Flexion–extension: range of motion

The accepted nomenclature for describing elbow flexion–extension range of motion is that of the American Academy of Orthopaedic Surgeons.¹ They defined the fully extended position, with a straight arm, as 0°, and the fully flexed position to be approximately 146°. Some earlier papers had used the ‘elbow angle’, which is the angle between the limb segments, but that convention has fallen from use among orthopaedic surgeons, although it remains common among veterinary surgeons. A review of the literature shows that many papers have merely estimated the range of normal motion, but those where it has actually been measured have reported a mean range from 0° to 142° of active elbow flexion.² The distribution of range of motion was investigated by West,³ who found that of 517 elbows that were measured 25% hyperextended by 5° or more, and more than 90% had maximum active flexion of between 140° and 150°.

There are several factors which can affect the range of elbow flexion–extension. Glanville and Kreezer⁴ examined the difference between active and passive elbow flexion, and found that passive movement allowed the range to increase from 141° to 146°. Active motion is limited by apposition of the soft tissues between the humerus and forearm and, as such, motion with the flexor muscles relaxed would be expected to be greater.

Several studies have looked for side-to-side differences in range of motion, and also examined the differences between dominant and non-dominant arms. These studies have found no significant differences in motion and therefore loss of motion can be calculated by comparison with the uninjured limb.

Other authors have noted that the build of the subject affects the range of motion, with slim subjects typically having approximately 10° more flexion than muscular or fat individuals.

There has also been agreement that the range of motion reduces with increasing age, with children hyperextending by a mean of 5°. After the fourth decade there is a loss of several degrees of flexion, with several more degrees of extension being lost after the sixth decade.

The final factor that affects the range of flexion–extension is the sex of the subject. Studies have shown that females have 5–8° more hyperextension than males.²

Flexion–extension and the carrying angle kinematics

Elbow motion results largely from the articular geometry, particularly the shape of the humeral trochlea and capitellum. These have been approximated as a spool and hemisphere placed coaxially across the distal humerus. In the past, some authors have produced diagrams which showed relative axial motion of the radius alongside the ulna, as the elbow flexed and extended, suggesting that this resulted from the capitellum being displaced from the trochlear axis, thus acting like a cam. However, there has been clear radiographic evidence of the coaxiality of the articular surfaces, as long as the X-ray beam is aligned accurately.⁵ Provided this is the case, the articular features are seen as a series of concentric circles and arcs of circles representing the waist of the trochlea, the capitellum and the medial lip of the trochlea. These structures are situated anterior to the shaft of the humerus, with the medial epicondyle hanging back posteriorly and proximal to the axis. The lateral view also shows the coronoid and olecranon fossae, which approach each other from anteriorly and posteriorly to produce a thin membrane of bone immediately proximal to the waist of the trochlea. This membrane may actually be perforated in elbows that hyperextend, and this supratrochlear foramen has an ethnic basis.⁶ The presence of these fossae increases the range of elbow flexion–extension before it is limited by bony impingement from the coronoid and olecranon processes. There is also a matching shallow anterior fossa that accommodates the rim of the radial head in terminal elbow flexion. However, in life elbow flexion is more likely to be limited by apposition of the bulky soft tissues between the shafts of the humerus and forearm bones. This tendency is reduced by the anterior offset of the distal humeral articulations and also the anterior position of the trochlear notch of the ulna, in relation to its shaft. These features combine to widen the gap between the bone shafts in deep elbow flexion. As much of the soft tissue is muscle, it follows that the range of passive elbow flexion is greater than that of active flexion.

