Bone

Part 1 · General anatomy

Living tissue that rebuilds along the lines of force: structure, cells, ossification, blood supply, remodelling and healing

Why bone is the first tissue you learn

Bone is the only tissue in the body that continuously destroys and rebuilds itself in response to the mechanical demands placed on it. That single property is why bone matters to a physiotherapist more than to almost any other clinician: the load you prescribe is a biological signal. A rehabilitation programme is, at the level of bone, a set of instructions to the osteocyte network about how much skeleton the patient is going to need.

Everything in this chapter serves three practical questions:

• What is bone made of, and why does that composition give it the mechanical properties it has?
• How does bone grow, and what does that mean for the injured child versus the injured adult?
• How does bone respond to loading, unloading and fracture — and how do I use that clinically?

Learning outcomes

By the end of this chapter you should be able to:

• State the functions of the skeleton and classify bones by shape, giving examples of each.
• Describe the gross architecture of a long bone, including the periosteum, endosteum and the diaphysis–metaphysis–physis–epiphysis sequence.
• Explain the organic and inorganic composition of bone matrix and relate each component to a specific mechanical property.
• Describe the four bone cell types, their lineages, and the RANK/RANKL/OPG signalling axis.
• Distinguish woven from lamellar bone, and describe the osteon in detail.
• Compare intramembranous and endochondral ossification, and describe the zones of the growth plate.
• Apply the Salter–Harris classification and explain its prognostic meaning.
• Describe the blood supply of a long bone and predict the sites at risk of avascular necrosis.
• Describe the remodelling cycle of the basic multicellular unit and its regulation, mechanical and hormonal.
• State Wolff’s law and Frost’s mechanostat, and use them to justify a loading prescription.
• Describe the stages and timeline of secondary fracture healing, distinguish it from primary healing, and list the factors that impair it.
• Explain the pathophysiology and rehabilitation implications of osteoporosis, bone stress injury and heterotopic ossification.

Functions of the skeleton

Figure 1 · What bone is for

The adult skeleton has 206 bones: 80 axial (skull 22, hyoid 1, auditory ossicles 6, vertebrae 26, sternum 1, ribs 24) and 126 appendicular. The newborn has around 270 ossification-derived elements, which fuse during growth.

Classification of bones

Figure 2 · Classification of bones by shape

2.3.1 By shape

Clinical note. The patella increases the quadriceps moment arm at the knee by roughly 30–50% depending on angle. After patellectomy, quadriceps force must rise by a comparable margin to generate the same extension torque — which is why patellar-tendon and extensor-mechanism injuries dominate the difficulty of knee rehabilitation.

2.3.2 By developmental origin

• Membranous (dermal) bones — form by intramembranous ossification directly within mesenchyme: skull vault, most of the facial skeleton, the clavicle (its medial part), the mandible.
• Cartilaginous (endochondral) bones — form by replacement of a hyaline cartilage model: all long bones, vertebrae, ribs, pelvis, skull base.
• Membrano-cartilaginous — both mechanisms contribute: the clavicle (the only bone with both, and the first to begin ossifying, in week 5–6), occipital, temporal, sphenoid, mandible.

2.3.3 By macroscopic texture

• Compact (cortical) bone — dense, forms the outer shell of every bone and the diaphyseal wall. Approximately 80% of skeletal mass but only ~20% of the surface area, so it turns over slowly (2–3% per year).
• Cancellous (trabecular, spongy) bone — a three-dimensional lattice of trabeculae, found in epiphyses, metaphyses, vertebral bodies, and the diploë. Approximately 20% of mass but 80% of surface area, so it turns over fast (~25% per year) — which is why metabolic bone disease and steroid-induced osteoporosis show themselves first in the vertebral body and femoral neck.

Gross structure of a long bone

Figure 3 · The structure of a long bone

Composition of bone matrix

Bone is a two-phase composite material, and every mechanical property it has is explained by that fact.

The classic demonstration

Decalcify a bone (soak in dilute acid) and the organic matrix remains: the bone becomes rubbery and can be tied in a knot — flexible but unable to bear compression. Deproteinise it (heat or hypochlorite) and the mineral remains: the bone keeps its exact shape but is chalky and shatters under minimal load. The living bone is neither: it is a composite whose toughness exceeds that of either phase alone.

The clinical corollaries follow directly: osteogenesis imperfecta is a type I collagen defect — the organic phase fails, so bones are brittle in a tensile sense and fracture with trivial trauma. Rickets and osteomalacia are mineralisation defects — the organic phase is present but soft, so bones bend (genu varum, coxa vara) rather than snap. Osteoporosis is a loss of both phases in normal proportion — there is simply less bone, of normal quality.

