a risk-factor model for anterior cruciate ligament injury

18
Sports Med 2006; 36 (5): 411-428 REVIEW ARTICLE 0112-1642/06/0005-0411/$39.95/0 © 2006 Adis Data Information BV. All rights reserved. A Risk-Factor Model for Anterior Cruciate Ligament Injury Gerwyn Hughes and James Watkins Department of Sports Science, University of Wales Swansea, Swansea, UK Contents Abstract .................................................................................... 411 1. Anterior Cruciate Ligament (ACL) Injuries ................................................... 412 2. Sex Differences in ACL Injuries ............................................................. 412 3. Anatomy of the Knee Complex ........................................................... 413 3.1 The Tibiofemoral Joint ................................................................ 413 3.2 The Function of the Cruciate Ligaments ................................................ 414 4. Risk-Factor Models ....................................................................... 414 5. Passive Stability Risk Factors ............................................................... 415 5.1 Geometry of the Articular Surfaces .................................................... 415 5.1.1 Alignment of the Femur with Respect to the Tibia ................................. 415 5.1.2 Congruence Between the Articular Surfaces of the Femur and Tibia ................ 416 5.2 Ligament Laxity ...................................................................... 417 6. Dynamic Stability Risk Factors ............................................................. 419 6.1 Patellar Tendon – Tibia Shaft Angle .................................................... 419 6.2 Whole-Body Movement Pattern ....................................................... 419 6.3 Muscle Activity Pattern ............................................................... 420 6.4 Muscle Reaction Time and Time to Peak Torque ........................................ 421 6.5 Muscle Stiffness and Muscle Strength .................................................. 422 6.6 Effects of Fatigue on Dynamic Stability ................................................. 423 7. Composite Model ........................................................................ 425 8. Conclusions and Future Research .......................................................... 425 The incidence of non-contact anterior cruciate ligament (ACL) injury is Abstract reported to be 6–8 times greater in females than males competing in the same activities. Injury to the ACL occurs as a result of insufficient stability of the tibiofemoral joint, which fails to prevent posterior dislocation of the femur on the tibia. The stability of the tibiofemoral joint is maintained by passive (non-contrac- tile) and dynamic (contractile) mechanisms. The passive mechanisms include the shape of the articular surfaces, the menisci, the ligaments and the joint capsule. The dynamic mechanisms consist of the muscle-tendon units that cross the joint, in particular, the quadriceps and hamstrings. The relative significance of the various passive and dynamic mechanisms in maintaining the stability of the

Upload: james-watkins

Post on 08-Dec-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: A Risk-Factor Model for Anterior Cruciate Ligament Injury

Sports Med 2006; 36 (5): 411-428REVIEW ARTICLE 0112-1642/06/0005-0411/$39.95/0

© 2006 Adis Data Information BV. All rights reserved.

A Risk-Factor Model for AnteriorCruciate Ligament InjuryGerwyn Hughes and James Watkins

Department of Sports Science, University of Wales Swansea, Swansea, UK

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4111. Anterior Cruciate Ligament (ACL) Injuries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4122. Sex Differences in ACL Injuries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4123. Anatomy of the Knee Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413

3.1 The Tibiofemoral Joint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4133.2 The Function of the Cruciate Ligaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414

4. Risk-Factor Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4145. Passive Stability Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

5.1 Geometry of the Articular Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4155.1.1 Alignment of the Femur with Respect to the Tibia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4155.1.2 Congruence Between the Articular Surfaces of the Femur and Tibia . . . . . . . . . . . . . . . . 416

5.2 Ligament Laxity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4176. Dynamic Stability Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419

6.1 Patellar Tendon – Tibia Shaft Angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4196.2 Whole-Body Movement Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4196.3 Muscle Activity Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4206.4 Muscle Reaction Time and Time to Peak Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4216.5 Muscle Stiffness and Muscle Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4226.6 Effects of Fatigue on Dynamic Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423

7. Composite Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4258. Conclusions and Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425

The incidence of non-contact anterior cruciate ligament (ACL) injury isAbstractreported to be 6–8 times greater in females than males competing in the sameactivities. Injury to the ACL occurs as a result of insufficient stability of thetibiofemoral joint, which fails to prevent posterior dislocation of the femur on thetibia. The stability of the tibiofemoral joint is maintained by passive (non-contrac-tile) and dynamic (contractile) mechanisms. The passive mechanisms include theshape of the articular surfaces, the menisci, the ligaments and the joint capsule.The dynamic mechanisms consist of the muscle-tendon units that cross the joint,in particular, the quadriceps and hamstrings. The relative significance of thevarious passive and dynamic mechanisms in maintaining the stability of the

Page 2: A Risk-Factor Model for Anterior Cruciate Ligament Injury

412 Hughes & Watkins

tibiofemoral joint is not clear. A number of intrinsic and extrinsic risk factors havebeen proposed to account for the sex difference in the incidence of ACL injuries.However, most of the proposed risk factors have arisen from univariate correla-tion studies based on relatively small samples.

