Written and maintained by CASRAI Editorial Board
Last updated
Short answer: DXA body composition is a whole-body dual-energy X-ray absorptiometry scan that estimates three compartments: bone mineral content, fat mass and lean mass. It infers them from how two X-ray energies are attenuated, so the numbers are model-based estimates, and they depend on the device, hydration and positioning.
This guide is general reference information for researchers and lab and clinic staff. It is not medical advice and does not interpret any individual’s scan.
Definition and scope
Dual-energy X-ray absorptiometry (DXA, also written DEXA) is best known for bone densitometry. The same whole-body scan can also be analysed for soft-tissue composition. In the whole-body mode, software splits the body into bone mineral content (BMC), fat mass (FM) and lean mass (LM), and reports them for the total body and for regions such as arms, legs and trunk (Messina et al., 2020).
A body composition scan is not the same report as a bone density (T-score) assessment. The related risk calculator is covered separately in the FRAX score guide, and broader context sits in what is nutrition science.
How DXA body composition works
The scanner emits X-rays at two energies. Different tissues attenuate the two beams in different proportions, and the ratio of attenuation (the R-value) lets software separate bone from soft tissue, and then fat from lean tissue within the soft-tissue pixels. Roughly 40-45% of the scan area is bone, so fat and lean mass are only estimated directly in the remaining pixels (Messina et al., 2020).
Early pencil-beam systems have largely given way to fan-beam systems, which scan faster; the same review notes pencil-beam designs offered higher precision. The whole-body radiation dose is described in that review as about 4-5 μSv, and pregnancy is listed as only a relative contraindication in ISCD guidance.
Required versus optional report parameters (ISCD)
The International Society for Clinical Densitometry (ISCD) 2013 Official Positions on body composition reporting set out what a DXA body composition report should contain (Petak et al., 2013). Use this as a checklist when reviewing a report format or a vendor template.
| Parameter | ISCD 2013 status |
|---|---|
| BMI | Required |
| Bone mineral density and bone mineral content | Required |
| Total mass | Required |
| Total lean mass | Required |
| Total fat mass | Required |
| Percent fat mass | Required |
| Visceral adipose tissue | Optional |
| Appendicular lean mass index | Optional |
| Android/gynoid percent-fat ratio | Optional |
| Trunk-to-leg fat mass ratio | Optional |
| Lean mass index and fat mass index | Optional |
For the United States, the ISCD position points to the NHANES 1999-2004 dataset as the age-, sex- and race-specific reference. It describes Z-scores and percentiles as useful when non-normal distributions are adjusted for, and treats DXA indices as useful for risk stratification in obese or sarcopenic patients while noting that thresholds still need validation.
What DXA assumes, and what affects the result
| Factor | What the sources state |
|---|---|
| Hydration of fat-free tissue | DXA assumes fat-free tissue is about 73% water. In simulations, a 5% change in hydration produced about a 1% error in fat mass; ascites and edema also affect fat mass (Messina et al., 2020). |
| Meals and exercise | These shift fluid, for example from trunk to limbs, and so move lean mass; the review suggests an overnight fast before scanning (Messina et al., 2020). |
| Metal and positioning | Metal should be removed, and region line-placement errors distort regional values (Messina et al., 2020). |
| Manufacturer and model | Different densitometers may not agree, so follow-up scans should use the same device (Messina et al., 2020). An earlier review reported marked systematic differences among Lunar, Hologic and Norland systems (Laskey, 1996). |
| Body size | Whole-body lean mass correlated with MRI skeletal muscle at r = 0.94, with lower correlations (r = 0.77-0.95) reported in markedly obese women (Messina et al., 2020). Precision is best in young healthy subjects and weaker in osteoporotic and obese subjects (Laskey, 1996). |
Where it matters for labs and clinics
For research protocols and clinic programmes, the practical consequence is consistency. Record the device make and model, software version and scan conditions (fasting status, clothing and metal removed) alongside each result, and compare longitudinal values only within one device. Check that the report template includes the ISCD required parameters before data are pooled or exported to a study database. For data-handling context, see the CT scanner explainer for another X-ray based imaging modality, and what is endocrinology for the clinical field that commonly uses these measures.
Common misunderstandings
- DXA is a gold standard. A 1996 review stated that DXA cannot be regarded as a gold standard for body composition (Laskey, 1996); that is dated, but the later review still lists hydration and device dependence as limits.
- Low lean mass means sarcopenia. Appendicular skeletal muscle index cut-offs of below 7.0 kg/m² in men and 6.0 kg/m² in women are cited, but low muscle mass alone does not diagnose sarcopenia (Messina et al., 2020).
- There is an agreed android/gynoid cut-off. The review reports no consensus cut-off for the android/gynoid ratio.
- Results are device-independent. They are not; see the table above.
FAQ
What does a DXA body composition scan measure?
Bone mineral content, fat mass and lean mass for the whole body and regions, estimated from two-energy X-ray attenuation.
Is DXA body composition the same as a bone density test?
No. The same scanner is used, but the body composition report centres on fat and lean mass, whereas bone densitometry reports bone mineral density against reference data.
Can results from different scanners be compared?
Not reliably. Manufacturers may disagree, so repeat measurements should use the same densitometer.
What should a DXA body composition report include?
Per the ISCD 2013 position: BMI, bone mineral density, BMC, total mass, total lean mass, total fat mass and percent fat mass, with several optional indices.
Sources
- Messina C, et al. Body composition with dual energy X-ray absorptiometry: from basics to new tools. Quant Imaging Med Surg. 2020;10(8):1687-1698. PMC7378094 (read 2026-11-11).
- Petak S, Barbu CG, Yu EW, et al. The Official Positions of the International Society for Clinical Densitometry: body composition analysis reporting. J Clin Densitom. 2013;16(4):508-519. PubMed 24183640 (read 2026-11-11).
- Laskey MA. Dual-energy X-ray absorptiometry and body composition. Nutrition. 1996;12(1):45-51. PubMed 8838836 (read 2026-11-11).