The early work that described the axes of the trochlea and capitellum as being eccentric when viewed laterally suffered from lack of appreciation of the orientation of the axis about which flexion occurs. The key to this is the carrying angle and how that angle varies with elbow flexion. The carrying angle has been defined in many specific ways, but all of these seek to describe the lateral deviation of the extended forearm from the sagittal plane, that is, aligned along the anatomical axis of the shaft of the humerus. Specific definitions have depended on how the axis of the forearm has been defined, either in terms of a line passing from the centre of the capitellum to the centre of the ulnar head distally, an axis based on bony anatomical features, or else as a line which passes along the centre of the forearm, irrespective of the structures within it.⁷ It is widely recognized that the carrying angle is maximal when the elbow is extended, and reduces until the forearm overlays the upper arm in maximum elbow flexion. Provided that the humero-ulnar joint acts as a uniaxial hinge this variation can be obtained by inclining the flexion axis so that it bisects the carrying angle. While the coaxial circles seen on an accurate lateral radiograph⁵ suggest this to be true, the literature has found it contentious until the recent advent of more modern 3-D kinematic measurement methods. In particular, it is possible to define the motion of one limb segment in relation to another in terms of the ‘instantaneous screw axis’, which allows a series of axes to be defined as the motion progresses incrementally. It has been shown⁸ that these instant axes are close to being parallel and intersect through a small zone at the centre of the trochlea. For this reason we can assume for all practical purposes that the elbow acts like a simple hinge under defined loads. The orientation of the hinge axis is not perpendicular to the sagittal plane of the humerus, but is abducted from that orientation by one-half of the carrying angle (Fig. 3.1).⁹

Figure 3.1 The axis of articulation is abducted from being perpendicular to the axis of the humerus, by approximately a half of the carrying angle. Thus the carrying angle reduces from maximal in extension close to zero in full flexion (this may be demonstrated by folding a strip of cardboard shaped like an extended upper limb). There is a range of abduction–adduction laxity allowed by the collateral ligaments, which is used during forearm rotation. This laxity reduces in terminal extension, due to tightening of the anterior soft tissues.

The literature shows that the carrying angle is larger in females than in males, typically being 14° and 11°, respectively.^2,7,10 It has also been shown that the carrying angle relates to build, with carrying angle increasing with heaviness of body type.¹⁰ Although Paraskevas et al¹⁰ found that the carrying angle was larger in the dominant limb than in the non-dominant limb, that difference was small, and most authors have found that they are effectively symmetrical. That observation is useful in children, because the carrying angle develops up to puberty and so there is no standard value. Smith¹¹ found that 9% of children aged less than 11 years had a zero carrying angle and a further 36% had less than 5°. The biomechanical and clinical importance of the carrying angle relates to both its effect on the appearance of the whole of the upper limb and its effect on the distribution of forces across the distal humerus. Inaccurate reduction of supracondylar humeral fractures will result in deformity and alteration of the carrying angle, while heavily loaded activities will result in significant force transmission across the elbow, which may affect the overall stability of the joint.

Pronation–supination motion

The second principal component of elbow motion is rotation of the forearm. The early literature assumed that this motion took place about a fixed axis, with the ulna not moving in relation to the humerus. This was thought to be due to the constraint of the congruent humero-ulnar joint, with the coronoid and olecranon processes gripping the waist of the trochlea. However, it has since been appreciated that the joint is not so closely constrained, with the carrying angle typically varying by ±5°, due to sloppiness of the fit of the articular surfaces and some slackness of the collateral ligaments.⁹ The classic paper that showed this variation was by Ray et al,¹² who took double-exposure radiographs of their own elbows and forearms in pronation and supination. They immobilized the humerus using transcutaneous pins into the epicondyles, and applied traction to the fingers using a Chinese finger basket. They showed that if the forearm rotation axis was aligned through the ulna, due to traction on the little finger, it resulted in the ulna remaining stationary, with the forearm swinging around it like the handle on a bucket. However, if the axis was aligned through the centre of the wrist, with traction on the long finger, then the centreline of the forearm did not move during forearm rotation and the ulna did move in relation to the fixed humerus. Thus, pronation from neutral rotation entailed ulnar abduction, and supination entailed adduction, with the range of ulnar ab–adduction motion corresponding to the ligamentous laxity noted above. It has been suggested that the ulnar abduction during pronation is a reason for the anconeus muscle. As the forearm rotates away from neutral rotation, so the ulna also flexes a little, as it moves alongside the radius. It was shown long ago¹³ that this was a circumduction motion, in which the distal ulna moves in a part-circular path, yet does not itself rotate (Fig. 3.2). This has been investigated again using modern technology, with the same findings.¹⁴

Figure 3.2 This is a stylized end-view of the radius and ulna during forearm rotation. The radius, at the top, rotates around the axis of rotation, which is near the centre of the wrist. At the same time, the ulna, at the bottom, is carried in an arc of circumduction: it moves in an arc of a circle yet does not rotate (indicated by the line drawn across it). The circumduction motion is a combination of abduction–adduction (Ab-Ad), plus flexion–extension (Fl-Ext) motion components.