1. Anterior Cruciate Ligament handball experts were used to identify possible risk(ACL) Injuries factors such as body position, type of movement,

with or without the ball, phase of play, balance,Anterior cruciate ligament (ACL) injuries are attention, speed and contact. Three physicians were

common and approximately 70% of these injuries used to identify factors relating to the knee positionoccur in sport.[1-3] ACL reconstruction was the sixth such as estimated flexion, rotation, varus-valgusmost common orthopaedic procedure performed in angle, foot position and degree of weight bearing onthe US in 1999 and 2000 with approximately the leg. The most common action being performed100 000 being performed annually at a cost of at the time of injury was a plant and cut movement>$US2 billion for surgery alone.[4-6]

(12 of the 20 cases analysed) accompanied by force-Up to two-thirds of patients who have complete ful valgus and external-internal rotation with the

ACL tears develop knee instability and, subsequent- knee close to full extension. Of these 12 injuries,ly, damage to the menisci and articular surfaces, four occurred during two-foot stance, eight were inwhich significantly affects knee function and leads single-foot stance and all occurred during the pushto a decrease in level of activity.[3,7-9] Noyes et al.[10]

off. The next most common injury mechanism was afound that in a group of individuals with rupture of one-legged jump shot landing, where athletes werethe ACL, 31% of patients reported overall difficulty jumping and landing on the same leg. This account-in walking alone, 44% had difficulties with activi- ed for four cases and all occurred with a forcefulties of daily living including walking and 77% had valgus, external rotation and knee close to full ex-difficulties with playing sport as a direct result of tension on landing. Therefore, it was concluded thattheir ACL injury. movements that put the ACL at risk involved a

Between 70% and 90% of ACL injuries have single foot stance, forceful valgus knee movementbeen reported to occur in non-contact situa- with the knee close to full extension accompaniedtions.[5,11,12] A non-contact situation is where there is with external or internal rotation of the tibia. Notno direct contact with the knee when the injury surprisingly, the incidence of ACL injury is relative-occurs.[13] Most ACL injuries appear to occur in ly high in sports such as basketball, netball, hand-situations involving one or more of the following ball, soccer and volleyball that are characterised by amanoeuvres: high frequency of landing, decelerating and rapid• foot strike with knee close to full extension[14,15]

changes of direction.[5,15,20]

• landing[16,17]

• deceleration[18]2. Sex Differences in ACL Injuries

• rapid change of direction.[19]

For example, in a study by Olsen et al.,[15] video The incidence of non-contact ACL injury in fe-tapes of game situations in which ACL injury oc- males has been reported to be 6–8 times greater thancurred in team handball were analysed to try to in males competing in the same sports.[20-28] Arendtidentify the mechanisms for ACL injury. Three and Dick[20] reported the incidence of ACL injury in

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 3: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 413

collegiate basketball and soccer for males and fe- tained and the role of the ACL within the tibi-males over the period 1989–93. Data were collected ofemoral joint.for 461 male and 278 female soccer teams and 531 The knee joint complex consists of two joints: themale and 576 female basketball teams. ACL injury patellofemoral joint and the tibiofemoral joint. Thewas reported in terms of athlete-exposure, where tibiofemoral joint is the articulation between theathlete-exposure took into account games and prac- proximal tibia and the distal femur. The patel-tise sessions. For soccer, female ACL injury inci- lofemoral joint is the articulation between the patelladence of 0.39 per 1000 athlete-exposures compared and the anterior femoral plateaus.with 0.13 per 1000 athlete-exposures for males werereported. For basketball, the incidence of ACL inju-

3.1 The Tibiofemoral Jointry was 0.29 per 1000 athlete-exposures for femalesand 0.07 per 1000 athlete-exposures for males. Ma-

The proximal articular surface of the tibiofemorallone et al.[25] documented the injuries of 402 malejoint is formed by large medial and lateral plateausand 385 female basketball players from 29 institu-of the distal end of the femur (figure 1). The twotions in three division 1 collegiate basketball confer-plateaus are separated by the intercondylar notch.ences over a 5-year period. Sixty-two females andThe articular surfaces of the proximal tibia thatnine males sustained ACL injury, which corre-correspond to the femoral articular surfaces are twosponded to an incidence of 16.1% in females andshallow concave medial and lateral plateaus.2.2% in males.

The tibiofemoral joint is stabilised by the interac-tion of dynamic and passive stabilisers. Dynamic3. Anatomy of the Knee Complexstability is provided by the muscles that cross the

To understand the possible mechanisms that con- tibiofemoral joint. These are primarily the quadri-tribute to ACL injury and therefore possible causes ceps, hamstrings and triceps surae. Passive stabilityof the sex difference in ACL injury, there is a need is provided by the non-contractile structures of theto have a brief understanding of the anatomy of the knee. These structures are the joint capsule, lateralknee complex, how stability of the joint is main- and medial menisci and four extracapsular liga-

Outlineof patella

Lateralligament

Lateralmeniscus

PCL

ACL

Medial ligament

Medial meniscus

PCLACL

a b

Fig. 1. (a) Anterior aspect of the right knee in flexion. (b) Normal orientation of the cruciate ligaments in the partially flexed knee. ACL =anterior cruciate ligament; PCL = posterior cruciate ligament.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 4: A Risk-Factor Model for Anterior Cruciate Ligament Injury

414 Hughes & Watkins

ments: lateral ligament, medial ligament, ACL and or dislocation of the tibia relative to the femur bythe posterior cruciate ligament (PCL) [see figure 1]. resisting forward movement of the tibial plateaus