Mechanical behaviour in brief

• Bone is anisotropic: strongest in compression along its long axis, weaker in tension, and weakest in shear. Torsional (shear) loading therefore produces the spiral fracture.
• Bone is viscoelastic: it is stiffer and stronger at high loading rates, and absorbs more energy before failure. High-energy trauma therefore stores more energy, which is released into the surrounding soft tissue — the reason a high-velocity fracture carries a worse soft-tissue prognosis than a low-velocity one of identical pattern.
• The stress–strain curve shows an elastic region (recoverable), a yield point, then a plastic region (permanent microdamage) before failure at roughly 2% strain. Repetitive loading into the plastic region without adequate recovery accumulates microdamage — the mechanism of bone stress injury.

Bone cells

2.6.1 The RANK / RANKL / OPG axis

This is the central control system of bone resorption and the target of modern osteoporosis drugs, so it must be understood rather than memorised.

• RANKL (receptor activator of nuclear factor κB ligand) is expressed by osteoblasts, osteocytes and stromal cells. Binding to RANK on osteoclast precursors, with M-CSF as the permissive signal, drives their fusion, differentiation, activation and survival. RANKL = resorption on.
• OPG (osteoprotegerin) is a soluble decoy receptor, also secreted by osteoblasts. It binds RANKL and prevents it reaching RANK. OPG = resorption off.
• The RANKL:OPG ratio, not either molecule alone, determines net resorption.

This single ratio explains an entire pharmacology and pathology:

Microscopic architecture

Figure 4 · Microscopic structure of compact bone

2.7.1 Woven versus lamellar bone

Woven bone is always the first bone laid down and is always abnormal in the adult skeleton outside healing. It is subsequently remodelled into lamellar bone.

2.7.2 The osteon (Haversian system)

The structural unit of compact bone: a cylinder 200–300 µm across running roughly parallel to the long axis.

• A central Haversian canal carrying one or two capillaries, lymphatics and nerve fibres.
• 4–20 concentric lamellae around it, collagen in each lamella running at an angle to its neighbours — a plywood arrangement that resists crack propagation in all directions.
• Osteocytes in lacunae at lamellar boundaries, linked by canaliculi radiating towards the canal; every osteocyte lies within ~200 µm of a vessel, the diffusion limit.
• A cement line at the periphery — a mineral-rich, collagen-poor boundary that deflects propagating microcracks. This is the microstructural basis of bone’s toughness.
• Volkmann’s (perforating) canals run transversely and obliquely, connecting Haversian canals to each other and to the periosteal and endosteal surfaces. They are not surrounded by concentric lamellae.
• Interstitial lamellae are the remnants of previous osteons, left between current ones — the fossil record of past remodelling.
• Circumferential lamellae run around the whole circumference beneath the periosteum and around the medullary cavity.

Cancellous bone has no osteons in the classical sense: its trabeculae are thin enough (typically <200 µm) to be nourished by diffusion from marrow, and are built of parallel lamellae organised into hemiosteons (trabecular packets).

2.7.3 Trabecular architecture and the stress trajectories

Trabeculae are not randomly arranged. In the proximal femur they form two clearly defined systems — a principal compressive group running from the medial cortex (calcar femorale) up into the head, and a principal tensile group arching from the lateral cortex to the head — crossing at approximately right angles and mirroring the theoretical principal stress trajectories of a loaded cantilever. Between them lies a relatively bone-poor region, Ward’s triangle, which is the first area to rarefy in osteoporosis and a standard DEXA region of interest.

This is the strongest visible evidence for Wolff’s law (§2.11): the internal architecture of bone is an engineered response to habitual loading.

Ossification (osteogenesis)

Figure 5 · Intramembranous and endochondral ossification

Bone forms in only two ways, and every bone in the body uses one or both.

2.8.1 Intramembranous ossification

Direct conversion of mesenchyme into bone, without a cartilage intermediate.

• Mesenchymal cells condense and differentiate into osteoblasts at a primary ossification centre.
• Osteoblasts secrete osteoid, which mineralises; trapped osteoblasts become osteocytes.
• Woven bone spicules form and fuse into trabeculae; the intervening vascular mesenchyme becomes red marrow.
• Surrounding mesenchyme condenses into periosteum; subperiosteal appositional growth lays down a compact outer shell over cancellous interior — the structure of a flat bone.
• Woven bone is progressively remodelled into lamellar bone.

Products: skull vault, most facial bones, mandible, medial clavicle. Because these bones grow by sutural and appositional growth rather than through a physis, cranial vault growth follows brain growth — hence the fontanelles (anterior closing at ~18 months, posterior at ~2–3 months) and the clinical significance of premature sutural fusion (craniosynostosis).

2.8.2 Endochondral ossification

Replacement of a hyaline cartilage model. This is the mechanism for all long bones and is examined constantly.

