Metabolic Bone Disease: Imaging Features

Metabolic bone disease encompasses systemic disorders that disturb the normal balance of bone formation, resorption and mineralisation. Because the abnormality is metabolic rather than focal, the imaging signs are usually generalised and bilateral, with characteristic regional accentuation that points to a specific diagnosis. The radiologist's task is to recognise the dominant abnormality — reduced quantity of mineralised bone (osteoporosis), defective mineralisation of osteoid (osteomalacia/rickets), excessive resorption driven by parathyroid hormone (hyperparathyroidism), disordered remodelling (Paget's disease), or the mixed picture of chronic kidney disease (renal osteodystrophy).

A disciplined search pattern converts a vague impression of low density into a specific diagnosis. Work through the DENSITY checklist (above): density, endplates/trabeculae, notches/resorption, soft tissue, insufficiency fractures/Looser zones, tumours, and then yield to biochemistry.

It is not enough to say "Genant grade" in a viva. Grade by approximate height loss against the adjacent vertebrae, and state the morphology separately. The method is visual and semiquantitative by design — that keeps it fast and reproducible, but it retains interobserver variability and, critically, it establishes that a deformity is fracture-shaped, not what caused it. Malignancy, Scheuermann's, and congenital deformity all deform vertebrae; grading does not exclude them.

The number that reframes the whole assessment: most people who fracture do NOT have osteoporosis by T-score. This is the single most counter-intuitive fact in bone densitometry, and it is the reason management moved on. The WHO definition (T-score at or below -2.5) is a diagnostic threshold, not a risk model, and because far more people sit in the osteopenic range than below it, the majority of low-trauma fractures arise in people who would never be labelled osteoporotic on DXA alone. The converse also holds — a young person with an osteoporotic-range T-score and no other risk factors fractures relatively rarely.

What DXA does tell you is gradient, not a cut-off. Fracture risk approximately DOUBLES for every one standard deviation fall in bone density in untreated postmenopausal women. That is why the T-score is a continuous risk variable that happens to have a diagnostic line drawn through it, and why a T-score of -2.4 is not meaningfully safer than -2.6. Quantitative loss must be measured by DXA and never estimated from a radiograph — 30-40% of bone mineral is gone before osteopenia is reliably visible on plain film.

This is what FRAX exists to fix. It integrates age, BMI, prior fracture, parental hip fracture, glucocorticoids, rheumatoid arthritis, smoking and alcohol, with or without femoral neck BMD, to output an absolute 10-year probability of hip and major osteoporotic fracture. The spread it produces is enormous — 10-year hip fracture probability in women ranges roughly 100-fold, from about 0.2% at age 50 with no risk factors to about 22% at age 80 with a parental hip fracture. Absolute probability, not T-score alone, drives intervention thresholds in AAOS, NOGG and most national guidance. Know its blind spots too: FRAX does not capture fall risk, glucocorticoid dose, the number or recency of prior fractures, or lumbar spine BMD, and it is calibrated to country-specific epidemiology — so outputs are not interchangeable between populations.

If you may order one film, order the hand — and here is the evidence for why. When whole-body radiographs were taken in 73 haemodialysis patients with severe secondary hyperparathyroidism, subperiosteal resorption was the commonest abnormality by a wide margin, present at the phalanges and distal clavicles in 94%. Ectopic and vascular calcification followed at 80%, brown tumours at 37%, and rugger jersey spine at only 27% — which is worth holding onto, because rugger jersey spine is taught as a classic and is in fact the least frequent of the four. The highest-yield films were the hands, wrists, lateral thoracolumbar spine and femurs, in that order.

Read those percentages against the population they came from. Every patient in that series had severe, dialysis-dependent secondary disease and was being referred for parathyroidectomy — the extreme end of the spectrum. The frequencies therefore represent a ceiling, not what you should expect in mild secondary or in primary hyperparathyroidism, where the skeleton is now usually radiographically normal because biochemical screening detects the disease long before bone disease develops. A normal hand radiograph does not exclude primary HPT; it is the biochemistry that makes that diagnosis.

The renal osteodystrophy table lists "amyloid deposition (dialysis-related)" as a cause of bone cysts and erosions, but this is a distinct process that deserves developing. After many years of haemodialysis (typically more than 5 to 10), beta-2-microglobulin is inadequately cleared and deposits as amyloid in synovium, tendon and bone — separate from the secondary hyperparathyroidism that dominates renal osteodystrophy.

Osteopetrosis is the sclerosing bone disorder that must be excluded whenever dense vertebral endplate bands raise the question of rugger jersey spine. It is caused by FAILURE of osteoclast-mediated resorption — most often a defect in the osteoclast proton pump or chloride channel that stops the cell acidifying the resorption lacuna — so bone that has been laid down is never taken away.

Dense bone that breaks easily is not a paradox once you know why. Bone strength depends on remodelling, not on mineral content alone. Normal osteoclasts continuously remove fatigue-damaged bone so that osteoblasts can replace it; they also carve the marrow cavity and sculpt the metaphyseal flare. When resorption fails, three things follow directly, and each is visible on the radiograph. Microdamage accumulates and is never repaired, so the bone is dense but structurally exhausted — it snaps transversely, like chalk. The primary spongiosa, calcified cartilage from the growth plate, is never converted into mature lamellar bone, leaving a brittle mineralised scaffold instead of load-bearing trabeculae. And because the marrow cavity is progressively obliterated, the severe infantile forms present with anaemia, thrombocytopenia and extramedullary haematopoiesis rather than with fractures at all. The same failure explains the surgical difficulty: this bone blunts drills and reamers, heals slowly, and osteomyelitis of the mandible is a recognised complication because the sclerotic bone is poorly vascularised.

The classic radiographic signs of metabolic bone disease overlap with focal pathology. The decisive discriminators are distribution (generalised/symmetric vs focal), the accompanying biochemistry, and whether the affected bone is enlarged.

Metabolic bone disease is diagnosed and managed worldwide; imaging interpretation is universal, but access to DXA, scintigraphy and biochemistry varies markedly by resource setting.