8. Joints and Articulations

Structure, Classification, and the Mechanics of Movement

Learning Objectives

8.1 What a Joint Has to Do

A joint, or articulation, is any place where two bones meet — or where bone meets cartilage, as at the costochondral junctions, or where bone meets tooth, as in the socket of a molar. That definition is deliberately loose, because the interesting question is not what counts as a joint but what a joint is for.

Start from the problem. Chapter 7 gave you 206 rigid segments. A rigid segment cannot bend, so a skeleton made of them can only move if it contains discontinuities — and every discontinuity is, by definition, a place where the structure is not continuous and therefore not as strong. Joints are the price of movement. The whole of joint anatomy is a set of engineering answers to one question: how do you interrupt a load-bearing structure without losing the ability to bear load?

Write the specification out properly, because everything in this chapter is a response to one of its four lines.

One — transmit load across a gap. The hip carries three to five times body weight during walking and up to eight during stair descent, across a contact area of a few square centimetres. Peak cartilage contact stresses of 3–5 MPa in level walking and up to 10–18 MPa in demanding activities are ordinary. Nothing about a gap makes that easy.

Two — permit a defined set of movements and forbid all the others. This is the half students underrate. A joint that permitted everything would be useless: your knee is valuable precisely because it refuses to abduct. Most of the tissue in a joint is there to say no.

Three — do it for about 10⁸ cycles. A person walking 6,000 steps a day flexes each knee roughly two million times a decade, and the joint is expected to last eighty years without service.

Four — and repair essentially nothing. This is the line that makes joints different from every other structure in the book, and it is where the chapter ends (§8.9).

Against that specification, the performance is remarkable. A healthy synovial joint runs at a coefficient of friction of about 0.003–0.02 — compared with roughly 0.03 for ice sliding on ice, and 0.05–0.1 for a well-oiled steel bearing. Your knee is, in the most literal engineering sense, more slippery than ice, and it achieves that with a fluid you could hold in a teaspoon.

8.2 Two Classifications, One Trade-Off

There are two systems for classifying joints and you need both, because they answer different questions.

• Structural classification asks what material binds the bones, and is there a cavity between them? Three answers: fibrous, cartilaginous, synovial.
• Functional classification asks how much movement does it permit? Three answers: synarthrosis (immovable), amphiarthrosis (slightly movable), diarthrosis (freely movable).

The two correlate but do not map one-to-one, and the mismatches are where the interesting anatomy lives.

                        HOW TO CLASSIFY ANY JOINT
                                    │
            ┌───────────────────────┴───────────────────────┐
     STRUCTURAL — what binds it?              FUNCTIONAL — how much movement?
            │                                               │
  ┌─────────┼──────────┬──────────────┐        ┌────────────┼────────────┐
FIBROUS  CARTILAGINOUS  SYNOVIAL              SYN-       AMPHI-      DI-
dense CT   cartilage    CAVITY +              ARTHROSIS  ARTHROSIS   ARTHROSIS
NO cavity  NO cavity    fluid                 immovable  slight      free
   │            │           │                     │          │          │
   │            │           └─────────────────────┼──────────┴──────────┤
   │            │                                 │                     │
   │            │            (all synovial joints are diarthroses ──────┘
   │            │             — this is the one clean 1:1 mapping)
   │            │
   │            ├── SYNCHONDROSIS ── hyaline ──────► SYNARTHROSIS
   │            │   epiphyseal plate; 1st sterno-     (temporary or
   │            │   costal joint; sphenooccipital      immobile)
   │            │
   │            └── SYMPHYSIS ────── fibrocartilage ► AMPHIARTHROSIS
   │                pubic symphysis; IV discs;         (always in the
   │                manubriosternal joint               MEDIAN PLANE)
   │
   ├── SUTURE ─────────── short fibres ──────────────► SYNARTHROSIS
   │   coronal, sagittal, lambdoid, squamous            (ossifies with
   │                                                     age → SYNOSTOSIS)
   ├── SYNDESMOSIS ────── ligament / membrane ────────► depends on FIBRE LENGTH
   │   distal tibiofibular (short fibres) ─────────────► synarthrosis
   │   radioulnar interosseous membrane (long) ───────► amphiarthrosis
   │
   └── GOMPHOSIS ──────── periodontal ligament ──────► SYNARTHROSIS
       tooth in alveolus — the ONLY joint in the body
       in which bone articulates with something that
       is not bone

  ═══════════════════════════════════════════════════════════════════════════
   THE RULE UNDERNEATH BOTH SYSTEMS
      more material between the bones  →  less movement  →  less risk
      MORE SPACE between the bones     →  more movement  →  MORE RISK
   Structure and function are two readings of the same measurement.

Fibrous joints — bone to bone by collagen, no cavity

Two things are worth more than the table.

Sutures are joints designed to disappear. Their interlocking margins are the growing edges of intramembranous bone (§7.2), and once growth stops they progressively ossify. By the eighth decade many are obliterated. A joint whose long-term plan is to stop being a joint is a strong hint that its function was never adult movement.

Syndesmoses are the clearest demonstration in the body that function follows a single parameter. Fibre length. The distal tibiofibular syndesmosis has short, tightly packed fibres and permits 1–2 mm of movement — enough for the ankle mortise to widen slightly as the wider front of the talus enters it in dorsiflexion, and no more. The interosseous membrane of the forearm is the same tissue with fibres several centimetres long, and it permits the 150° of pronation and supination that make the hand a manipulator. Same histology, same structural class, opposite functional class, and the only variable is how long the collagen is.

Cartilaginous joints — bone to bone by cartilage, no cavity

The distinction is functional, not merely histological. Synchondroses are temporary or immobile. The epiphyseal plate is a joint whose entire purpose is to be destroyed on schedule; the first sternocostal joint is immobile because rib 1 must be a fixed anchor (§7.4). Symphyses are permanent, midline, and built to deform: fibrocartilage takes compression like cartilage and tension like tendon, so a symphysis absorbs shock while resisting separation.

8.3 The Synovial Joint in Detail

Every synovial joint in the body — from the tiny articulations between the auditory ossicles to the hip — has the same six components. Learn them once and you have learned every diarthrosis you will ever be asked about.

        A SYNOVIAL JOINT IN SECTION — the six components, plus accessories

            ┌───────────────────────────────────────────┐
            │                 BONE  1                   │
            │        (periosteum continuous with ───────┼──┐
            │         the fibrous capsule)              │  │
            └────────────┬──────────────────┬───────────┘  │
                         │                  │              │
     ①  ARTICULAR ══════▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓              │
        CARTILAGE        ░░░░░░░░░░░░░░░░░░░░  2–4 mm      │
        hyaline          ↑                                 │
        avascular        │  ②  ARTICULAR CAVITY            │
        aneural          │      0.5–4 mL of fluid          │
        alymphatic       │      a POTENTIAL space          │
                         ▼                                 │
     ①  ARTICULAR ══════▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓               │
        CARTILAGE        ░░░░░░░░░░░░░░░░░░░░               │
            ┌────────────┴──────────────────┴───────────┐  │
            │                 BONE  2                   │  │
            └───────────────────────────────────────────┘  │
                                                           │
   ③  ARTICULAR CAPSULE — two layers ──────────────────────┘
        OUTER  fibrous layer   dense irregular CT; the mechanical restraint;
                               continuous with periosteum of both bones
        INNER  SYNOVIAL MEMBRANE  1–3 cells thick; NOT an epithelium — no
                               basement membrane; type A (macrophage-like)
                               + type B (fibroblast-like) synoviocytes;
                               FENESTRATED capillaries immediately beneath

   ④  SYNOVIAL FLUID  plasma dialysate + hyaluronan (3–4 mg/mL) + lubricin
        viscous · shear-thinning · thixotropic · FIBRINOLYTIC
        jobs: LUBRICATE · ABSORB SHOCK · FEED THE CARTILAGE (nothing else can)

   ⑤  REINFORCING LIGAMENTS
        CAPSULAR ───── thickenings of the capsule itself (iliofemoral)
        EXTRACAPSULAR ─ outside the capsule (fibular collateral of the knee)
        INTRACAPSULAR ─ INSIDE the capsule (the cruciates)  ← remember this one
                        — it is the answer to Case File question 3

   ⑥  NERVES + VESSELS
        nerves → richly supply CAPSULE and LIGAMENTS: pain + PROPRIOCEPTION
        vessels → supply synovial membrane and capsule
                  ✗ NOT the articular cartilage.  Not ever.

  ═══════════════════════════════════════════════════════════════════════════
   ACCESSORY STRUCTURES (present at some joints only)
      MENISCI / ARTICULAR DISCS   fibrocartilage wedges; improve the fit of
                                  mismatched surfaces, spread load, absorb shock
      FAT PADS                    fill dead space, deform to follow the joint
      BURSAE                      flat synovial-lined sacs wherever two things rub
      TENDON SHEATHS              a bursa rolled into a tube around a tendon

① Articular cartilage

A layer of hyaline cartilage 2–4 mm thick in most joints, reaching 5–7 mm on the posterior surface of the patella, which faces the highest contact stresses in the body. It has no perichondrium, because a perichondrium would be worn off within days. Its composition is worth committing to memory, because every one of its mechanical properties follows from it:

The mechanics are those of a biphasic material — a stiff porous solid saturated with fluid. Aggrecan's fixed negative charges attract counter-ions, which draw water in, which pressurizes the tissue from within; the collagen mesh resists that swelling, so the unloaded tissue sits permanently pre-tensioned, like an inflated tyre. Load it and fluid cannot escape quickly, because the pore size in the matrix is on the order of nanometres. So the fluid takes the load — interstitial fluid pressurization carries up to 90% of the applied stress in the first seconds of loading — and the solid matrix is spared. Only under sustained load does fluid slowly exude and the solid matrix begin to carry the stress. This is why cartilage tolerates a footstep beautifully and standing still for three hours much less well.