• Cartilage model forms from condensed mesenchyme in the fetal limb bud.
• Periosteal (bone) collar forms around the mid-diaphysis by intramembranous ossification of the perichondrium — which now becomes periosteum.
• Central chondrocytes hypertrophy, the surrounding matrix calcifies, and the chondrocytes undergo apoptosis, leaving cavities.
• A periosteal bud — capillaries, osteoprogenitors, osteoclasts, haemopoietic cells — invades through the collar, establishing the primary ossification centre (appearing by week 8–12 of fetal life in most long bones).
• Ossification spreads towards both ends; osteoclastic resorption of the central trabeculae creates the medullary cavity.
• Secondary ossification centres appear in the epiphyses, mostly after birth (a few before: the distal femoral and proximal tibial centres are present at term, which is medico-legally used as evidence of full-term birth).
• Cartilage persists in two places only: the articular cartilage (permanently) and the epiphyseal plate (until fusion).
• Growth in length occurs at the physis; growth in width (girth) occurs by subperiosteal appositional deposition with simultaneous endosteal resorption.

2.8.3 Zones of the epiphyseal growth plate

Learn these in order from epiphysis to metaphysis, with the pathology attached to each — this is the highest-yield table in the chapter.

Physeal closure is hormonally driven, principally by oestrogen in both sexes (which is why aromatase deficiency causes continued growth, and why girls, with earlier oestrogen exposure, fuse earlier — typically 14–16 years versus 16–18 in boys). Growth hormone acts via IGF-1 on the proliferative zone; thyroid hormone is required for hypertrophic differentiation.

2.8.4 Salter–Harris classification of physeal injury

Physeal fractures account for roughly 15–20% of paediatric fractures. The classification predicts growth disturbance.

Physiotherapy implications. In a skeletally immature patient, a mechanism that would produce a ligament sprain in an adult far more often produces a physeal injury, because the physis is weaker than the ligament. A child with a “sprained ankle” and tenderness precisely over the distal fibular physis should be treated as a Salter–Harris I until proved otherwise, regardless of normal radiographs. Similarly, apophyseal traction injuries (Osgood–Schlatter at the tibial tuberosity, Sinding-Larsen–Johansson at the inferior patellar pole, Sever’s at the calcaneal apophysis, and avulsions at the ASIS, AIIS and ischial tuberosity) are the adolescent equivalents of adult tendinopathy and require load modification rather than aggressive stretching into the painful apophysis.

Blood supply of a long bone

Figure 6 · Blood supply of a long bone

Four sources, and the way they overlap determines which bones die when their supply is interrupted.

Blood flows centrifugally in the normal diaphysis (medulla → cortex → periosteal capillaries). After the medullary supply is disrupted, flow reverses to centripetal, driven by the periosteum. This is why stripping the periosteum during surgery is far more damaging than reaming the canal, and why periosteal preservation is a principle of modern fracture fixation.

Sites at risk of avascular (osteo)necrosis

Wherever a bone segment has a single, retrograde, intracapsular supply with poor collateral, interruption causes necrosis:

Nerve supply. Periosteum is richly supplied with sensory (nociceptive) fibres — the most pain-sensitive part of bone. Articular branches, medullary nerves accompanying the nutrient artery, and sympathetic vasomotor fibres complete the picture. This explains periosteal pain from a direct blow, the deep boring pain of intraosseous pathology, and the effectiveness of periosteal anaesthesia.

Bone remodelling

Figure 7 · The remodelling cycle and the cells that run it

Bone turnover in the adult is not random; it occurs in discrete, coupled cycles executed by a basic multicellular unit (BMU) — a travelling team of osteoclasts followed by osteoblasts, working within a canal in cortical bone or across a trench on a trabecular surface.

The five phases

Two consequences of this timetable matter clinically:

• Resorption is fast and formation is slow. Any stimulus that increases activation frequency therefore causes transient net bone loss simply by opening more resorption spaces — the “remodelling transient”. This is why bone density can dip in the first months of a new high-turnover state.
• A full cycle takes 4–6 months. Bone adaptation to a training programme cannot be assessed at six weeks. Expect radiographic and densitometric change on a timescale of months to a year.

Regulation

Wolff’s law and the mechanostat

Figure 8 · How bone answers to load

2.11.1 Wolff’s law

Julius Wolff (1892): bone adapts its internal architecture and external form to the loads habitually placed upon it. Increase the load and bone is laid down; remove it and bone is resorbed. The trabecular systems of the femoral neck are the visual proof.