At the elbow, forearm rotation is accommodated by the radial head rotating in the ulnar notch on the lateral side of the coronoid. This notch is less than 180° in extent and therefore the radial head is vulnerable to subluxation under the anterior force vector applied to the radial tuberosity by the biceps tendon. This vulnerability is controlled by the tension in the surrounding annular ligament, which envelops the radial head and attaches securely to the ridges at the anterior and posterior edges of the ulnar notch. At the wrist, the axis of rotation passes through the centre of the head of the ulna, so the axis of rotation slants across from the capitellum proximally to the ulna distally, when viewed in the coronal plane. In the sagittal plane, the axis of forearm rotation is relatively anterior proximally, again being centred in the capitellum, and is anterior to the interosseous border of the ulna. This has great significance for the function of the interosseous membrane which, when the forearm is in neutral rotation, will be tightest when the radius is furthest away from the interosseous border of the ulna,. It will then slacken when the forearm rotates away from that posture.¹⁵ In particular, the interosseous membrane is slack when the forearm is in pronation, which is the posture taken up during falls onto the outstretched hand, when load transmission along the forearm is important. This will be discussed below.

Medial collateral ligament complex

The principal structures crossing the joint are the anterior and posterior bands of the medial collateral ligament (MCL). The humeral attachment is spread around the base of the medial epicondyle, across its distal and posterior–distal aspects, thus spanning across the site of the flexion–extension axis, which emerges from the centre of the circular medial end-face of the trochlea, at the distal edge of the epicondyle. This means that the most anterior fibres of the anterior band of the MCL are close to being isometric, and therefore tight throughout the range of elbow flexion–extension. Moving posteriorly across the width of the MCL complex, there is an increasing tendency for the ligament fibres to slacken as the elbow is extended.²² Thus the posterior band of the MCL is only tightened completely when it becomes wrapped around the distal aspect of the epicondyle beyond 90° of elbow flexion.²³ Ciccotti et al²² found that the progressive tightening of the MCL fibres as the elbow flexed correlated with a reducing range of valgus laxity. Morrey and An²⁴ found that the MCL resisted approximately one-third of the valgus moment applied to the elbow when it was extended. In addition, they noted that because the anterior joint capsule was tight in this position it also took one-third of the load, with the articular surfaces taking the remainder. As the elbow flexed, so the MCL took a larger proportion of the load, reaching 54% at 90° flexion. Thus they felt that the MCL was the primary restraint to valgus laxity. This was consistent with the findings of Sojbjerg et al,²⁵ who reported that cutting the anterior band of the MCL led to 14° of valgus laxity at 70° of elbow flexion. At this point the posterior band became tight and if this was also cut the elbow became unstable, with up to 31° of valgus deformity.

Lateral collateral ligament complex

The functional anatomy of the lateral collateral ligament is complex, with part of the insertion of the ligament dissipating into the annular ligament rather than attaching to bone. An accessory ligament lies posteriorly, usually known as the lateral ulnar collateral ligament. This corresponds to the anterior margin of the anconeus, thus linking the lateral epicondyle directly to the supinator ridge of the ulna.

In general, the lateral collateral ligament structures are less important for elbow stability than the medial collateral ligament complex. Morrey and An²⁴ found that the trochlea was more important than the lateral ligaments for resisting varus moments, contributing 75% at 90° of flexion. As the elbow reached full extension, the tightening of the anterior capsule caused the soft tissues to become more dominant, with the lateral collateral ligaments contributing 14% and the anterior capsule 32%.