(figure 3a and b). Similarly, the PCL works with theThe ACL and PCL are located in the intercondy-quadriceps to help prevent posterior dislocation oflar notch and cross each other with the ACL passingthe tibia relative to the femur by restricting back-laterally to the PCL. The distal end of the ACL isward movement of the tibial plateaus (figure 3a andattached to the posterior aspect of the anterior inter-c). When the knee is at deeper flexion angles, thecondylar area of the tibial table. The proximal end isline of action of the quadriceps changes so that itattached to the posterior medial aspect of the lateralacts with the hamstrings and the ACL in resistingfemoral plateau. The distal end of the PCL is at-forward movement of the tibial plateaus. This maytached to the posterior aspect of the intercondylarbe why ACL injury is uncommon when the knee isarea of the tibial table, and the proximal end ismore flexed.attached to the anterior inferior lateral aspect of the

medial femoral plateau (figure 1). In addition to helping to prevent posterior dislo-cation of the femur on the tibia (ACL) and anteriordislocation of the femur on the tibia (PCL), the3.2 The Function of the Cruciate Ligamentscruciate ligaments also help prevent hyperextensionof the tibiofemoral joint, medial and lateral displace-The function of the cruciate ligaments is to en-ment of the tibia relative to the femur and internalsure normal movement between the articular sur-rotation of the tibia relative to the femur. At present,faces of the femoral and tibial plateaus. Figure 2there is little empirical information on the extent toshows a free-body diagram of the forces acting onwhich landing, decelerating and cutting movementsthe proximal tibia when the knee is close to fullcause abnormal movement of the tibiofemoral jointextension.and, therefore, abnormal strain on the ACL andFrom figure 2, it can be seen that when the kneeother knee ligaments.is close to full extension, i.e. the position in which

most non-contact ACL injuries occur, the ACL and4. Risk-Factor Modelsthe hamstrings work together to help prevent anteri-

The risk factors associated with ACL injury ingeneral and the sex difference in ACL injury inci-dence in particular have been grouped in variousways.[5,6,13,29-31] One of the most frequently usedcategorisations of injury risk factors is intrinsic-extrinsic. Intrinsic factors are personal, physical andpsychological characteristics that distinguish indi-viduals from each other and extrinsic factors con-cern environmental conditions and the manner inwhich activities are administered.[32] With regard toACL injuries, Griffin et al.[5] categorised risk factorsinto three intrinsic groups (anatomical, hormonaland biomechanical) and one extrinsic group (envi-ronmental). While these categories may be of somehelp in identifying the possible cause of an injury, in

H A P Q

Fig. 2. Free-body diagram of forces acting on the proximal tibia inthe sagittal plane. Forces due to the hamstrings (H), quadriceps(Q), anterior cruciate ligament (A) and posterior cruciate ligament(P).

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 5: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 415

a b c

PCL ACL

Fig. 3. (a) Normal orientation of the cruciate ligaments in the partially flexed knee; (b) anterior cruciate ligament (ACL) tear due to posteriordislocation of the femur on the tibia; (c) posterior cruciate ligament (PCL) tear due to anterior dislocation of the femur on the tibia.

most cases the cause of the injury is likely to be the between the articular surfaces of the femur and tibiaresult of a complex interaction of intrinsic and ex- (figure 5).trinsic factors.[29] The main limitation of current

5.1.1 Alignment of the Femur with Respect tomodels of the aetiology of injury in general and the TibiaACL injury in particular is the failure to show how There is considerable evidence that the Q angle,the various risk factors interact. A reasonable first i.e. the acute angle between the line connectingstep in trying to understand the interaction between anterior superior iliac spine to the middle of thefactors would be to describe the relationship be- patella and the line connecting the tibial tuberositytween the various factors. ACL injury is caused by to the centre of the patella[33] is, on average, larger inexcessive load on the ACL, which is due to abnor- females than males[34-38] (figure 6).mal movement of the tibiofemoral joint. The latter is

For example, Herrington and Nester[38] measureddue to the failure of the passive and/or dynamic

the Q angles of 51 male and 58 female physicallysupport mechanisms to adequately stabilise the joint

active subjects with no history of lower-limb injury.(figure 4). This article presents a risk-factor model

Q angle was measured with subjects standing andfor ACL injury based upon a review of the compo-

quadriceps relaxed. The results showed mean Qnents of the passive and dynamic support mecha-

angle to be significantly greater in females (mean Qnisms.

angle 13.9°) than in males (mean Q angle 11.5°).

5. Passive Stability Risk Factors

As shown in figure 5, the passive stability of thetibiofemoral joint depends upon the geometry of thearticular surfaces and ligament laxity.

5.1 Geometry of the Articular Surfaces

The geometry of the articular surfaces of thetibiofemoral joint depends upon the alignment of thefemur with respect to the tibia and the congruence

ACL injury

Load on ACL

Passive stability Dynamic stability

Stability oftibiofemoral joint

Type and range of motionin tibiofemoral joint

Fig. 4. Model of anterior cruciate ligament (ACL) injury based onpassive and dynamic stability of the tibiofemoral joint.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 6: A Risk-Factor Model for Anterior Cruciate Ligament Injury

416 Hughes & Watkins

Passive stability

Q angle Size and shape ofthe condylar surfaces

Size and shapeof the menisci

Cross-sectionalarea of ligaments

Material propertiesof the ligaments

Circulatinghormones

Ligamentlaxity

Geometry of thearticular surfaces

Alignment of thetibia with respect

to the femur

Congruence between thearticular surfaces of the

femur and the tibia

Length ofligaments

Tensile stiffnessof ligaments

Fig. 5. Factors affecting passive stability of the tibiofemoral joint.