That single fact also explains the tissue's most infamous property. It is avascular, aneural, and alymphatic. There is not a capillary in it. Chondrocytes are fed by diffusion from synovial fluid, over a working distance of roughly 100–200 µm, assisted by convection — fluid squeezed out under load and drawn back in on unloading. Articular cartilage is, quite literally, fed by being used, which is why immobilizing a joint thins its cartilage and why "rest it completely" is bad advice for almost every joint problem.

② The articular cavity

A potential space, not an actual one, exactly like the pleural and pericardial cavities of Chapter 1. A resting adult knee — the largest joint cavity in the body — holds perhaps 0.5–4 mL of fluid, a film thin enough that the opposing cartilage surfaces are essentially in contact. The clinical logic of a potential space is always the same: it is unremarkable until something fills it, and then the volume tells you what happened. Toby's 55 mL is between fifteen and a hundred times normal.

③ The articular capsule

Two layers, with two entirely different jobs.

The outer fibrous layer is dense irregular connective tissue, continuous with the periosteum of both bones. It is the mechanical restraint — a sleeve strong enough that in many joints it is the primary limit to extreme movement — and it is where most of the capsule's sensory innervation lives. Its thickness varies enormously and purposefully: at the hip it is thick and reinforced everywhere, while at the shoulder it is thin and redundantly folded inferiorly so that the arm can be raised.

The inner synovial membrane is not an epithelium at all, and this is a favourite examination point. It has no basement membrane and no tight junctions; it is modified areolar connective tissue whose surface cells simply lie side by side, with gaps between them.

④ Synovial fluid

A plasma dialysate with two molecules added by the type B synoviocytes.

Hyaluronan — an unbranched glycosaminoglycan of enormous molecular weight (6–7 million daltons) at a concentration of 3–4 mg/mL. It is what makes the fluid viscous and gives it its name: synovia, from Latin, meaning "with egg," because it looks and behaves like egg white.

Lubricin (proteoglycan 4, PRG4) — a mucinous glycoprotein that adsorbs onto the cartilage surface and provides boundary lubrication. Mutations in the PRG4 gene cause camptodactyly–arthropathy–coxa vara–pericarditis syndrome, in which joints fail early — a natural experiment proving that one molecule does most of the low-speed lubrication.

Non-Newtonian behaviour. A Newtonian fluid — water, for instance — has a viscosity that does not depend on how fast you shear it. Synovial fluid is not Newtonian in two separate ways.

• It is shear-thinning (pseudoplastic): viscosity falls as shear rate rises, because the long hyaluronan chains align with the flow instead of tangling across it. Viscosity can fall by more than two orders of magnitude between slow and fast movement.
• It is thixotropic: viscosity also falls with time under shear and recovers over minutes at rest.

Both properties point the same way and the design intent is obvious once you state it. At rest, high viscosity keeps the fluid where it is put and resists being squeezed out from between loaded surfaces, preserving the lubricating film during long static loading — standing, or sleeping. In motion, low viscosity means the joint does not have to work against its own lubricant. The system gets a durable film and cheap movement, which a Newtonian fluid could not deliver.

The symptom is one everybody has: morning stiffness that eases within a few minutes of moving. That is thixotropic recovery being undone. A few minutes is normal. Stiffness lasting more than about an hour is a red flag for inflammatory arthritis (§8.8), and the difference between "a few minutes" and "more than an hour" is one of the most useful questions in rheumatology.

⑤ Reinforcing ligaments

Three positions, and the position is what matters clinically.

• Capsular ligaments are thickenings of the fibrous capsule itself, not separate structures — the iliofemoral ligament of the hip, the glenohumeral ligaments.
• Extracapsular ligaments lie outside the capsule — the fibular (lateral) collateral ligament of the knee is the standard example, running as a free cord separated from the capsule by fat.
• Intracapsular ligaments lie inside the capsule — the cruciates of the knee, the ligament of the head of the femur.

Note the odd arrangement of the cruciates specifically: they are intracapsular but extrasynovial. They lie within the fibrous capsule, but the synovial membrane folds backward around them from behind, so that they are technically outside the synovial cavity while being completely surrounded by synovial fluid. Hold that sentence. It becomes decisive in §8.9, and it is the reason a structure can be simultaneously "not in the joint fluid" anatomically and completely at its mercy biologically.

⑥ Nerves and vessels

Nerves. The capsule and ligaments are richly innervated; the cartilage is not innervated at all. Four classes of articular receptor are described: type I Ruffini endings in the superficial capsule, slowly adapting, signalling static joint position; type II Pacinian corpuscles deeper in the capsule and in fat pads, rapidly adapting, signalling acceleration and deceleration; type III Golgi-like endings in the ligaments, very high threshold, signalling tension near end-range; and type IV free nerve endings everywhere, signalling pain.

Hilton's law is the organizing principle and it is worth knowing by name: the nerve supplying a muscle that crosses a joint also supplies the joint and the skin over its distal attachment. That is why hip pathology is felt in the knee (both are supplied by branches of the femoral and obturator nerves), and it is one of the commonest sources of diagnostic error in a limping child.

Vessels. A periarticular arterial anastomosis supplies the capsule and synovial membrane richly — the genicular anastomosis around the knee, the cruciate anastomosis around the hip — and vessels penetrate as far as the peripheral menisci and the synovial folds. They stop at the cartilage. Figure 8.6 in §8.9 maps this exactly, because the map is the prognosis.

Accessory structures: menisci, fat pads, bursae, sheaths

Menisci and articular discs are fibrocartilage wedges that improve congruence between mismatched surfaces. A meniscus is C-shaped and incomplete (the knee); a disc is complete and divides the cavity in two (the temporomandibular joint, the sternoclavicular joint, the distal radioulnar joint). Their functions are load distribution — the knee menisci spread load over roughly three times the contact area — shock absorption, and secondary stabilization.

Bursae are flattened, synovial-lined sacs containing a film of synovial fluid, positioned wherever a tendon, ligament, muscle, skin, or bone would otherwise rub. There are roughly 160 in the body and they are named for what they lie between: subcutaneous between skin and bone, as at the prepatellar and olecranon bursae; subtendinous between tendon and bone, as at the subacromial and retrocalcaneal bursae; submuscular between muscle and bone, as at the greater trochanter; and subfascial between fascia and bone.

A tendon sheath is simply a bursa rolled into a tube around a tendon, found where tendons pass through tight fibro-osseous tunnels — the flexor tendons of the fingers, the tendons crossing the wrist. Inflame a bursa and you have bursitis; inflame a sheath and you have tenosynovitis; inflame the joint lining itself and you have synovitis. The same word turns up in three diagnoses because the same tissue is in three places.

8.4 The Six Shape Classes and Degrees of Freedom

Synovial joints are subclassified by the shape of their articular surfaces, and shape sets a hard ceiling on what the joint can do. The measure is degrees of freedom: the number of independent axes about which one bone can rotate on the other.

     THE SIX SHAPE CLASSES OF SYNOVIAL JOINT AND THEIR DEGREES OF FREEDOM

  ┌───────────────┬────────────────────┬───────────┬──────────────────────┐
  │ CLASS         │ SURFACES           │ DEGREES   │ EXAMPLES / MOVEMENTS │
  │               │                    │ OF FREEDOM│                      │
  ├───────────────┼────────────────────┼───────────┼──────────────────────┤
  │ PLANE         │ flat on flat       │ NON-AXIAL │ intercarpal, inter-  │
  │  ▭ ── ▭       │                    │ (gliding  │ tarsal, facet joints,│
  │               │                    │  only)    │ acromioclavicular    │
  ├───────────────┼────────────────────┼───────────┼──────────────────────┤
  │ HINGE         │ cylinder in a      │ 1         │ elbow (humeroulnar), │
  │  ╭─╮          │ trough             │ UNIAXIAL  │ interphalangeal,     │
  │  ╰─╯──▭       │                    │           │ ankle (talocrural)   │
  ├───────────────┼────────────────────┼───────────┼──────────────────────┤
  │ PIVOT         │ rounded end inside │ 1         │ atlantoaxial (dens), │
  │  ●            │ a ring of bone +   │ UNIAXIAL  │ proximal radioulnar —│
  │ (○)           │ ligament           │           │ rotation about a     │
  │               │                    │           │ LONG axis            │
  ├───────────────┼────────────────────┼───────────┼──────────────────────┤
  │ CONDYLAR      │ oval convex into   │ 2         │ metacarpophalangeal, │
  │  ⬭            │ oval concave       │ BIAXIAL   │ wrist (radiocarpal), │
  │  ╰─╯          │                    │           │ knee (bicondylar) —  │
  │               │                    │           │ flex/ext + abd/add   │
  │               │                    │           │ → CIRCUMDUCTION      │
  ├───────────────┼────────────────────┼───────────┼──────────────────────┤
  │ SADDLE        │ each surface is    │ 2         │ 1st carpometacarpal  │
  │  ╭╮  ╰╯       │ concave one way,   │ BIAXIAL   │ (thumb base) and     │
  │  ╰╯  ╭╮       │ convex the other;  │ + some    │ sternoclavicular.    │
  │               │ they nest at 90°   │ axial spin│ Enables OPPOSITION.  │
  ├───────────────┼────────────────────┼───────────┼──────────────────────┤
  │ BALL-AND-     │ sphere in a cup    │ 3         │ shoulder (shallow cup│
  │ SOCKET        │                    │ MULTI-    │ → huge range, low    │
  │  ●            │                    │ AXIAL     │ stability); hip (deep│
  │ (◡)           │                    │           │ cup → less range,    │
  │               │                    │           │ high stability)      │
  └───────────────┴────────────────────┴───────────┴──────────────────────┘