The principal clinical problem associated with disruption of the lateral ligaments is posterolateral subluxation or instability. This is associated with the radial head subluxing posterior to the capitellum as a result of translation and rotation of the forearm. O’Driscoll et al²⁶ found that it required the ulna to supinate 41° out of articulation with the trochlea, supinating the whole forearm and carrying the head of the radius posteriorly while pivoting on the intact medial collateral ligament. The proximal translation of the radial head that occurred took the forearm into 15° of valgus from its normal carrying angle. Dunning et al²⁷ showed that all of the lateral ligamentous stabilizing structures had to be transected before they could produce posterolateral rotatory instability, reflecting the large posterior displacement of the proximal radius that must occur at the time of injury.

The strength of the elbow ligaments

The mechanical properties of the collateral ligaments were measured by Regan et al.²⁸ They found that the anterior band of the MCL was the strongest, with a tensile strength of 261±71 N, while the posterior band had a tensile strength of 159±40 N. The lateral collateral ligament had a tensile strength of 233±116 N. By comparison, the palmaris longus tendon strength was 358±88 N and its stiffness was more than six times greater than the ligaments.

Flexion strength

Many publications have measured elbow flexion strength because of its relevance to ergonomic design of controls and also because this action is amenable to analysis when researching muscle actions. The values reported below are for isometric strength, when the subject exerts maximum strength against a fixed handle or wrist strap. Maximum elbow flexion strength occurs at approximately 60° of flexion, declining towards both greater flexion and extension. Hunsicker,³⁷ who produced comprehensive data, found the mean isometric elbow flexion strength was 55 N m at 0° elbow flexion, 80 N m at 60° and 65 N m at 120° flexion (80 N m flexion torque results from a force of approximately 250 N at the hand) (Fig. 3.4). There have been few studies in which male strength was compared with female, but the data available³⁸ found that females had approximately 55% of male flexion strength.

Figure 3.4 Both flexion and extension strength peak around 60° elbow flexion, at approximately 250 N load on the hand. These graphs show maximal isometric strength of young adult males.

Adapted from Hunsicker P. Arm strength at selected degrees of elbow flexion. Technical Report 54-548. Wright Air Development Centre, Wright-Patterson Air Force Base; 1955.

Several studies have examined the effect of forearm rotation posture on elbow flexion strength. In general, this does not have a large effect, although Provins and Salter³⁹ reported that strength was reduced when the hand was pulling on a handle in a direction that required dorsiflexion of the wrist, and that the handle tended to slip out of the grasp of the fingers. Thus the elbow flexion strength in mid rotation was reduced 22% when the forearm was in pronation.

The data above may be modified if the elbow flexors are not contracting isometrically: Elftman⁴⁰ showed that muscles have a force–velocity relationship such that, as the speed of shortening increases (concentric muscle contraction), the tension reduces and, if the muscle is lengthening, or stretched while contracting (eccentric muscle contraction), the maximum tension increases. Following this trend, Doss and Karpovich⁴¹ reported that the mean force at the hand was reduced by 23% during concentric contraction but increased by 14% in eccentric contraction.

Elbow strength may be reduced by injury or disease. Amis et al³⁶ measured the isometric elbow flexion strength of 102 patients with rheumatoid arthritis. A wrist strap was used in order that elbow strength was not under-assessed due to disability of the fingers. The mean force at 90° of flexion was 146 N for male outpatients, 98 N for male inpatients, and 57 N for both groups of female patients. Thus, when compared with the data for normal subjects, these rheumatoid patients had approximately 45% of normal strength. Some of this reduction in strength would have been secondary to the older age of the patients compared to the fit young adults who typically are recruited for research studies of ‘normal’ people. In addition, the rheumatoid group also included some, mostly female, subjects who had no flexion strength, due to pain.