The larger the Q angle, the larger the knee valgus (shallow concave in the sagittal plane) [figure 3] areangle (figure 6). Since increased valgus angle during not very congruent, but in the healthy knee, thedynamic movement has been associated with an congruence between the femoral and tibial plateausincreased likelihood of ACL injury,[14,15] some stud- is normally quite high due to the menisci (figure 1a).ies have investigated the relationship between Q Damage to the menisci, especially complete radialangle and ACL incidence. For example, Shambaugh tears, has been shown to decrease congruence andet al.[39] investigated the relationship between lower decrease passive stability of the joint.[40] However,extremity alignment and knee injury in 45 recrea- there would appear to be no empirical evidence oftional athletes. The results showed that athletes who damaged menisci increasing the incidence of ACLhad sustained knee injuries had significantly larger injury. The congruence between the femoral andQ angles than the players who had not been injured tibial plateaus is also affected by the width of the(mean Q angle in knee injured 14°, non-injured 10°). intercondylar notch (INW); the wider the notch, theHowever, other studies have demonstrated no appar- lower the congruence (figure 7).ent relationship between Q angle and ACL injury,[22]

However, the narrower the notch, the smaller theindicating little evidence that the increased Q anglespace available for the cruciate ligaments. Somein females increases the risk of ACL injury. Also,studies have reported that females have a smallerthe apparent valgus angle of the knee observed whenINW[41-43] and NWI (ratio of INW to width of femo-subjects are performing cutting and landing ma-ral epicondyles)[44] than males. Also, a number ofnoeuvres is likely to be different than the static Qstudies have reported greater incidence of unilateralangle measured anatomically.and bilateral ACL injury in subjects with smallerINW[41,43,45,46] and NWI.[44,46] For example, Uhor-5.1.2 Congruence Between the Articular Surfaceschak et al.[43] carried out a 4-year study of 711 maleof the Femur and Tibiaand 113 female army cadets. INW was measured atCongruence between the articular surfaces of thethe start of the 4-year period via radiographic assess-femur and tibia depends upon the size and shape ofment using digital callipers. During the 4-year peri-the plateau surfaces of the femur and tibia and theod, 24 non-contact ACL injuries occurred, 16 tosize and shape of the menisci. The femoral plateaus

(convex in the sagittal plane) and tibial plateaus males and 8 to females. The results showed INW to

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 7: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 417

be significantly smaller in females (mean INW

15.6mm) compared with males (mean INW

17.7mm) and INW to be significantly smaller in

non-contact ACL-injured subjects (mean INW

13.8mm) than in non-injured subjects (mean INW

17.5mm). Also, Shelbourne et al.,[41] investigated

the relationship of the INW with bilateral ACL

injury incidence and the rupture rate of the recon-

structed ACL. This was to identify whether subjects

who had a narrow INW were more likely to rupture

either the reconstructed ACL or the ACL of the

other knee than those subjects with a wide INW. 714

a

b

Fig. 7. Intercondylar notch width (a) and notch width index (a/b).

subjects who underwent patella tendon graft ACLreconstruction were investigated and divided intonarrow (<15mm) and wide (>16mm) INW groups.The findings showed females to have narrower INWthan males of the same height and the incidence ofbilateral rupture after initial rupture of the otherACL was found to be higher for the narrow INWgroup than for the wide INW group. However, thefindings showed no difference between males andfemales for contralateral ACL injury incidence andno difference in rupture rate of the reconstructedACL between the narrow and wide groups. Also,other studies have not found an association betweenACL injury and INW or NWI[47-49] and there is noclear evidence that differences in INW or NWIinfluence the incidence of ACL injury.

5.2 Ligament Laxity

Joint laxity refers to the degree of instability in ajoint, i.e. the range of movement in directions thatare considered abnormal for that joint when nomuscles are active.[32] Several studies have reportedthat females exhibit greater joint laxity thanmales[50,51] and knee joint laxity has been proposedas potentially contributing to the greater incidenceof ACL injuries in females.[22,43,50] It is not clearwhat causes differences in joint laxity in healthy

Anterior superiorilliac spine

Qangle

Valgus angleCentre ofpatella

Tibialtuberosity

Fig. 6. Q angle and valgus angle of the right leg.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 8: A Risk-Factor Model for Anterior Cruciate Ligament Injury

