell you why. Low bone density has many possible causes and the scan sees none of them. Sorting them out takes a history and blood tests, and that is what turns a number into a plan.

It cannot measure bone quality. It sees mineral in a projected area, not the internal architecture of the bone, and both decide whether a bone breaks.

It does not measure whether a muscle works. Lean mass is a quantity, not a capacity: a scan says how much lean tissue is in your legs and nothing about what those legs can do, which is why assessments pair one with a VO2 max test.

It is not a screening scan for anything else, and does not replace one. A whole-body DEXA is not looking for tumors, organ disease or vascular problems, should never be mistaken for whole-body imaging, and changes nothing about what your age and history already call for.

And a single number is not a verdict. A T-score or a fat percentage is a measurement placed against a reference population that may not resemble you, the same reading problem as biological age scores.

Who tends to get something out of it

For the bone scan, the clearest cases are the ones national screening already names: women 65 and older, postmenopausal women under 65 with added risk, and anyone of any age on long-term steroids, with a fracture from a minor fall, or with a condition that affects bone. If you are in one of those groups this is not a longevity purchase at all, it is ordinary care.

For the body-composition scan, it suits someone whose weight has been stable while they suspect the composition has not, someone starting a training or nutrition change who wants a before, and anyone who wants a number to train against rather than a mirror. It has least to offer if you would not do anything differently on the strength of it, or if you cannot commit to retesting on the same machine, because a single body-composition scan without a comparison is close to a curiosity.

How often, and what to ask before you book

Bone changes slowly. In one study of women aged 67 and older, researchers estimated it would take roughly 15 years for 10% of those starting from normal or mildly low density to develop osteoporosis. Annual rescanning from that starting point measures a change that has not had time to happen. Someone with substantially lower density, or on treatment, is a different case, which is why the professional positions call for individualised intervals.

Body composition moves faster than bone and slower than the bathroom scale. A retest a few months after a real change in training shows something. Three weeks later, on a different machine, after a large breakfast, it cannot.

Every city page groups clinics by what they measure, including strength, fitness and body composition.

Sources

• MedlinePlus (NIH/NLM), Bone Density Scan
• NIAMS (NIH), Bone Mineral Density Tests: What the Numbers Mean
• ACR and RSNA, RadiologyInfo: Bone Density Scan (DEXA)
• U.S. Preventive Services Task Force, Osteoporosis to Prevent Fractures: Screening
• International Society for Clinical Densitometry, 2023 Adult Official Positions
• Pasco JA et al., The population burden of fractures originates in women with osteopenia, not osteoporosis. Osteoporosis International
• Bouxsein ML et al., Change in Bone Density and Reduction in Fracture Risk: A Meta-Regression of Published Trials. Journal of Bone and Mineral Research
• NIH News in Health, How Often Should Women Have Bone Tests?
• Systematic review of quantitative ultrasound and DXA, Frontiers in Endocrinology
• Cross-calibration of GE Lunar Prodigy and Hologic Horizon W, Journal of Bone Metabolism
• Neeland IJ et al., Comparison of visceral fat mass measurement by DXA and MRI in a multiethnic cohort: the Dallas Heart Study. Nutrition & Diabetes
• Influence of passive, diet-based hydration on muscle quantity assessment, Frontiers in Physiology