Extension strength

There have been fewer studies leading to elbow extension strength curves, not all of which have reached the same conclusions. This disparity may reflect the study methods, particularly whether the subject was restrained in a test rig that allowed greater forces to be imposed, such as having a back rest when pushing away to extend the elbow. A study with a large number of male subjects³⁷ found that extension strength was maximal at 60° of flexion, when a mean force of 250 N was applied to a handle, and an elbow extension torque of 75 N m. Elkins et al³⁸ also found that the maximum extension strength was at 60°, while Currier⁴² found the maximum at 90° flexion. However, Hunsicker³⁷ reported a mean strength approximately 60% higher than the other studies, and this may represent a truer attainment of maximum strength (Fig. 3.4). These papers have also included the effects of forearm rotation and shoulder position. The latter variable may have an effect because of the long head of the triceps crossing the shoulder, but the differing postures do not have significant effects on elbow extension strength.

Forearm rotation strength

Forearm rotation strength has been extensively studied by Darcus and Salter,^43,44 who used hand grip dynamometers. They investigated the effects of a wide range of postures and in general found that shoulder position had very little effect on forearm rotation, despite its effect on the long head of the biceps. They also found that the greatest strength of mid-position rotation torques occurred at 90° of elbow flexion for both pronation and supination. Supination strength decreased with elbow flexion, reflecting the reduced length and hence tension of the biceps muscle. A much larger effect resulted from testing rotation strength at different positions of rotation. This related to stretching of the forearm rotation muscles. Thus pronation was strongest when the forearm had been supinated, stretching the pronator muscles, and vice versa for supination. If the hand had already been pronated 60° the pronation strength was only 2 N m, whereas if the forearm was in 60° of supination the pronation strength was 10 N m. As a reference for normal forearm rotation strength, both pronation and supination strengths were a mean of 6 N m with the forearm in neutral rotation, the elbow flexed 90° and with the humerus adducted alongside the trunk.

Forces during elbow flexion

Nordin and Frankel⁴⁵ have described the forces acting across the elbow during elbow flexion. The arm is illustrated with the elbow flexed 90° and with the hand supporting a weight. To maintain that position the extending moment (N m) caused by the weight is resisted by muscular activity equal to the force (N) times the distance (m) from the flexion axis of the elbow.

The muscle action is usually shown diagrammatically as only supplied by the biceps, and its tension (N) times the distance (m) of its line of action from the elbow axis gives the flexing moment (N m), which opposes the extending moment of the weight in the hand. Thus, taking typical approximate values, with the weight 350 mm from the elbow axis and a biceps tendon of 50 mm, rotational equilibrium of the moments acting about the flexion axis demands that the biceps tension must equal seven times the external load (Fig. 3.5). Further, with the load acting downwards and the biceps pulling upwards, parallel to the humerus, there is a net resultant force of six times the external load acting upwards onto the distal end of the humerus. While this is a gross simplification, it does nevertheless show how the lever arm effect causes the internal forces to be multiples of the external load.

Figure 3.5 Simplified representation of elbow flexion, to show how the muscle tension must be a multiple of the load in the hand, resulting in most of the elbow joint force being caused by the muscle, not the load itself.

In reality, elbow flexion actions usually entail actions by the hand, to grip the object being lifted or the handle being pulled, which means that elbow flexion inherently requires activity from the muscles which act to flex the finger and to stabilize the wrist against the combination of transverse shearing load and angulation.⁴⁶ This is a difficult situation to analyse mechanically because of the large number of muscles acting simultaneously, and so a scheme for apportioning the muscle actions is needed. There are a number of well-accepted optimization criteria for predicting muscle actions when walking which are based on energy efficiency, but they are not appropriate for most upper limb actions. A starting point is electromyography, which shows which muscles are active and also indicates their relative levels of stimulation. A key paper by Dempster and Finerty⁴⁷ showed how all the muscles crossing the wrist were co-contracted in order to stabilize the joint when the hand was loaded. There was a range of stimulation from most intense in the direction of the load, through to approximately 50% to stabilize the wrist in the perpendicular direction, and approximately 33% in the antagonistic direction. The net result was both a force action in the desired direction and wrist stability. Most of those muscle tensions cross the elbow, so the action of simply clenching the fist, or pulling on the handle of a suitcase, causes the elbow to be compressed along the direction of the forearm and with the load imposed on both the radial head and the coronoid process. This result is perhaps initially surprising, but a similar simple analysis to that of elbow flexion applied to the metacarpophalangeal joints shows that the finger flexor tendons are also working at multiples of the external load; thus, even if one were to hang from an overhead bar, the wrist and elbow would still be compressed.