418 Hughes & Watkins

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Study 1

Lower incidenceHigher incidence Days of the menstrual cycle

3

4

2

Fig. 8. Reported days of significantly higher and lower incidence of anterior cruciate ligament injury. Study 1 = Mykelbust et al.;[12] study 2 =Wojtys et al.;[57] study 3 = Wojtys et al.;[58] study 4 = Slauterbeck et al.[59]

knees, but it is reasonable to assume that it is due to injuries in females. Due to the identification ofdifferences in the length and tensile stiffness (resis- oestrogen and progesterone receptors in the humantance to stretching) of ligaments (figure 5). The ACL,[53] some studies have investigated the effectslonger the ligaments and the lower the tensile stiff- of the variations in the levels of female sex hor-ness, the greater the laxity. The tensile stiffness of a mones on the material properties of ligaments.[54-56]

ligament will depend upon the cross-sectional area For example, Slauterbeck et al.[55] reported that theand the effect of circulating hormones on the materi- administration of oestrogen significantly decreasedal properties of the ligament. Some studies have the tensile strength of the ACL in rabbits. Thesereported that the cross-sectional area of the ACL is findings suggest that the levels of female sex hor-relatively larger in males than in females.[42,52] For mones may affect the strength of the ACL, whichexample, Charlton et al.[42] used magnetic resonance may provide some explanation for the greater inci-imaging to measure ACL volume in 98 knees of dence of ACL injury in females. Consequently, sev-males and females and found that ACL volume was eral studies have investigated the time of occurrencesignificantly relatively smaller in females than of non-contact ACL injury in females in relation tomales. However, there is no evidence that cross- the phase of the menstrual cycle;[12,57-59] however,sectional area of the ACL influences the incidence the findings are inconsistent (figure 8). Some studiesof ACL injury. The effect of hormones on the mate- reported significantly higher incidence of ACL inju-rial properties of the ACL has been identified as a ry between days 10–14,[57,58] whereas others report-potential cause of the greater incidence of ACL ed significantly higher incidence during days 1–2 of

Patellar tendon tibiashaft angle

Alignment of thetibiofemoral joint

Whole-bodymovement pattern

Muscle stiffness

Muscle strength

Fatigue

Time topeak torque

Hormones

Dynamic stability

Musclereaction time

Muscle function

Proprioception

Muscleactivity pattern

Training

Material properties ofconnective tissues

Fig. 9. Factors that affect dynamic stability of the tibiofemoral joint.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 9: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 419

force applied to the proximal end of the tibia(ASPT). The greater the PTTSA, the greater theASPT and the greater the potential strain on theACL.

Nunley et al.[60] investigated the effect of kneeangle on the PTTSA. Regression analysis showedthat the PTTSA decreased linearly in both males andfemales from 0 to 90° of knee flexion (table I), i.e.the PTTSA and, therefore, the ASPT was largestwhen the knee was close to full extension. A numberof studies including Boden et al.[14] and Olsen etal.[15] have reported that non-contact ACL injury

a

ba

b

Fig. 10. Effects of knee flexion angle (a) on patella tendon tibiashaft angle (b).

appears to occur more frequently when the knee isclose to full extension than when flexed. Nunley etthe menstrual cycle.[59] Also, days of significantlyal.[60] also showed that the PTTSA was, on average,lower incidence of ACL injury have been reported3.6° greater in females than in males over the 0–90°between days 1–9,[57] days 8–14[12] and days 15–28range of knee flexion, i.e. for a given angle of kneeof the menstrual cycle.[58]

flexion, the ASPT is likely to be 13.0% greater inThe inconsistency of these findings may be at-females than in males over the 0–90° range of kneetributable to the relatively small samples used. How-flexion. Therefore, the PTTSA may contribute aever, on current evidence, the influence of changesgreater risk of ACL injury in females than in males.in hormone concentrations on the incidence of ACL

injuries in females is not clear.6.2 Whole-Body Movement Pattern

6. Dynamic Stability Risk FactorsThe patellar tendon-tibia shaft angle depends up-

During dynamic movements such as landing and on the knee angle which, in turn, depends upon thecutting (side-stepping), dynamic stability in the whole-body movement pattern. Kinematic and ki-form of muscle activity is necessary to provide netic analyses of drop-jumps, stop-jumps and cut-adequate joint stability. Figure 9 shows the factors ting movements have been undertaken in order to trythat affect muscle function and, therefore, the dy- to identify movement characteristics that may ac-namic stability of the tibiofemoral joint: the patellar count for sex differences in the incidence of ACLtendon-tibia shaft angle, muscle activity pattern (interms of net joint torque), muscle reaction time, timeto peak torque and muscle stiffness.

6.1 Patellar Tendon – Tibia Shaft Angle

The patella tendon-tibia shaft angle (PTTSA) isthe angle (in the sagittal plane) between the longaxis of the tibia and the line of action of the patellarligament (figure 10). When the knee is close to fullextension, contraction of the quadriceps, which actsthrough the patellar tendon, causes an anterior shear

Table I. Mean and range of patellar tendon-tibia shaft angle (PTT-SA) in male and female recreational athletes over the 0–90° rangeof knee flexion[60]

Mean (°)a Knee flexion (°) Range of PTTSA (°)

Males

22.0 0 12.2 to 27.8

90 –11.3 to –0.1

Females

25.7 0 13.3 to 34.8

90 2.1 to 5.4

a Mean PTTSA over the 0–90° range of knee flexion.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 10: A Risk-Factor Model for Anterior Cruciate Ligament Injury

420 Hughes & Watkins

Table II. Summary of recent studies that investigated the kinematics and kinetics of landing and cutting manoeuvres in males and females