   DEGREES OF FREEDOM are set by SHAPE and are a HARD CEILING.  No amount of
   ligament laxity converts a hinge into a ball-and-socket.  Everything past
   the ceiling is not motion — it is INJURY.  ("Hyperextension" of a knee
   past 0–5° is the anterior cruciate ligament being asked to be a bone.)

  ═══════════════════════════════════════════════════════════════════════════
   ARTHROKINEMATICS — what the surfaces do, as opposed to what the bone does

     ROLL    new points on surface A meet new points on surface B
             (a tyre rolling along a road)
     GLIDE   the SAME point on A meets new points on B  (a skidding tyre)
     SPIN    rotation about a fixed point of contact

   CONVEX-ON-CONCAVE  →  roll and glide go in OPPOSITE directions
     (humeral head rolls up, glides DOWN, as the arm abducts — which is
      the only reason the head does not smash into the acromion)
   CONCAVE-ON-CONVEX  →  roll and glide go in the SAME direction
     (tibial plateau on femoral condyles in an open-chain knee extension)

Working through the classes

Plane (gliding) joints. Flat or very slightly curved surfaces that slide across one another without any angular change. The intercarpal, intertarsal, and vertebral facet joints, plus the acromioclavicular and sternocostal joints. They look trivial and are not: a few millimetres of glide at each of the facet joints, summed over the whole lumbar spine, is most of what allows you to bend forward, and a few millimetres at each intercarpal joint is most of what allows a wrist to follow a hand.

Hinge (ginglymus) joints. A cylindrical surface in a trough, permitting flexion and extension about a single transverse axis. The humeroulnar elbow is the purest example — the trochlear notch of the ulna wraps the trochlea of the humerus through nearly 180° of arc, which is why the elbow has genuine bony stability, unusual for a limb joint. The interphalangeal joints and the talocrural ankle are the others. Note the ankle carefully: the mortise (§7.6) permits dorsiflexion and plantar flexion and nothing else, which is why inversion and eversion are not ankle movements (§8.6).

Pivot (trochoid) joints. A rounded process rotating within a ring formed partly of bone and partly of ligament — one degree of freedom, but rotation about a long axis rather than a transverse one. The atlantoaxial joint, where the atlas and skull rotate around the dens of the axis, supplies about half of all cervical rotation. The proximal radioulnar joint, where the disc- shaped radial head spins within the anular ligament, supplies pronation and supination together with its distal partner.

Condylar (ellipsoid) joints. An oval convex surface in an oval concavity: two degrees of freedom, flexion–extension and abduction–adduction, and therefore circumduction. The metacarpophalangeal joints are the model — you can spread your fingers and bend them, but you cannot voluntarily rotate one about its own axis. The radiocarpal wrist is condylar. So, by most classifications, is the knee, which is a bicondylar joint with two side-by-side condyles and therefore two degrees of freedom, even though its second degree — axial rotation — is only available in flexion. The knee is often called a modified hinge for teaching convenience; that description is useful and slightly false, and §8.7 explains the difference the second degree of freedom makes.

Saddle (sellar) joints. Each surface is concave in one direction and convex in the other, and the two nest at right angles like a rider in a saddle. Two degrees of freedom plus a limited conjunct rotation. The first carpometacarpal joint — trapezium against the first metacarpal — is the one that matters, because it is what makes opposition possible and therefore what makes the human hand a human hand. The sternoclavicular joint is the other, and its saddle geometry plus an intra-articular disc is why the only bony link between arm and trunk (§7.5) can move in three directions without dislocating.

Ball-and-socket (spheroidal) joints. A hemispherical head in a cup: three degrees of freedom. Only two exist in the body, the shoulder and the hip, and they sit at opposite ends of the stability–mobility axis for the reason established in §8.1 — socket depth.

Degrees of freedom are a ceiling, not a target

The most useful thing about the shape classes is what they forbid. A hinge joint has one degree of freedom because its surfaces physically cannot travel in any other direction while remaining in contact. No amount of ligament laxity converts a hinge into a ball-and-socket. So when a joint moves in a direction its shape does not permit, the movement is not motion — it is injury, and the tissue absorbing it is the tissue that was forbidding it. Knee "hyperextension" past about 5° is the anterior cruciate ligament being asked to do the job of a bone. Elbow hyperextension past 0–10° is the olecranon being levered against its fossa. Valgus at the knee is the medial collateral ligament, and beyond it the ACL.

That is a diagnostic tool. Ask what direction the joint was forced in, look up which structure forbids that direction, and you have a differential diagnosis before you touch the patient.

Close-packed and loose-packed positions

One further property of shape, and it explains a great deal of clinical practice.

Every synovial joint has a single close-packed position in which the articular surfaces are maximally congruent, the capsule and ligaments are maximally taut, the joint surfaces cannot be separated by traction, and the two bones behave temporarily as one. Every other position is loose-packed (open-packed), with the surfaces less congruent, the capsule slack, and some accessory movement available.

Three consequences. Injuries cluster in close-packed positions, because a joint locked into congruence transmits force to bone and ligament rather than absorbing it by moving — which is why the ankle fractures rather than sprains when it is loaded in dorsiflexion, and sprains rather than fractures in plantar flexion. Joints are immobilized in loose-packed positions when the aim is comfort and in close-packed positions when the aim is preventing contracture — hence the "position of safe immobilization" for a hand, with the metacarpophalangeal joints flexed 70–90° (near close-packed, so their collateral ligaments are long) and the interphalangeal joints nearly straight. Splint a hand the intuitive way, flat, and the MCP collaterals shorten and the hand never opens properly again. And swollen joints adopt the loose-packed position spontaneously, because that is the position of maximum capsular volume and therefore lowest intra-articular pressure. Toby's knee resting at 10–75° rather than straight is not stiffness. It is 55 mL of fluid finding the position of greatest capacity, and it is why his measured range is restricted at both ends.

Arthrokinematics: roll, glide, and spin

The movement you can see — the tibia swinging on the femur — is osteokinematics. What the surfaces are doing to each other is arthrokinematics, and there are only three possibilities.

• Roll. New points on one surface meet new points on the other, like a tyre on a road.
• Glide (slide). The same point on the moving surface sweeps across new points on the stationary one, like a skidding tyre.
• Spin. Rotation about a single fixed point of contact.

Pure rolling would be a disaster in a joint, because the moving bone would roll straight off the end of the stationary surface within a few degrees. Every real joint therefore combines roll with a simultaneous glide in a direction that keeps the surfaces centred — and the direction is given by the convex–concave rule:

• A convex surface moving on a concave one rolls and glides in opposite directions.
• A concave surface moving on a convex one rolls and glides in the same direction.

The shoulder is the demonstration. As the arm abducts, the convex humeral head rolls superiorly on the concave glenoid — and if that were all, it would travel into the acromion within about 20° of abduction. It does not, because it simultaneously glides inferiorly. The rotator cuff, and specifically supraspinatus and the depressors, produce that glide. Lose it, as in cuff pathology, and the head migrates upward on the film — superior migration of the humeral head is a radiographic sign of a massive cuff tear, and it is the convex–concave rule failing in a living patient.

This is not an abstraction for physical therapy students. Restoring accessory glide is what joint mobilization techniques do, and the direction of the mobilization is chosen from this rule.

8.5 What Actually Holds a Joint Together

Three factors stabilize a synovial joint, and students almost always rank them wrongly. Rank them correctly and a great deal of clinical reasoning falls out.

Factor 1 — the shape of the articular surfaces

Real, occasionally dominant, and completely outside your control.

A deep socket helps enormously. The acetabulum encloses more than half the femoral head and is deepened further by a fibrocartilaginous labrum that adds roughly 10% to the coverage and, by sealing against the head, maintains a negative intra-articular pressure. That seal is not a metaphor: the hip's negative pressure contributes measurable suspension, and if you vent a cadaveric hip capsule with a needle the head visibly drops away from the socket under the weight of the limb alone. Atmospheric pressure is one of the hip's stabilizers.