The principal muscles acting to flex the elbow are the biceps, brachialis, brachioradialis and pronator teres. The electromyography literature shows that biceps tend to be more active when the forearm is supinated,¹⁸ but that is of little relevance when the limb is loaded heavily, when all these muscles contract maximally. It is worth noting that three of these four muscles insert distally into the radius, and a force analysis shows that the humeroradial joint is loaded heavily, as is the humero-ulnar joint, during elbow flexion.⁴⁸ A further factor is the tendency for some of these muscles to cause forearm rotation, and so the supinating effect of biceps must be counterbalanced by the action of the pronator teres to ensure rotational equilibrium. The effect of flexing the elbow with the forearm in different rotational positions has not been analysed, but it can be predicted that trying to flex with the hand pronated will require the wrist extensors to work very hard, while flexing in supination will load the wrist flexors. The result of these contrasting cases is that the load on the elbow will tend to move from the radius to the ulna as the forearm is supinated.

The combination of all the muscle tensions leads to a resultant force on the elbow that moves around the end of the humerus as the elbow flexes. The resultant force acts on the distal aspect of the humerus in extension, to approximately 60° from the axis of the humerus at 90° of flexion⁴⁸ (Fig. 3.6). This resultant force direction follows from the combination of all the muscle actions, with the hand and wrist muscles, plus the pronator teres and brachioradialis acting along the forearm, while the biceps and brachialis act closer to the direction of the humerus. Near extension, the flexor muscles wrap closely over the anterior aspect of the elbow, and so they have a large mechanical disadvantage and flexion strength is reduced. That disadvantage reduces with flexion, as the lines of action of the muscles bridge across the decreasing gap between the humerus and forearm,⁴⁹ and result in the elbow forces tending to reduce as a proportion of the external load being lifted.

Figure 3.6 The forces acting on the distal humerus, trochlear notch and head of radius during maximal isometric elbow flexion activity. The arrows pointing at the joint surfaces indicate both the magnitude and direction of the resultant force.

Redrawn from Amis, AA, Miller, JH. Design, development and clinical trial of a modular elbow replacement incorporating cement-free fixation. Eng Med 1984; 13:175–179.

Taking all the muscle actions into account, and compounding this by the large mechanical disadvantage caused by the distance from the elbow to the hand, it is not surprising to find that the elbow forces are predicted to be up to 15 times the external load held in the hand. If the maximum isometric force which can be imposed at the hand (noted above to be approximately 250 N) is multiplied by this ratio, it leads to an elbow joint load of 3 kN (four times body weight). Although this calculation relates to young adult males imposing their maximum possible strength, it indicates the size of the forces that are possible and shows that it is incorrect to assume that the elbow is not a load-bearing joint. Noting also the paths of the muscles and their attachments, particularly the distal attachments to the radius of the biceps, pronator teres and brachioradialis, it also follows that both the humero-ulnar and humeroradial joints will be loaded heavily by forceful elbow flexion activity. The joint force may be even higher than shown above, because there is also electromyographic evidence⁵⁰ of antagonistic triceps activity during elbow flexion, acting to stabilize the joint. This has been estimated to raise the force on the coronoid process by up to 20% near elbow extension.

The clinical importance of these biomechanical aspects relates not only to the size of the loads on the elbow but also the directions and points of application of the loads. At the distal humerus the compressive joint force is spread across both the capitellum and trochlea, and both are compressed in a proximal/posterior direction. This matches the migration of loosened elbow prostheses, where the distal end of the humeral component migrates posteriorly, with an accompanying rotation into extension. Associated with this the proximal tip of the humeral stem migrates anteriorly.