Study Subjects Level Task Results

Salci et al.[63] 8 F, 8 M University Vertical landing F displayed reduced knee and hip flexion angles at groundvolleyball contact, greater peak knee extension moment and greater

normalised ground reaction force

Decker et al.[65] 9 F, 12 M Recreational Vertical landing F had reduced knee flexion at ground contact, greater rangevolleyball and of motion and greater peak angular velocities of ankle, hipbasketball and knee in sagittal plane. F showed greater energy

absorption and peak powers in the knee extensors andankle plantar flexors

Chappell et al.[62] 10 F, 10 M Recreational Forward, backward F landed with reduced knee flexion and had greater proximaland vertical jump tibia anterior shear force, greater knee extension momentslanding and greater knee valgus moments

Ford et al.[66] 47 F, 34 M High-school Vertical drop jump Increased knee valgus motion and maximum angle in Fbasketball

Malinzak et al.[61] 9 F, 11 M Recreational Cutting F displayed reduced flexion, greater knee valgus angle andgreater quad activation

James et al.[64] 19 M, 19 F High-school Cutting F displayed greater ground reaction force at maximum kneeand collegiate flexion, reduced knee flexion at ground contact and greaterbasketball range of knee flexion during stance phase

F = females; M = males.

injuries. A summary of these studies can be seen in tibia and, therefore, minimal strain on the kneetable II. In general, the studies indicate that in land- ligaments. However, if the shear load exerted by theing and cutting manoeuvres, females tend to land quadriceps is greater than the shear load exerted bywith less knee flexion,[61-65] exhibit greater knee the hamstrings, a resultant anteriorly directed shearvalgus,[61,66] display greater peak knee extension force may be exerted on the proximal end of themoments[62,63] and produce greater normalised tibia, which will increase ACL strain (figure 11).ground reaction force[63,64] when compared with This is known as quadriceps dominance and is de-males. As described in section 1, these types of fined as an imbalance between the recruitment pat-movement are likely to increase the load on the terns of the knee flexors and extensors.[66] A numberACL. These findings tend to indicate differences in of studies have found that females exhibit greaterlanding and cutting movement patterns between quadriceps dominance than males in activities asso-males and females. However, lack of appropriate ciated with ACL injury.[51,61,67]

standardisation in task demands may have invalidat- Malinzak et al.[61] investigated 3-dimensionaled meaningful comparison between females and knee joint motion and electromyographic (EMG)males. For example, dropping down from a raised

activities of the hamstrings and quadriceps in 9platform set at the same height for both males and

female and 11 male recreational athletes while run-females[63,65,66] may result in significantly different

ning, cross-cutting and side-cutting. Surface EMGtask demands.

signals were recorded during maximum voluntarycontraction (MVC) tests for vastus lateralis and vas-

6.3 Muscle Activity Patterntus medialis oblique and for the biceps femoris andmedial hamstrings so that EMG data could be ex-When providing dynamic stability of the knee,pressed as a percentage of the EMG recorded forthe activity of the knee flexors and extensors should,

ideally, result in a zero shear load on the proximal MVC. The EMG data indicated that females had

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 11: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 421

greater quadriceps activation (17–40% MVC information and further research is necessary toclarify this possible sex difference.greater) yet reduced hamstring activation (20%

MVC less) than males. Huston and Wojtys[51] inves-6.4 Muscle Reaction Time and Time totigated knee activation patterns in response to anteri-Peak Torqueor tibial translation in elite basketball and volleyball

players. Females athletes tended to respond to ante-Muscle reaction time (RT) and time to peakrior tibial displacement of the knee by firstly con-

torque (TPT) clearly affect the speed with whichtracting the quadriceps, whereas male athletes anddynamic stability can be achieved and, as such, maymale and female non-athletes responded by firstlyaffect the likelihood of joint injury.[68] Cowling andcontracting the hamstrings. Zeller at al.[67] investi-Steele[69] investigated the effect of sex on lower limbgated the kinematics and EMG activity of the quad-muscle synchrony during landing in 7 male and 11riceps and hamstrings of nine male and nine femalefemale subjects. The task involved subjects acceler-collegiate athletes while performing a one-leggedating forwards three steps to receive a netball chestsquat using the dominant limb. Analysis of all mus-pass from 3m directly in front and then land on theircles tested showed females to display greater totaldominant limb on a force platform in single-limbmuscle activation than males. Analysis of each mus-stance. EMG data were recorded, from the rectuscle independently showed females to produce sig-femoris, vastus medialis, vastus lateralis, semimem-nificantly greater total and maximal activation of thebranosus, biceps femoris and the medial head of therectus femoris compared with males.gastrocnemius muscles. The main findings were that

The findings of these studies[51,61,67] suggest that, males exhibited a delayed semimembranosus onsetcompared with males, females tend to exhibit a relative to ground contact (males 113 ± 46ms, fe-quadriceps dominant mode of producing dynamic males 173 ± 54ms) and delayed peak activity rela-joint stability of the knee, which may increase the tive to peak tibiofemoral shear force (males 54 ±risk of ACL injury. However, there is little empirical 27ms, females 77 ± 15ms) compared with females.

It was suggested that the delayed onset of semimem-branosus activity in males allows peak semimem-branosus activity to better coincide with high anteri-or force exerted on the proximal tibia by the quadri-ceps, thereby acting as an ACL synergist viaincreased joint compression and posterior tibialdrawer, reducing the chance of ACL injury.