A shallow socket contributes almost nothing. The glenoid holds about a third of the humeral head, and even with its labrum the concavity is slight.

The clinical consequence is stark: the shoulder accounts for roughly half of all major joint dislocations, and dislocating a hip requires an axial force through a flexed femur — a dashboard in a head-on collision. Same joint class, same tissues, thirty-fold difference in incidence, from geometry.

Factor 2 — ligaments

Important, and the less the shape helps, the more the ligaments are asked to do. But ligaments are a last defence, not a first one, and the reason is mechanical.

A ligament is dense regular connective tissue: parallel bundles of type I collagen with fibroblasts strung between them, and — this is the key detail — the collagen is not straight at rest. It is crimped, folded in a regular zig-zag. Load a ligament and its stress–strain curve has a characteristic shape:

Read that table as a timeline and the problem becomes obvious. Through the whole toe region — the first few per cent of stretch, which corresponds to the joint travelling toward the end of its normal range — the ligament is contributing almost nothing. It begins to resist only once the joint is already near its limit, and it sustains damage only a couple of per cent of strain later. The window between "starts helping" and "starts tearing" is narrow.

Two further properties make it worse. Ligaments are viscoelastic: held under constant load they creep, lengthening over minutes to hours, and held at constant length they show stress relaxation, with tension falling. And they do not shorten back. A stretched ligament stays stretched, which is why every dislocation makes the next one more likely and why recurrence after a first traumatic shoulder dislocation before the age of 20 exceeds 70%.

Factor 3 — muscle tone

The most important stabilizer at most joints, and the only one you can train.

Tendons crossing a joint act as active, adjustable, anticipatory restraints, and each of those three words is doing work.

Active: a muscle generates force on demand rather than only when stretched past a threshold. Adjustable: the amount of force is set by the nervous system according to what it expects to happen next. Anticipatory: this is the decisive one. A muscle spindle detects the beginning of an unwanted movement, and the reflex increase in tension arrives in roughly 30–50 ms — before the ligament has taken any appreciable load at all. And feedforward activation is faster still, because it precedes the event: a trained athlete pre-tensions the hamstrings before the foot contacts the ground, so the restraint is already in place when the load arrives.

Compare the numbers. Ligament: engages at end-range, damaged 2% of strain later, does not recover. Muscle: engages 30–50 ms after perturbation onset or before it, adjustable in magnitude, and strengthens with use. It is not a close contest.

Ranking the three, with evidence

The ranking is not fixed across all joints — at the hip, shape beats muscle; at the interphalangeal joints, shape and collateral ligaments do nearly all of it — but for the joints that actually get injured in life, muscle is first.

8.6 Movements at Synovial Joints

Movement terminology is the anatomical language of Chapter 1 applied to motion, and it obeys the same rule: everything is defined from the anatomical position. A movement is named for the direction the distal bone travels away from that starting posture, not from wherever the limb happens to be at the time. Get that rule wrong and half the vocabulary stops making sense.

Four categories: gliding, angular, rotational, and a set of specials.

   ANGULAR MOVEMENTS — the angle between two bones changes

     FLEXION / EXTENSION            ABDUCTION / ADDUCTION
     (sagittal plane, about a       (frontal plane, about an
      frontal axis)                  anteroposterior axis)

       ╲                                    │
        ╲  flexion  ← angle                 │  ab = AWAY from midline
      ───●───  DECREASES                ────●────  (think "abduct" =
        ╱   extension → angle               │╲      to carry away)
       ╱     INCREASES                      │ ╲   ad = TOWARD midline
                                            │  ╲
     HYPEREXTENSION = extension beyond    For FINGERS the reference midline
     the anatomical position (normal at   is the 3rd digit; for TOES it is
     the shoulder, neck, hip; abnormal    the 2nd.  Spreading fingers = ABduct.
     past ~0–5° at the knee and elbow —
     at the knee that is the ACL          CIRCUMDUCTION = flexion + abduction
     stopping the tibia sliding forward)  + extension + adduction in sequence
                                          → the limb sweeps a CONE.  Not a
                                          separate movement; four movements
                                          blended, so it REQUIRES a biaxial or
                                          multiaxial joint.  A finger
                                          circumducts (condylar); an elbow
                                          cannot (hinge).

   ROTATION — a bone turns about its OWN long axis
     medial (internal) rotation = anterior surface turns toward the midline
     lateral (external) rotation = anterior surface turns away
     Only at multiaxial and pivot joints: shoulder, hip, atlantoaxial, and
     the vertebral column as a whole.

   GLIDING (translation) — flat surfaces slide, no angular change
     intercarpals, intertarsals, vertebral facets, acromioclavicular.
     Humble, and it is most of what lets you bend forward.

  ═══════════════════════════════════════════════════════════════════════════
   SPECIAL MOVEMENTS — named because they happen at only one or two places

   SUPINATION / PRONATION   forearm.  Supinated = palm anterior (anatomical
     ╲│╱   ╱│╲              position; radius parallel to ulna).  Pronated =
      ●     ●               palm posterior; the radius CROSSES the ulna.
     "soup"  "pour"         ~150° total.  Mnemonic: you carry SOUP supinated.

   DORSIFLEXION / PLANTAR FLEXION   ankle.  Dorsiflex = toes toward shin
      ╱                             (standing on heels), 0–20°.  Plantar flex
     ●───   dorsiflexion            = toes away (tiptoe), 0–50°.  These are
      ╲                             the ONLY movements the mortise permits.
     ●───   plantar flexion

   INVERSION / EVERSION    at the SUBTALAR and transverse tarsal joints,
     the sole turns MEDIALLY (inversion, 0–35°) or LATERALLY (eversion,
     0–15°).  NOT the ankle.  Inversion + plantar flexion = the sprain
     position, and the anterior talofibular ligament is what tears.

   PROTRACTION / RETRACTION   anterior / posterior movement in the
     transverse plane.  Jut the jaw forward; round the shoulders forward.

   ELEVATION / DEPRESSION   superior / inferior movement.  Shrug (elevate
     the scapula); close the mouth (elevate the mandible).

   OPPOSITION   thumb tip to the tip of any other digit.  Requires the
     SADDLE joint at the 1st carpometacarpal (§8.4).  It is what makes a
     precision grip possible; losing it costs ~40–50% of hand function.

The four traps

One. Gliding is a genuine fourth category and the humblest one: flat surfaces sliding with no angular change. It looks trivial and is not. Several millimetres of glide at each facet joint, summed over twenty-four vertebral levels, is most of what lets you bend forward.

Two. Flexion and extension reverse at the knee relative to intuition. Knee flexion moves the leg posteriorly; hip flexion moves the thigh anteriorly. Both decrease the angle at the joint, which is what the definition actually says. Trust the definition, not the direction. The same trap catches people at the shoulder and the ankle, where dorsiflexion is anatomically "extension" of the foot and is given its own name precisely to avoid the argument.

Three. Inversion and eversion are not ankle movements. The ankle mortise is a hinge and can only dorsiflex and plantar flex (§7.6). Turning the sole inward happens below the ankle, at the subtalar and transverse tarsal joints. Getting this right is the difference between describing an injury and locating it: "he inverted his ankle" means the talus stayed in the mortise and the lateral ligaments took the load.

Four. Circumduction is not rotation. Rotation spins a bone about its own axis; circumduction sweeps it around a cone while its own axis stays put. You can circumduct a finger, which cannot rotate at all.

How much movement is normal

Naming a movement is half the skill. The other half is knowing how much of it there should be, because "reduced range" is meaningless without a reference. These are American Academy of Orthopaedic Surgeons reference values for adults, measured from the anatomical position (0°), and they vary by 10° or so between individuals and by more with age.

Two things are worth noticing in that table. The lumbar rotation value of about 5° is the facet-orientation argument from §7.3 stated numerically: sagittal rails forbid rotation, and the number is not "small," it is essentially zero. And knee flexion exceeds 135° only passively, because active flexion is limited by the bulk of the calf against the thigh — a soft tissue approximation, which is a different kind of limit from the ligamentous one that stops extension.

Goniometry and end-feel

A goniometer is two arms and a protractor, and using one correctly is a matter of three decisions: where the axis goes, where the stationary arm goes, and where the moving arm goes. For knee flexion, for example, the axis sits over the lateral femoral epicondyle, the stationary arm points along the femur toward the greater trochanter, and the moving arm points along the fibula toward the lateral malleolus. Measured this way, intra-rater reliability is generally within about ±5° and inter-rater reliability is worse, which is why the useful comparison is nearly always the same examiner measuring the same patient over time, or the injured side against the uninjured side.

Toby's 10°–75° against the other knee's 0°–140° is exactly that comparison, and the loss at both ends is the informative part.

End-feel is the quality of the resistance you feel at the end of a passive movement, and it tells you which tissue stopped the movement. Learning normal end-feels is the entire skill, because the abnormal ones are defined against them.

Two of Toby's findings are end-feel findings. His Lachman with a soft endpoint means the tibia kept translating and nothing caught it: a firm endpoint would say the ligament, or what is left of it, took the load. His restricted range with a boggy resistance at both ends is the end-feel of a joint too full of fluid to reach either extreme.