It is the muscle tensions crossing the joint line that cause compressive forces on the articular surfaces. At the distal humerus, those muscle tensions focus onto the multiple attachments on and around the epicondyles, with tensions acting distally along the forearm. These loads cause fractured condyles to have their most superficial aspects pulled distally. The resultant motion means that the lateral condyle will tend to rotate internally and the medial condyle will tend to rotate externally.

The combination of forces acting distally/anteriorly at the two extremities of the width of the distal humerus, allied to the central joint loads acting proximally/posteriorly, means that the bar of bone across the distal humerus is effectively under a bending load, which tends to break the central bar, the trochlea, in a posterior direction. Condylar fracture fixation must be designed to resist this tendency.

These actions on the humerus are reflected in the reaction forces acting on the radius and ulna. These are in a distal and posterior direction, which means that the joint load acts towards the base of the coronoid process and onto the posterior lip of the concave end-face of the radial head (Fig. 3.6). The head of the radius is not circular when viewed end-on. The largest dimension is oriented so that the elliptical extension beyond the circle is positioned posteriorly when the forearm is supine. This means that the posterior lip is deepest in supination, making the articulation of the radial head most stable against the anterior subluxation tendency of the biceps, which is strongest in the supinated posture. If the tendency towards anterior subluxation overcomes the stability of the articulation, then further subluxation is resisted by the annular ligament, which encircles the head of the radius attaching to the anterior and posterior edges of the articular surface of the superior radio-ulnar joint. Finally, the balance of forces across both the ulna and radius during elbow flexion is a fundamental reason for considering radial head replacement, whether as part of a total joint arthroplasty or following trauma. While load sharing clearly suggests the mechanical logic for this approach, there is no prosthesis currently available that appears to achieve reliable long-term outcomes.

Forces during elbow extension

Elbow extension is caused almost entirely by triceps tension with the humero-ulnar joint bearing most of the load. The medial head of the triceps is the principal extensor, after which the lateral and then the long heads are recruited as the load increases.⁵¹ There is also a small contribution from the anconeus. This does not mean that the radiocapitellar joint remains unloaded since the long head of triceps partly passes lateral to the olecranon with its tension dissipating into the fascia overlying the anconeus. In addition, when an object is grasped by the hand there will be contraction of the forearm muscles to stabilize the wrist. The elbow is also stabilized by antagonistic activity in the biceps, estimated to equal 7% of the triceps tension.⁵⁰

The elbow forces are very large when the joint extends from a flexed posture. The triceps tendon wraps around the tip of the olecranon process and trochlea, and therefore has a small moment arm of approximately 20 mm.⁴⁹ This great mechanical disadvantage means that the elbow joint force may be in excess of 20 times the external load at the hand, acting onto the distal aspect of the trochlea.⁴⁸ The magnitude of the triceps tension explains the importance of obtaining secure fixation of olecranon fractures. It also means that there is a practical limit on the size of any fixation keel of the ulnar component of joint replacements, since excessive bone excavation will reduce the tensile strength of the olecranon.⁵²

Forces during forearm rotation

The forearm rotation muscles that act transversely between the radius and ulna are relatively small and only a minor component of the tensions in the larger muscles acts to compress these bones together. This means that, despite the accompanying muscle actions when the hand grasps the object that must be rotated, such as the handle of a screwdriver, the forces on the radio-ulnar joints are much smaller than the axial joint loads. The axial compressive forces from the hand and wrist muscles tend to maintain the central articulation of the head of the radius onto the capitellum.

One aspect of the out-of-roundness of the head of the radius is that there is an articular surface on the side of the radial head at the opposite side to the deeper lip that resists the biceps subluxation. This articular surface is oriented medially when the forearm is in mid rotation, so it then presses against the matching articular surface of the ulnar notch.

There has been some work to examine the question of whether the out-of-roundness of the radial head has a significant effect on the mechanics of the elbow⁵³ and, although statistically significant differences were noted, these were all less than 1° difference of ulnar rotation. This is therefore functionally likely to be negligible.