2.11.2 Frost’s mechanostat — the version you can actually prescribe from

Harold Frost reframed Wolff’s law as a negative-feedback control system with strain thresholds. Bone “measures” peak strain (in microstrain, µε) and responds:

Three empirical rules follow, and these are what you build a programme from:

• Magnitude beats duration. A few high-magnitude loading cycles are far more osteogenic than thousands of low-magnitude ones. Ten to fifty impacts is enough; a thousand adds little.
• Rate matters. High strain rate (impact, jumping, change of direction) is more osteogenic than slow loading of the same magnitude.
• Bone desensitises rapidly and re-sensitises with rest. The osteogenic response saturates within about 20–40 loading cycles; separating loading into short bouts with 4–8 hours between them produces more bone than one long session of the same total volume.

2.11.3 Disuse osteopenia

The mechanostat runs in both directions, and the losses are fast and clinically severe.

• Spinal cord injury: bone loss below the lesion of roughly 1–4% per month initially, with 25–50% of trabecular bone lost at the distal femur and proximal tibia within 1–2 years, plateauing at a new steady state. Fragility fracture on transfer or during passive movement is a real and under-recognised risk.
• Spaceflight: approximately 1–1.5% loss per month at weight-bearing sites, despite exercise countermeasures.
• Bed rest and immobilisation: measurable trabecular loss within weeks.
• Recovery of bone is far slower than its loss, and may be incomplete.

The practical translation. For the ambulant patient at risk of osteoporosis, the osteogenic prescription is weight-bearing, impact-containing, and progressively resisted: jumping, hopping, skipping, stair climbing, and heavy-load resistance training (which loads bone through muscle pull, the largest force bone experiences). Swimming and cycling, whatever their cardiovascular merit, are essentially non-osteogenic at the lower limb. For the patient with established vertebral fragility, this is tempered by fracture risk: avoid loaded end-range spinal flexion and rotation, and progress impact cautiously. For the non-ambulant patient, functional electrical stimulation cycling and standing frames produce modest, site-specific effects at best; expectations should be set honestly.

Fracture healing

2.12.1 Secondary (indirect) healing — the default

Occurs when there is relative stability and some interfragmentary motion: cast treatment, intramedullary nailing, bridge plating, external fixation. It proceeds through callus, recapitulating endochondral ossification.

2.12.2 Primary (direct) healing

Occurs only with absolute stability and anatomical reduction — compression plating, lag screws. There is no callus. Cutting cones of osteoclasts cross the fracture line directly and osteons are re-established across it (contact healing where the gap is <0.01 mm; gap healing where it is <1 mm, with lamellar bone deposited first and then remodelled). It is slower to achieve mechanical strength and is unforgiving of any residual motion: interfragmentary movement in a construct designed for primary healing produces non-union rather than callus.

Clinically this matters: the absence of visible callus on a radiograph after rigid internal fixation is expected, not a sign of failure. The appearance of callus around a compression plate, conversely, indicates unwanted motion.

2.12.3 Factors impairing healing

Delayed union = healing slower than expected for that site. Non-union = healing has ceased, conventionally by 6–9 months, with no radiographic progress over 3 consecutive months. Non-unions are hypertrophic (abundant callus, “elephant foot” — good biology, insufficient stability; treat mechanically) or atrophic (no callus — poor biology; treat biologically, with bone graft, and address vascularity and systemic factors). Recognising which type you are looking at explains why one patient is told to load more and another less.

Clinically important bone disorders

Where students consistently go wrong

• Treating bone as inert. It is a metabolically active organ with an endocrine function and a mechanosensory network. Rehabilitation acts on it.
• Confusing osteoblast and osteoclast lineages. Osteoblast = mesenchymal. Osteoclast = haemopoietic (monocyte–macrophage). This is why marrow disease affects resorption.
• Forgetting that the osteocyte is the mechanosensor. The osteoblast builds; the osteocyte decides.
• Confusing the pathologies of the two phases. Loss of collagen quality = brittle (OI). Loss of mineralisation = soft (osteomalacia). Loss of quantity with normal quality = osteoporosis.
• Assuming the epiphysis is the weak point in a child. The hypertrophic zone of the physis is.
• Expecting callus after rigid internal fixation. Absolute stability gives primary healing without callus.
• Prescribing volume when bone needs magnitude and rate. Long slow distance work is poor osteogenic stimulus; a few high-impact loads, distributed across the day, is far better.
• Ignoring the periosteum in surgery-related prognosis. Periosteal stripping compromises healing more than medullary reaming does.
• Stretching aggressively into a hot, firm, restricted post-traumatic joint. Consider heterotopic ossification before you push.
• Assessing bone adaptation too early. A remodelling cycle is 4–6 months.

Check yourself

15 questions on this chapter. Tap one to see the answer and the reasoning.

Quick review

Further reading

Chapter 2 of 24 · Human Anatomy · Physiotherapist India Next: Chapter 3 — Cartilage: no vessels, no nerves, no lymphatics — and everything difficult follows from that.

Reviewed by the Physiotherapist India Team. · Human Anatomy contents