Wojtys et al.[70] investigated the effect of threetypes of training on RT and TPT in 16 male and 16female healthy subjects. Subjects were placed intoone of four groups; isokinetic, isotonic, agility andcontrol. The three training groups exercised for 30minutes, three times a week for 6 weeks, whereasthe control group participated only in activities theywere involved in prior to the study. RT was assessedas the muscular response to an anteriorly directed133.42N (30-pound) force applied to the posterior

H

Q

PHH

HV ACL

PV

PH

a b

Fig. 11. Forces exerted by the quadriceps and hamstrings on thetibia. (a) Relationship of the hamstrings (H), quadriceps (Q) andpatellar ligament (P) to the anterior cruciate ligament (ACL). (b)Components of H and P in relation to the ACL: if PH is greater thanHH, the ACL will be put under strain. HH = horizontal component ofH; HV = vertical component of H; PH = horizontal component of P;PV = vertical component of P.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 12: A Risk-Factor Model for Anterior Cruciate Ligament Injury

422 Hughes & Watkins

aspect of the tibia, with the subject sitting with the ry. As a result of this, a number of studies haveknee flexed at 30°. Surface EMG was recorded for investigated the muscular stiffness of males andthe gastrocnemius, lateral hamstring, medial ham- females to determine its link to the sex difference instring, lateral quadriceps and medial quadriceps ACL injury incidence.[71-73]

muscles. The readings from these muscles were then Granata et al.[71] investigated the muscle stiffnessused to calculate three components of RT; spinal of the quadriceps and hamstrings in healthy subjectsreflex (between 26–130ms after tibial displace-

(12 male, 11 female). Subjects were required toment), intermediate response (between 110–216ms

support their lower leg at 45° in two positions. Theafter tibial displacement), and voluntary muscle ac-

first required the hamstring to support the weight oftivity (between 156–431ms after tibial displace-

the leg and the other required the quadriceps toment). The findings showed that the agility exer-

support the weight of the leg. Weights of 0kg, 6kgcises resulted in a significant reduction in all threeand a weight corresponding to 20% of maximumcomponents of RT. Isokinetic training resulted in avoluntary exertion were added to a fixed ankle-footsignificant decrease in the voluntary muscle activityorthosis. A sudden transient perturbation was thencomponent of RT. No significant improvements inapplied to the lower leg and the resulting kneeany of the components of reaction time were seen asflexion-extension motion recorded by an accelerom-a result of isotonic training. All groups showed aeter attached to the heel of the orthosis. Subjectsreduction in TPT in knee extension isokineticwere asked not to attempt to control the naturalstrength test, with the agility group showing themotion of the tibia after the perturbation and tobiggest reduction (39ms). For knee flexion, the agil-maintain constant muscle activity, which was mea-ity group again showed the greatest reductionsured by EMG electrodes on the quadriceps and(38ms) after training. For ankle plantar flexion dur-

ing isokinetic strength testing, TPT reduced by hamstrings. The flexion-extension oscillations of15ms for the isokinetic group, with the other two the knee were then used to calculate muscle stiff-training groups showing an increase. The findings of ness. After accounting for differences in appliedthis study show that agility and isokinetic training moment resulting from leg mass, females demon-may reduce anterior tibial translation and therefore strated reduced muscle stiffness for the quadricepsreduce the risk of ACL injury. and the hamstrings for all torque levels compared

with males, with results showing females’ stiffness6.5 Muscle Stiffness and Muscle Strength to be between 55.8–73.9% of males.

Granata et al.[72] continued the work of the previ-As the quadriceps and hamstring muscles con-ous study to evaluate whether females also demon-tract, they act in a way to increase the joint contactstrated reduced leg stiffness in functional tasks, suchforces and limit shear movement within the tibi-as two-legged hopping, compared with males. Fif-ofemoral joint. The ability of the muscles to resistteen male and fifteen female healthy subjects tookmovement within the tibiofemoral joint (maintain apart in the study. Subjects were required to performparticular joint angle) refers to muscle stiffness. Therepeated two-legged hopping on a force platform.greater the ability of the muscles of the knee toSubjects were required to hop at three differentprevent tibiofemoral shear movement, the less likelyhopping frequencies of 2.5Hz, 3.0Hz and preferredthe passive structures of the knee, such as the ACL,frequency. Leg stiffness was found to be significant-will be put under strain. Therefore, muscle stiffnessly greater in males than females at all hoppingmay be an important factor in preventing ACL inju-

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 13: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 423

frequencies with females exhibiting 73–81% of the ings and the quadriceps in females comparedleg stiffness of males with males, even when normalised to body-

weight.[51,63,74,75] This suggests that the lower levelsWojtys et al.[73] investigated the hypothesis thatof muscle stiffness exhibited by females comparedfemales are less able to volitionally increase thewith males may be due, at least in part, to lowerapparent torsional stiffness of the knee by maximal-levels of muscle strength (absolute and relative).ly activating the knee muscles. Torsional stiffness isLower levels of strength may increase the risk ofthe ability of the muscles of the knee to preventACL injury.rotation of the knee joint. There were two groups of

subjects. The first group consisted of 12 male and 126.6 Effects of Fatigue on Dynamic Stabilityfemale National Collegiate Athletic Association Di-

vision 1 athletes competing in basketball, volleyballFatigue has been suggested as a risk factor for

and soccer. The second group of subjects was 14non-contact ACL injury[76,77] because of its affect on

male and 14 female collegiate endurance athleteslower extremity muscle activity. However, the find-