8.7 Four Joints in Depth

Four joints, four different mechanical problems, four different solutions. Read them as a set: the knee is stability without bone, the shoulder is range without stability, the hip is stability without effort, and the temporomandibular joint is two incompatible movements in one capsule.

The knee — stability with no bony help at all

The knee is the largest and most complicated joint in the body, and it is built on a contradiction. It must be rigid enough to carry three to seven times body weight in running and landing, and mobile enough to flex 140° — and it must do that with articular surfaces that provide essentially no stability. Two convex femoral condyles sit on two nearly flat tibial plateaus. It is a ball resting on a plate. Everything holding it together is soft tissue.

It is also two joints in one capsule: the tibiofemoral joint, which is bicondylar, and the patellofemoral joint, which is a sellar articulation between the patella and the femoral trochlea. They share a cavity, a capsule, and a synovial membrane, which is why a problem in one produces an effusion that affects the other.

   THE KNEE — coronal (front) and sagittal (side) schematic

   ══ CORONAL VIEW (looking at the front, knee extended) ══════════════════
                        FEMUR
              ┌────────────────────────┐
              │                        │
              │   PCL ╲      ╱ ACL     │
       MCL    │        ╲    ╱          │    LCL (fibular collateral)
    (tibial   │         ╲  ╱           │    - cord-like, EXTRAcapsular
   collateral)│          ╲╱            │    - NOT attached to the
    - broad,  │          ╱╲            │      lateral meniscus
      flat    │         ╱  ╲           │    - resists VARUS
    - ATTACHED│    ╭───╱────╲───╮      │
      to the  ├────┤ MED│ LAT  │ ├─────┤
      MEDIAL  │    │MENISCUS   │ │     │
      MENISCUS│    ╰───────────╯ │     │
    - resists │      TIBIA       │  FIBULA
      VALGUS  └────────┬─────────┘   (carries ~10% of load)
                       │
      ← the MCL–medial meniscus attachment is why they tear together →

   ══ SAGITTAL VIEW (from the side, knee flexed ~30°) ═════════════════════

        FEMUR                          ACL — ANTERIOR CRUCIATE LIGAMENT
      ╱‾‾‾‾‾‾‾╲                        origin: posteromedial aspect of the
     │  ╲      │                       LATERAL femoral condyle, inside the
     │   ╲ PCL │   quadriceps          intercondylar notch
     │    ╲    │   tendon              insertion: ANTERIOR intercondylar
     │  ACL╲   │      │                area of the tibia
     │  ╱   ╲  │   PATELLA             length ~32 mm, width ~10 mm
     │ ╱     ╲ │   (sesamoid — it      failure load ~2,000 N
     │╱       ╲│    increases the      JOB: stops the TIBIA sliding
     ├─────────┤    quadriceps         ANTERIORLY on the femur, and
     │  TIBIA  │    moment arm ~30%)   resists internal rotation of the
     └────┬────┘      │                tibia + hyperextension
          │        patellar tendon
          │           │                PCL — POSTERIOR CRUCIATE
     tibial tuberosity                 thicker, ~2× the strength.
                                       Stops the tibia sliding POSTERIORLY.
   NAMING RULE: cruciates are named    Torn by a dashboard striking the
   for their TIBIAL attachment.        front of a flexed tibia.
   ACL = anterior on the tibia.
   They CROSS (L. crux) — hence        MENISCI — C-shaped fibrocartilage
   "cruciate" — and one bundle is      wedges. Deepen the plateau, spread
   always taut at every angle of       load over ~3× the area, absorb shock.
   flexion.                            Blood supply: outer 1/3 RED (healable)
                                       middle RED-WHITE, inner 1/3 WHITE
                                       (avascular — cannot heal, is trimmed)

Four structures deserve a sentence each.

The cruciate ligaments are the knee's central pivot. They are intracapsular but extrasynovial: they lie inside the joint capsule, but the synovial membrane folds backward around them from behind, so they are technically outside the synovial cavity while being completely bathed in synovial fluid. That sentence becomes decisive in §8.9.

The menisci deepen the tibial plateau from nearly flat to gently dished, converting point contact into area contact. Removing a whole meniscus increases peak contact stress on the articular cartilage by two to three times, which is why total meniscectomy — routine in the 1970s — is now avoided; it reliably produced osteoarthritis a decade later.

The patella is the body's largest sesamoid, embedded in the quadriceps tendon, and it does one thing: it holds the tendon away from the joint's axis of rotation, increasing the quadriceps' moment arm by roughly 30%. It is a pulley made of bone.

The screw-home mechanism is the knee's second degree of freedom doing something useful. In the last 20° of extension the tibia rotates laterally about 10° on the femur — obligate, automatic, driven by the larger medial femoral condyle and the tightening ACL. It locks the knee into its close-packed position, so that standing with the knee straight requires almost no quadriceps activity. Popliteus unlocks it to begin flexion.

The shoulder — range bought at the price of everything else

The glenohumeral joint is the mobility extreme (§7.5): a shallow glenoid deepened perhaps 50% by a fibrocartilaginous labrum, a lax capsule with redundant inferior folds that unfold as the arm elevates, three weak glenohumeral ligaments that tighten only at end-range, and a rotator cuff doing the real work by compressing the humeral head into the socket — a mechanism called concavity–compression. It is the most mobile and the most dislocated joint in the body, and those are the same fact.

Total elevation of the arm is not a glenohumeral movement alone. Scapulohumeral rhythm divides 180° of elevation roughly 2:1 between glenohumeral motion (120°) and scapular upward rotation on the thorax (60°), with the clavicle elevating and rotating at the sternoclavicular and acromioclavicular joints to permit it. Disturb the scapular half — weak serratus anterior or lower trapezius, a painful shoulder held guarded — and the glenohumeral half is asked to do more than 120°, which is where impingement begins.

The hip — stability that costs nothing to maintain

The hip is the stability extreme. A deep acetabulum plus a labrum encloses more than half the femoral head; the capsule is thick; and three capsular ligaments — the iliofemoral (the Y ligament of Bigelow, the strongest ligament in the body), the pubofemoral, and the ischiofemoral — spiral around the neck in such a way that they wind tight in extension.

That spiral is the elegant part. Standing upright extends the hip, which tightens the ligaments, which locks the joint. You can stand for an hour with almost no gluteal activity, balanced on ligaments. Standing still is nearly free because a ligament is doing it — and the same arrangement means the hip is least stable in flexion, when the spiral unwinds, which is exactly the position in which it dislocates.

The temporomandibular joint — two joints in one capsule

The TMJ is the strangest joint in the body, and it is worth a paragraph because it violates several rules just established.

8.8 How Joints Fail

Joints fail in a small number of stereotyped ways, and each one is a different tissue giving out. Name the tissue and the rest follows.

Sprain, strain, and the grading system

The commonest confusion in the whole vocabulary: a sprain is a ligament, a strain is a muscle or tendon. Sprains are graded by how much of the ligament has failed, and the grade is a statement about laxity and endpoint rather than about pain.

Grade III injuries are often less painful than grade II, because a completely torn ligament has no intact fibres left to stretch. Pain is a poor guide to severity here; endpoint is the useful sign, which is why Toby's soft Lachman endpoint mattered more than anything he reported.

The lateral ankle sprain is the worked example and the commonest musculoskeletal injury in the world. The mechanism is inversion with plantar flexion (§8.6), and the ligaments fail in a fixed order — anterior talofibular first, then calcaneofibular, then, rarely, posterior talofibular. A high ankle sprain is a different injury: external rotation of the foot forces the talus to splay the mortise apart, tearing the tibiofibular syndesmosis. It is much slower to recover, because a syndesmosis is a load-bearing structural link rather than a motion-limiting strap.

Dislocation and subluxation

Dislocation (luxation) is complete separation of the articular surfaces, which requires tearing the capsule and usually some ligaments. Subluxation is partial separation that reduces spontaneously — which is exactly what Toby described as the knee "coming apart and going back together," and it is the same event the pivot-shift test reproduces deliberately in clinic.

Every dislocation stretches the capsule, and stretched capsules stay stretched (§8.5). That is why recurrence rates after a first traumatic anterior shoulder dislocation under the age of 20 exceed 70%, and why the second dislocation typically takes less force than the first.

A paediatric special case worth knowing: nursemaid's elbow, subluxation of the radial head. A child under about five is lifted by a pronated forearm; the head, still small and round rather than disc-shaped, slips partly out from under the anular ligament. The child holds the arm pronated and refuses to use it, with no swelling and a normal radiograph. Supination with flexion reduces it in seconds. The condition disappears after about age five for a purely anatomical reason: the radial head grows into its adult flared shape and can no longer escape the ring.

Meniscal and labral tears

Fibrocartilage tears by pattern, and the pattern predicts the symptom. A longitudinal or bucket-handle tear splits the meniscus parallel to its circumference, and the inner fragment can flip into the notch and produce true locking. A radial tear cuts perpendicular to the fibres from the free edge outward and destroys hoop tension, so the meniscus stops distributing load at all. A horizontal cleavage tear splits it into upper and lower leaves and is usually degenerative. A flap tear leaves a mobile tag that catches and clicks. And a root tear avulses the meniscal attachment from the tibia, which unhooks the hoop and is functionally equivalent to removing the whole meniscus.