Forces during abduction and adduction actions

The forces that load the elbow into abduction or adduction are most easily visualized when the elbow is flexed 90° and the shoulder is rotated internally or externally. The strongest action is internal rotation, caused largely by tensions in the pectoralis muscles. This is the situation when the palms of the hands are pressed against the sides of a box, pushing towards the centreline of the body. Hunsicker³⁷ found that inward rotation resulted in a maximum force of 218±100 N, while the external rotation force at the hand was 156±81 N.

In young adult males internal rotation of the humerus induces a torque of 65 N m in the bone. At the elbow, this torque must be countered by an equal and opposite reaction, and this arises from a coupled action between tension in the MCL and compression of the radial head onto the capitellum (Fig. 3.7). These internal forces are large multiples of the external load, because of the length of the forearm. As a result the MCL tension will be approximately six times the external force and the load on the radial head approximately eight times.

Figure 3.7 Internal rotation action: the pectoral muscles induce an inward rotation torque in the humerus, so that the palm of the hand presses towards the centreline of the body. This causes tension in the MCL and compression of the head of the radius.

If the head of the radius is excised, it reduces the width of the base of the forearm on the distal humerus. The torque must then be resisted by the MCL tension coupled with compression onto the lateral facet of the coronoid process. The forces will be much higher than with the elbow intact because of the reduced distance between them, and can lead to increased valgus of the elbow after trauma.⁵⁴

Force transmission along the forearm and during falls onto the outstretched hand

The mechanism of force transmission along the forearm was misunderstood for many years because of the influence of anatomical teaching that noted the ulna became more robust from distal to proximal, while the reverse occurred in the radius. When this was coupled with the observation that the interosseous membrane had fibres oriented predominantly from proximal on the radius to distal on the ulna, it seemed that there was a mechanism for gradually transferring any axial load, applied to the distal radius by the carpus, to the ulna. There has more recently been greater understanding. In particular, it is recognized that the bones are much stiffer than the interosseous membrane, and so the majority of the force will still travel along the radius and will not be offloaded by the membrane. Additionally the membrane does not have the opportunity to be elongated sufficiently for a large tension to build up in its elastic fibres. A more sustainable argument is that the membrane acts as an extensive area of origin for the deep finger flexor muscles and their tensions are reacted across the wrist, necessitating the large distal radius.

Classical studies of force transmission along the forearm found that 60% of the load applied to the hand passed to the capitellum via the radial head, with 40% passing to the trochlea from the coronoid process.⁵⁵ It has more recently been shown that this load share is disrupted by the interposition of an excessively thick and stiff radial head prosthesis,⁵⁶ which offloads the coronoid. These experiments applied loads axially along the forearm and although they have provided some useful data the real-life situation is more complex. Experimental work on live subjects falling forwards onto their outstretched hands⁵⁷ found that they took up a characteristic posture, in which the forearm was semi-pronated, the elbow partly flexed (approximately 15°) and the shoulder flexed, abducted and partly inwardly rotated. This combination of arm posture caused the palm of the hand to be placed flat on the floor in front of the falling body (Fig. 3.8). The flexed elbow in this voluntary situation allowed energy to be dissipated by forced flexion of the elbow, stretching the triceps, plus forced retraction of the shoulder, stretching the pectoral muscles. If the subject fell onto a fully extended arm, it would act as a rigid strut, so could not then dissipate the energy of falling. It would have to be released by creating bone fractures. The characteristic falling posture means that the impact force on the palm of the hand causes a combination of axial compression and valgus moment on the forearm, which may cause a fracture of the radial head and tension in the medial collateral ligament. In that situation, the force on the head of the radius will be larger than the external load because of the MCL tension, while the load on the coronoid process may be zero.^58,59

Figure 3.8 The posture assumed during a fall forwards onto the outstretched hand: the forearm is semi-pronated and the elbow flexed approximately 15°. The shoulder is partly internally rotated, so the palm of the hand is flat on the floor. This brings the radius to the superior aspect and the ulna inferior, so the external load then causes axial compression along the radius and tension in the MCL. Energy is dissipated by stretching the triceps, otherwise the head of the radius may be fractured.