DEXA Scan

DXA (dual-energy X-ray absorptiometry) passes two X-ray beams of different photon energies through the region of interest. Bone and soft tissue attenuate the two energies differently, allowing the machine to subtract soft tissue and compute bone mineral content (g) and areal bone mineral density (g/cm²).

It is a two-dimensional areal measurement, so it is influenced by bone size — small bones read falsely low and large bones falsely high. It does not measure true volumetric density.

Radiation dose is very low (roughly 1–10 microsievert per site), less than one day of natural background radiation and far less than a chest radiograph. Scanning takes 10–20 minutes.

T-score = the number of standard deviations from the mean peak bone mass of a young healthy adult of the same sex. Used in postmenopausal women and men aged ≥50.

Z-score = standard deviations from the mean for age, sex and ethnicity-matched controls. Used in premenopausal women, men under 50, and children. A Z-score ≤ −2.0 is reported as “below the expected range for age”.

WHO categories (T-score): normal ≥ −1.0; osteopenia/low bone mass −1.0 to −2.5; osteoporosis ≤ −2.5; severe or established osteoporosis ≤ −2.5 plus a fragility fracture.

Standard sites: PA lumbar spine (L1–L4) and proximal femur (femoral neck and total hip). The lowest T-score of the valid sites is used for diagnosis.

The 33% (one-third) radius is the accepted third site — used when spine and hip are unmeasurable or invalid: weight above table limit, bilateral hip replacement, or hyperparathyroidism (cortical bone is preferentially lost).

Vertebral exclusion rules: exclude vertebrae that are fractured, have focal structural change, or differ by more than 1.0 T-score from adjacent vertebrae; at least two evaluable vertebrae are required, otherwise the spine is not reportable.

Women ≥65 and men ≥70; younger postmenopausal women and men aged 50–69 with risk factors; any adult with a fragility fracture; glucocorticoid therapy or any disease or drug associated with bone loss; radiographic osteopenia or vertebral deformity; and monitoring of osteoporosis treatment.

Vertebral fracture assessment (VFA) — lateral spine imaging at the same sitting at very low dose, detecting clinically silent vertebral fractures that immediately upgrade the patient to treatment.

Trabecular bone score (TBS) — grey-level texture analysis of the lumbar spine DXA image reflecting trabecular microarchitecture. It predicts fracture independently of BMD and can adjust FRAX probability.

Whole-body DXA measures appendicular lean mass, used to confirm low muscle mass in the diagnosis of sarcopenia (EWGSOP2, AWGS 2019).

Hip structural analysis and hip axis length add geometric information to fracture prediction; DXA also gives body composition and can screen for abdominal aortic calcification.

Degenerative change falsely raises lumbar spine BMD — osteophytes, facet joint sclerosis, vertebral compression fracture, scoliosis and aortic calcification. In older people the hip is the more reliable site.

Artefacts: recent oral or intravenous contrast, radionuclide studies, calcium tablets in the gut, surgical clips, spinal instrumentation, vertebroplasty cement, coins or metal in clothing, and previous fractures.

Technical/patient factors: poor positioning, severe kyphosis or inability to lie flat, tremor or movement, obesity beyond the table limit, and inconsistent analysis between operators.

DXA cannot distinguish osteoporosis from osteomalacia — both show low BMD. Interpretation must always be with biochemistry.

Repeat DXA on the same machine, with the same software and technologist, and compare the absolute BMD in g/cm², not T-scores.

Change is meaningful only if it exceeds the least significant change (LSC), which each centre must calculate — typically about 3% at the lumbar spine and 5% at the hip.

Usual interval is 1–2 years on treatment, longer (2–5 years) if stable or untreated. ISCD 2023 emphasises that the clinical context prompting the repeat should drive the interval rather than a fixed rule.