The decisive variable is not the pattern but the zone. The meniscus is vascularized only from its periphery by the perimeniscal capillary plexus: the outer third is the red zone (vascular, repairable), the middle is red–white, and the inner third is the white zone (avascular, not repairable, trimmed). Toby's tear is peripheral, longitudinal, and 2 mm from the capsular margin — red zone, and therefore repairable, which is why it will be sutured rather than trimmed at the time of his reconstruction.

The labrum is the same tissue playing the same role at the shoulder and hip: a fibrocartilage rim that deepens a shallow socket and creates a seal. Bankart lesions (anteroinferior, from dislocation) and SLAP lesions (superior, from biceps traction) are its two classic failures, and acetabular labral tears — often with femoroacetabular impingement — are the hip's version.

Bursitis and tendinopathy

Bursitis is inflammation of a bursa, and it is almost always mechanical: repeated pressure or friction at a named site produces a named condition. Prepatellar bursitis from kneeling ("housemaid's knee"), olecranon bursitis from leaning on elbows, trochanteric pain from lying on one side, retrocalcaneal bursitis from a stiff heel counter, subacromial bursitis from overhead work. The diagnosis is usually made by asking what the patient does with that part of the body.

Tendinopathy deserves a correction that took the field thirty years to accept. The old name was "tendinitis," implying inflammation. Histologically, chronic tendon pain shows collagen disarray, increased ground substance, neovascularization with accompanying nerve ingrowth, and very few inflammatory cells — a failed healing response rather than an inflammatory one. Hence tendinosis, and hence the shift in treatment from rest and anti-inflammatories toward progressive loading, particularly eccentric and heavy slow resistance work, which appears to drive matrix remodelling. It is the same principle as everywhere else in this chapter: connective tissue is organized by the loads applied to it, so the treatment for disorganized connective tissue is usually a carefully chosen load.

8.9 Advanced Topic · Why Cartilage and Intra-Articular Ligaments Do Not Heal

Chapter 6 established that bone is the best-healing tissue in the body: a fractured radius in a child reconstitutes itself so completely that the fracture line disappears from the radiograph within a couple of years. This section explains why two tissues sitting a few centimetres away, in the same limb, with the same blood pressure and the same circulating cells, cannot do it at all.

Three reasons, and they are cumulative.

   THE BLOOD SUPPLY OF A SYNOVIAL JOINT — and where it stops

     periarticular arterial anastomosis (e.g. genicular anastomosis)
     ══════════════════════════════════════════════════════════════
        │                    │                        │
        ▼                    ▼                        ▼
    ┌────────────┐   ┌──────────────────┐   ┌────────────────────┐
    │  CAPSULE   │   │ SYNOVIAL MEMBRANE│   │  SUBCHONDRAL BONE  │
    │  vascular  │   │  richly vascular │   │  richly vascular   │
    │  innervated│   │  fenestrated caps│   │  marrow + vessels  │
    └─────┬──────┘   └─────────┬────────┘   └─────────┬──────────┘
          │                    │                      │
          │            makes SYNOVIAL FLUID           │
          │                    │                      ▲
          │                    ▼                      │  ✗ BLOCKED by the
          │       ┌───────────────────────┐           │    CALCIFIED ZONE
          │       │   JOINT CAVITY        │           │    and the TIDEMARK
          │       │   fibrinolytic fluid  │           │
          │       └───────────┬───────────┘           │
          │                   │ diffusion +           │
          │                   │ convection only       │
          ▼                   ▼                       │
   ╔══════════════════════════════════════════════════╪═══════════╗
   ║   ARTICULAR CARTILAGE          NO VESSELS  ──────┘           ║
   ║   ✗ avascular  ✗ aneural  ✗ alymphatic                       ║
   ║   chondrocytes 1–5% of volume, TRAPPED in lacunae,           ║
   ║   cannot migrate through a solid matrix                      ║
   ╚══════════════════════════════════════════════════════════════╝

   ╔══════════════════════════════════════════════════════════════╗
   ║   ACL — intracapsular, EXTRAsynovial                         ║
   ║   ✓ HAS an artery (middle genicular a., through the          ║
   ║     posterior capsule)  → so it BLEEDS when torn             ║
   ║   ✗ but its torn ends hang in SYNOVIAL FLUID, which is       ║
   ║     FIBRINOLYTIC → the blood never organizes into a clot     ║
   ║   → NO SCAFFOLD → no repair.  Ends retract and are resorbed. ║
   ╚══════════════════════════════════════════════════════════════╝

   ┌──────────────────────────────────────────────────────────────┐
   │  MCL — EXTRAcapsular, embedded in vascular soft tissue       │
   │  ✓ artery   ✓ clot forms normally   ✓ HEALS without surgery  │
   └──────────────────────────────────────────────────────────────┘

   MENISCUS  outer 1/3 RED (perimeniscal plexus) → repairable
             inner 1/3 WHITE (no vessels)        → resected

Reason one — avascularity

Every repair process in the body begins with the arrival of cells that are not already there: neutrophils, then macrophages, then fibroblasts and their progenitors. They arrive in blood. Articular cartilage has no blood vessels at any point, in any zone, at any age. There is no route by which the inflammatory phase of healing can even begin.

Note that this is not the ACL's problem. The ACL does have an artery — the middle genicular artery, entering through the posterior capsule — which is why tearing it produces 55 mL of blood. Avascularity explains cartilage. It does not explain the ligament, which is why there is a second reason.

Reason two — no fibrin scaffold

Repair does not merely need cells; it needs something for them to crawl into. In bone that something is the fracture hematoma (§6.8): blood fills the gap, clots, and the fibrin mesh becomes the provisional scaffold onto which inflammatory cells, then fibroblasts and chondroblasts, then osteoblasts, are laid in sequence. Every stage depends on the first one. The clot is the starting gun.

A torn ACL cannot fire that gun, and the reason is the fluid it is sitting in. Synovial fluid is fibrinolytic: it contains plasminogen activator activity and comparatively little antiplasmin, so fibrin formed within the joint cavity is dissolved rather than organized. As established in §8.3, that is a design feature — it is what stops a joint filling with adhesions after every minor bruise, and it is why the cavity remains a cavity. But the consequence for a ruptured intracapsular ligament is fatal. The blood pouring from the middle genicular artery disperses through 55 mL of fluid and is lysed. No bridging clot forms between the torn ends. Without a scaffold there is no matrix for fibroblasts to invade; the ends retract, are resorbed, and within weeks the intercondylar notch is simply empty.

Contrast the medial collateral ligament, which Toby sprained in the same instant. It is extracapsular — outside the synovial cavity, embedded in vascular soft tissue where a clot forms normally. Isolated MCL injuries heal without surgery in the great majority of cases. Two ligaments, one joint, one event, opposite outcomes, decided entirely by which side of the synovial membrane each one lies on.

Reason three — the cells cannot move

Even given cells and a scaffold, cartilage would still be in trouble, because its cells are immobilized. Chondrocytes make up 1–5% of tissue volume, sit individually in lacunae, and are surrounded by a solid matrix through which they cannot migrate. A fibroblast in loose connective tissue crawls to a wound. A chondrocyte cannot crawl anywhere. It can increase its synthetic output locally — and it does, in early osteoarthritis — but it cannot travel to a defect, and there is no reservoir of progenitor cells inside the tissue to be recruited.

There is one revealing exception. A partial-thickness cartilage defect, which stays above the tidemark, does not heal at all. A full-thickness defect, which penetrates the calcified zone into subchondral bone, does fill — because it has breached into a vascular space, and marrow cells and blood enter it. What fills it is fibrocartilage, type I collagen rather than type II, with inferior stiffness and wear resistance, and it typically degrades within two to five years. That exception is the basis of microfracture surgery, and it proves the argument: give cartilage a blood supply and a clot and it produces repair tissue, just not the right tissue.

What surgery restores, and what it cannot

Reconstruction exists precisely because repair does not work. A torn ACL is not sewn back together; it is replaced with a graft, usually the patient's own hamstring tendons, patellar tendon with bone blocks, or quadriceps tendon.

The graft then undergoes ligamentization — remodelling in the sense of §6.7, applied to dense regular connective tissue:

What reconstruction restores: anterior translational restraint, most rotational control, and the subjective sense of a stable knee. Lachman and pivot-shift tests normalize in most patients.

What it cannot restore, and this is the part worth carrying:

• The mechanoreceptors. The native ACL contains Ruffini and Pacinian endings feeding the ACL–hamstring reflex arc (§8.5). A graft is reinnervated slowly, partially, and never to native density. The strut comes back; the sensor does not.
• The exact geometry. The native ACL has bundles that tighten at different flexion angles, so some part of it is taut throughout the arc. A single-bundle graft approximates this.
• Material properties. Grafts reach perhaps 50–80% of native ACL properties and stop there.
• The damage already done. The bone bruises, the meniscal tear, and the cartilage insult delivered in the original 50 milliseconds are not undone by a graft. Post-traumatic osteoarthritis risk after ACL injury is elevated roughly fourfold at 10–20 years, and reconstruction reduces it only modestly.

One development is worth mentioning because it is the exception that proves the whole argument. Bridge-enhanced ACL restoration places a resorbable collagen scaffold, loaded with the patient's own blood, between the torn ends — an engineered substitute for the clot that synovial fluid dissolves. Early trials show a repaired ligament that is genuinely the patient's own. Whatever becomes of the technique, the design logic confirms the diagnosis: the problem was never the ligament's biology. It was the absence of a scaffold.

Resolving Toby's knee

Everything needed is now on the table, and it comes down to three sentences.

The structure. Deceleration on a planted foot with a near-extended knee, dynamic valgus, and internal tibial rotation loads all three of the restraints the ACL provides, simultaneously, in under 50 ms. A collateral ligament would require a force across the knee from outside, and nobody touched him.

The timing. The ACL has an artery. Tearing the ligament tears the artery, inside a capsule that will not expand. Arterial bleeding fills a joint in one to three hours — hemarthrosis — while an effusion, produced by an irritated synovium, takes 12–36 hours. His swelling at two hours and 55 mL of frank blood name the tissue before any scan.

The prognosis. His radius healed at eight because a clot formed. His ACL will not heal because one cannot. His MCL will heal because it lies outside the fluid that prevents it. His medial meniscus will heal if the tear is in the red zone, and his is, by 2 mm.

Chapter Summary

§8.1 A joint is any place two bones meet, and every joint is an answer to one engineering problem: how to interrupt a load-bearing structure without losing the ability to bear load. The specification has four lines — transmit load across a gap, permit some movements and forbid the rest, survive about 10⁸ cycles, and repair essentially nothing. Every joint sits somewhere on a single axis running from stability to mobility, and no joint gets both, because stability is constraint and constraint is the absence of mobility.

§8.2 Structurally, joints are fibrous (suture, syndesmosis, gomphosis), cartilaginous (synchondrosis, symphysis), or synovial; functionally they are synarthroses, amphiarthroses, or diarthroses. All synovial joints are diarthroses and that is the only clean mapping. The mismatches are informative: a syndesmosis is a synarthrosis or an amphiarthrosis depending on nothing but fibre length. Joint cavities are excavated in weeks 7–8 by hyaluronan-driven cavitation of the GDF5-expressing interzone, and the process requires embryonic movement — which is why fetal akinesia produces arthrogryposis.

§8.3 Six components: articular cartilage, articular cavity, two-layered capsule, synovial fluid, reinforcing ligaments, and nerves and vessels. Articular cartilage is 65–80% water in a type II collagen mesh pressurized by aggrecan, arranged in four zones whose collagen arcs from vertical at the tidemark to horizontal at the surface; it is avascular, aneural, alymphatic, and fed by diffusion and convection — which is to say it is fed by being used. Synovial fluid is a plasma dialysate plus hyaluronan and lubricin; it is shear-thinning, thixotropic, and fibrinolytic, and it lubricates by boundary, fluid-film, and interstitial-pressurization mechanisms. The synovial membrane is not an epithelium: no basement membrane, type A macrophage-like and type B fibroblast-like cells over fenestrated capillaries.

§8.4 Six shape classes — plane, hinge, pivot, condylar, saddle, ball-and-socket — with degrees of freedom set by shape and functioning as a hard ceiling. Movement beyond the ceiling is injury, and the tissue absorbing it is whichever structure forbade that direction. Every joint has one close-packed position of maximum congruence and taut ligaments, which is where it fractures rather than sprains and which dictates how a hand is splinted. Arthrokinematically, joints combine roll with glide according to the convex–concave rule, which is why the humeral head glides down as it rolls up and why losing that glide produces superior migration on a radiograph.

§8.5 Stability comes from articular shape, ligaments, and muscle tone, and the correct ranking is usually the reverse of the intuitive one. Ligaments are crimped and contribute almost nothing until roughly 4% strain, sustain damage by 6%, and never shorten back. Muscle is active, adjustable, and anticipatory, arriving in 30–50 ms or pre-activated before the load. Joint stability is therefore largely a nervous system property, which is why neuromuscular training cuts non-contact ACL injury by half while bracing does much less, and why re-injury after reconstruction is as likely in the other knee.

§8.6 Movements are named from the anatomical position: gliding, the angular movements (flexion/extension, abduction/adduction, circumduction), rotation, and the specials — supination/pronation, dorsiflexion/plantar flexion, inversion/eversion, protraction/retraction, elevation/depression, and opposition. Ranges are measured by goniometry against the opposite side to about ±5°, and the end-feel identifies which tissue stopped the movement: bony, soft tissue approximation, tissue stretch, or the always-abnormal empty, springy block, and guarding. Limb rotation in weeks 7–8, lateral above and medial below, explains most of the apparent inconsistencies in the terminology.

§8.7 Four joints, four problems. The knee has no bony stability and is held entirely by ligaments and menisci, with the patella as a bony pulley adding 30% to the quadriceps moment arm and the screw-home mechanism locking it in extension. The shoulder trades stability for range, depends on concavity–compression by the rotator cuff, and divides elevation 2:1 between glenohumeral and scapular motion. The hip achieves stability for free, because its capsular ligaments spiral tight in extension. The temporomandibular joint houses a hinge and a glide in two separate compartments divided by a disc, which is why its own click is diagnostic.

§8.8 Sprains are ligaments and strains are muscles; grade III is often less painful than grade II, and the endpoint matters more than the pain. Dislocation stretches capsule permanently, which is why recurrence exceeds 70% under 20. Meniscal and labral tears are classified by pattern but decided by zone. Osteoarthritis and rheumatoid arthritis are near-opposites, and the decisive difference is that bone builds in one and is destroyed in the other, from opposite signals reaching the same osteoclast. Gout is a solubility problem expressed in the coldest joint in the body, amplified by the NLRP3 inflammasome and IL-1β. Tendinopathy is degenerative, not inflammatory, which is why it is treated with load.

§8.9 Three reasons a joint does not heal itself: articular cartilage has no blood supply; torn intracapsular structures cannot form a fibrin scaffold, because synovial fluid is fibrinolytic; and chondrocytes cannot migrate through a solid matrix. Fracture repair has all three and works; the ACL has an artery but no scaffold and fails; the MCL, extracapsular, heals normally. Reconstruction restores the strut, most of the rotational control, and the sense of stability — but not the mechanoreceptors, not the native material properties, and not the cartilage and meniscal damage already delivered.

Case File 8 · Resolution

Question 1 — Which structure failed, and why does non-contact deceleration-and-pivot predict it?

The anterior cruciate ligament of the left knee, with a grade I–II sprain of the medial collateral ligament and a peripheral tear of the medial meniscus.

The mechanism predicts the ACL because it loads all three of the things the ACL restrains at once. The ACL runs from the posteromedial aspect of the lateral femoral condyle to the anterior intercondylar area of the tibia (Figure 8.5), and it resists (a) anterior translation of the tibia on the femur, (b) internal rotation of the tibia, and (c) hyperextension.

Take the mechanism apart:

• Deceleration on a nearly extended knee (10–30° of flexion). Near extension, the patellar tendon meets the tibia at an angle that converts quadriceps force into a large forward pull on the tibia, and at that flexion angle the hamstrings have almost no line of pull with which to oppose it. The ACL is the only remaining restraint against anterior drawer.
• Pivot on a planted foot. The studs fix the foot, the body rotates over it, and the tibia rotates internally beneath the femur. That is load (b).
• Dynamic valgus as the hip adducts and internally rotates — the knee falls medially, stressing the MCL and, because the MCL is attached to the medial meniscus, dragging the meniscus with it.

The whole event occupies 17–50 ms, shorter than any voluntary correction, and the ACL's failure load of about 2,000 N is exceeded. No opponent is required, which is why 70% of these injuries are non-contact.

Why not a collateral ligament? Because the collaterals resist forces applied across the knee from outside — the MCL resists valgus, the LCL resists varus — and an isolated collateral rupture essentially requires an external blow. Nobody touched him. A self-generated deceleration cannot produce a pure frontal-plane force large enough to rupture the MCL, though it produced enough valgus to sprain it. His varus stress test is firm, which excludes the LCL, and his valgus laxity has a firm endpoint, which says enough MCL fibres remain to catch the movement.

The examination confirms the reasoning rather than replacing it. Lachman positive with a soft endpoint at 20–30° isolates the ACL, because the secondary restraints that confound the anterior drawer at 90° are slack there. Pivot shift positive reproduces the injury itself: the tibia subluxes anteriorly in extension and reduces with a clunk as the knee flexes past 30°. McMurray positive with a medial joint-line click indicates a meniscal fragment caught between condyle and plateau. And the restricted arc of 10°–75° is not stiffness — it is a joint too full of fluid to reach either end of its range, resting in its loose-packed position of maximum capsular volume (§8.4).

Question 2 — Why two hours rather than two days, and what does 55 mL of blood mean?

Because the swelling is blood, not effusion, and the two arrive on different timetables because they arrive by different routes.

The ACL is intracapsular, and although the synovial membrane folds around it, it carries its own blood supply: the middle genicular artery, entering through the posterior capsule. Rupturing the ligament ruptures its artery. Bleeding at arterial pressure into a capsule that will not expand fills the joint in one to three hours. That is hemarthrosis, and it produces exactly what Toby has: a tense, warm knee, a 3 cm circumference difference, and a range limited at both ends by volume.

An effusion, by contrast, is synovial fluid overproduced by an irritated synovial membrane. That is a cellular response requiring the synoviocytes to be stimulated and to secrete, and it takes 12–36 hours to accumulate. A knee that swells overnight is a different injury — an isolated meniscal tear, a capsular sprain, a contusion — with a different prognosis.

So the timeline is diagnostic before any imaging, and the aspirate confirms it: 55 mL of frank blood, against a normal joint volume of 0.5–4 mL. In acute traumatic knee hemarthrosis, 70–80% of cases prove to be ACL tears; the remainder are mostly osteochondral fractures, peripheral meniscal tears in the vascular red zone, and patellar dislocations. The absence of fat globules matters too: marrow fat floating on the blood would indicate an intra-articular fracture releasing marrow, and there is none — consistent with an MRI showing bruised but unbroken bone.

Question 3 — Why will this not heal when his broken wrist did?

Because repair begins with a clot, and the ACL cannot form one.

When Toby broke his distal radius at eight, the fracture tore vessels in the periosteum and medullary cavity; blood filled the gap and clotted; and that fracture hematoma (§6.8) was the scaffold onto which everything else was built — inflammatory cells, then fibroblasts and chondroblasts laying a soft callus, then osteoblasts converting it to bone, then months of remodelling along the new lines of stress. Remove the first step and none of the others happen.

The ruptured ACL sits in the joint cavity, bathed in synovial fluid, which is fibrinolytic. That is a feature, not a defect: it is what stops a joint filling with adhesions after every minor bruise (§8.3). But it means the blood pouring in from the middle genicular artery never organizes into a bridging clot between the torn ends. It disperses and is lysed. Without a scaffold there is no matrix for fibroblasts to invade, the retracted ends are progressively resorbed, and within weeks the intercondylar notch is empty. Reconstruction with a graft exists precisely because repair does not work.

His MCL, injured in the same instant, is extracapsular. It bleeds into vascular soft tissue where a clot forms normally, and isolated MCL injuries heal without surgery in the large majority of cases. Two ligaments, one joint, one event, opposite outcomes — decided entirely by which side of the synovial membrane each lies on.

The same principle sorts the rest of his knee and much of the last two chapters. His medial meniscus tear is 2 mm from the capsular margin, in the vascular red zone, so it can be repaired and will heal; a tear 8 mm further in would be trimmed instead. Articular cartilage never heals at all, because it has no blood supply anywhere and its cells cannot move. The scaphoid (§7.5) and the femoral neck (§7.6) fail to heal when the fracture line crosses their retrograde blood supply. Whether a tissue can repair itself is a question about its vasculature and its scaffold, not about how tough it is.

Systems Integration Case File · Entry 8

Review

Level 1 · Recall

8.1 Which of the following is a synchondrosis?

a) the pubic symphysis    b) an intervertebral disc    c) the epiphyseal plate    d) the distal tibiofibular joint

8.2 Which structure prevents the tibia from sliding anteriorly on the femur?

a) posterior cruciate ligament    b) medial collateral ligament    c) anterior cruciate ligament    d) medial meniscus

8.3 Articular cartilage receives its nutrition by:

a) branches of the periarticular anastomosis    b) subchondral vessels crossing the tidemark    c) diffusion and convection from synovial fluid    d) lymphatics of the capsule

8.4 The most important stabilizer of the glenohumeral joint is:

a) the glenoid labrum    b) the glenohumeral ligaments    c) the rotator cuff    d) the shape of the glenoid

8.5 In a sprain, a soft endpoint on stress testing indicates:

a) grade I    b) grade II    c) grade III    d) a strain rather than a sprain

8.6 Which radiographic feature distinguishes osteoarthritis from rheumatoid arthritis?

a) joint space narrowing    b) osteophytes    c) soft tissue swelling    d) symmetrical distribution

Level 2 · Comprehension

8.7 Why is muscle tone ranked above ligaments as a stabilizer of most joints? Give the timing argument.

8.8 Explain why the hip is stable and the shoulder is not, and state precisely what the shoulder gets in return.

8.9 A knee is aspirated two hours after injury and yields 55 mL of blood with no fat globules. Another is aspirated 24 hours after a twisting injury and yields 30 mL of clear yellow fluid. Interpret both.

Level 3 · Clinical Application

8.10 A 24-year-old lands from a rebound, feels the knee give way, and cannot straighten it fully. There is a springy block to the last 20° of extension. Explain the finding and name the lesion.

8.11 A 62-year-old has knee pain worse after gardening, 15 minutes of morning stiffness, and a weight-bearing X-ray showing narrowing of the medial compartment only, with osteophytes. Name the disease and explain the asymmetry.

8.12 A 3-year-old is lifted by one hand to cross a road and afterwards refuses to use that arm, holding it at her side, pronated. There is no swelling, no bruising, and no tenderness over any bone. The radiograph is normal. What has happened, and why does this condition disappear after about age five?

Level 4 · Integration and Synthesis

8.13 Toby's surgeon proposes reconstruction with a hamstring tendon graft. Using Chapters 4, 6, and 8, explain what must happen to that graft over the following year, and why return to sport is delayed to nine to twelve months rather than the six weeks that healed his radius.

8.14 Toby asks whether his other knee is at risk. Construct the argument that it is, using the material of this chapter, and state what should be done about it.

Concept Map to Complete

Copy this onto blank paper and fill every bracket from memory before checking the chapter.

                                A JOINT
                                   │
             ┌─────────────────────┴─────────────────────┐
      STRUCTURAL class                            FUNCTIONAL class
             │                                            │
   ┌─────────┼─────────┐                ┌─────────────────┼─────────────────┐
[ _______ ] [ _______ ] SYNOVIAL   SYNARTHROSIS    [ ___________ ]    [ _________ ]
   │           │           │        = [ _______ ]   = slightly          = freely
subtypes:   subtypes:   6 parts:                      movable             movable
[ _____ ]   [ _______ ]  1 [ ______________ ]  ← 65–80% [ ___ ], type [ __ ]
[ _____ ]   [ _______ ]  2 articular cavity       collagen, aggrecan; 4 zones
[ _____ ]                3 [ ______________ ]  (2 layers: [ ___ ] + [ ___ ])
                         4 synovial fluid → lubricate, absorb, and
                                            [ _______________ ]
                                          → and it is [ _____________ ]
                         5 [ ______________ ]  types: capsular /
                                                extracapsular / [ __________ ]
                         6 nerves + vessels → vessels supply [ ___ ],
                                              NOT [ ______________ ]
                                   │
                  ┌────────────────┴────────────────┐
             6 SHAPE CLASSES                  STABILITY comes from
             (degrees of freedom)              1 [ ______________ ] (fixed)
             plane      [ _ ]                  2 [ ______________ ] (passive;
             hinge      [ _ ]                     acts only when [ _______ ];
             pivot      [ _ ]                     damaged past [ __ ]% strain)
             condylar   [ _ ]                  3 [ ______________ ] (active;
             saddle     [ _ ]                     arrives in [ __ ]–[ __ ] ms;
             ball/socket[ _ ]                     the ONLY trainable one)
                                   │
                        WHY TISSUE DOES OR DOES NOT HEAL
                                   │
            ┌──────────────────────┴──────────────────────┐
     CAN heal                                        CANNOT heal
     bone (clot → [ ________ ] → bony                articular cartilage:
     callus → remodel)                               ✗ [ __________ ]
     MCL: because it is [ _____________ ]            ✗ cells [ ____________ ]
     outer meniscus: [ ___ ] zone                    ACL: torn ends lie in
     full-thickness cartilage defect →               [ ______________ ] fluid,
     fills with [ ______________ ]                   so no [ ______ ] forms

Lab / Self-Exploration

1. Map degrees of freedom on your own hand. At the metacarpophalangeal joint, flex, extend, abduct, adduct, and circumduct — two axes. At the proximal interphalangeal joint, try the same: only flexion and extension. One axis. Then find the third movement at your thumb base, opposition, that no other digit has, and name the joint class that permits it.
2. Feel your own end-feels. Passively extend your elbow to its limit — abrupt and bony. Passively flex it — soft and squashy, muscle meeting muscle. Passively dorsiflex your ankle with the knee straight — firm with a little give, which is tissue stretch. Three normal end-feels in ninety seconds; every abnormal one is defined against them.
3. Perform a Lachman test on a willing, uninjured partner. With the knee flexed 20–30°, stabilize the femur with one hand and draw the tibia forward with the other. Feel the firm endpoint — the ACL catching. Learning what normal feels like is the entire skill.
4. Watch scapulohumeral rhythm. Stand side-on to a mirror, place your opposite hand on the inferior angle of your scapula, and raise your arm slowly to full elevation. Feel the scapula stay nearly still for the first 30° and then rotate upward steadily. That is the 2:1 rhythm, and it is why a stiff scapula produces a painful shoulder.

Key Terms

Next: Chapter 9 · The Muscular System I — how the muscles that stabilize and move every joint in this chapter actually generate force, why a muscle atrophies within days of a joint effusion, and why cardiac muscle cannot be replaced when it dies.