competing in cycling, crew and running. Testingings are conflicting. Chappell et al.[76] investigated

was performed using a weighted pendulum, whichthe effects of fatigue on knee-joint kinetics and

applied an 80N force, directed medially, to the later-kinematics of ten male and ten female recreational

al aspect of the right forefoot. The resulting internalathletes whilst performing drop-jump tasks. The

rotation was measured optically. Trials were carriedsubjects performed three drop-jump tasks before

out with and without maximal activation of theand after undertaking repeated 30m sprints and ver-

quadriceps and hamstring. The maximal rotations oftical jumps to fatigue the subjects’ lower limbs. The

the leg were greater in females than males for trialsfindings showed that both males and females signifi-

with (27% greater) and without (16% greater) acti-cantly increased the peak proximal tibia anterior

vation of the knee muscles. Also, females displayedshear force and decreased knee flexion on landing

significantly smaller (18%) volitional increase inwhen fatigued compared with non-fatigued. Also,

torsional stiffness of the knee than males. This find-after fatigue, males displayed decreased knee varus

ing was particularly evident in those athletes inmoment, whereas females showed increased knee

group 1, where there was a 42% difference betweenvalgus moment.

sexes.Fagenbaum and Darling[77] investigated the effect

The findings of the studies examining the muscle of fatigue on knee kinematics and muscle activationstiffness of the knee suggest that females exhibit less during jump landings for six male and eight femalemuscular protection of the knee than males when the university basketball players. Three landing tasksknee is externally loaded. This suggests that sex were performed, one where subjects jumped fromdifferences in the muscular stiffness of the knee may both feet as high as possible and landed on theaccount, at least in part, for the greater incidence of dominant leg, one where subjects jumped downACL injury in females. However, further investiga- from a 25.4cm high platform and landed on thetion into the apparent variation in the stiffness of the dominant leg, and finally where subjects jumpedknee between males and females of varying athletic down from a 50.8cm high platform and landed onlevels is needed to validate this. the dominant leg. Each of these tasks were per-

One factor that may influence muscle stiffness is formed before and after subjects were fatigued bymuscle strength. A number of studies have reported using an isokinetic dynamometer. Subjects under-significantly lower muscular strength of the hamstr- took knee flexion-extension exercises on the dyna-

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 14: A Risk-Factor Model for Anterior Cruciate Ligament Injury

424 Hughes & Watkins

Pat

ella

r te

ndon

tibia

sha

ftan

gle

Mus

cle

activ

itypa

ttern

Alig

nmen

t of t

hetib

iofe

mor

al jo

int

Who

le-b

ody

mov

emen

t pat

tern

Fatig

ue

Tim

e to

peak

torq

ueM

uscl

est

iffne

ss

Mus

cle

stre

ngth

Hor

mon

es

AC

L in

jury

Load

on

AC

L

Type

and

ran

ge o

f mot

ion

in ti

biof

emor

al jo

int

Sta

bilit

y of

tibi

ofem

oral

join

t

Dyn

amic

sta

bilit

y

Mus

cle

func

tion

Pro

prio

cept

ion

Trai

ning

Mat

eria

lpr

oper

ties

of c

onne

ctiv

etis

sues

Pas

sive

sta

bilit

y

Q a

ngle

Siz

e an

dsh

ape

of th

eco

ndyl

ar s

urfa

ces

Siz

e an

dsh

ape

ofth

e m

enis

ci

Cro

ss-s

ectio

nal

area

of

ligam

ents

Mat

eria

lpr

oper

ties

ofth

e lig

amen

ts

Circ

ulat

ing

horm

ones

Liga

men

tla

xity

Geo

met

ry o

f the

artic

ular

sur

face

s

Alig

nmen

t of t

hetib

ia w

ith r

espe

ctto

the

fem

ur

Con

grue

nce

betw

een

the

artic

ular

surf

aces

of t

hefe

mur

and

the

tibia

Leng

th o

flig

amen

tsTe

nsile

stiff

ness

of

ligam

ents

Mus

cle

reac

tion

time

Fig

. 12.

Pro

pose

d ris

k-fa

ctor

mod

el o

f ant

erio

r cr

ucia

te li

gam

ent (

AC

L) in

jury

bas

ed o

n pa

ssiv

e an

d dy

nam

ic s

tabi

lity

of th

e tib

iofe

mor

al jo

int.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 15: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 425

mometer until they were fatigued to two levels. that fatigue has a greater effect on the incidence ofACL injury in females compared with males.These were when subjects could no longer produce

50% and then 25% of maximum extensor torque.7. Composite ModelEMG data for the hamstrings and quadriceps were

collected along with knee flexion angle, which was Figure 12 shows a composite model of the pas-measured using a potentiometer incorporated into sive and dynamic risk factors that affect the stabilitytwo freely pivoting orthoplast lever arms placed on of the tibiofemoral joint. It is reasonable to assumethe lateral aspect of the dominant knee. The findings that the apparent greater incidence of ACL injury inshowed that females exhibited greater knee flexion females compared with males is due to the sexangle at ground contact than males. Also, females differences regarding some or all of the passive and

dynamic stability risk factors. The only evidencedisplayed greater knee flexion accelerations as a(univariate correlation based on small samples) inresult of ground contact than males, contradictory tosupport of sex differences regarding some risk fac-a number of other studies.[61-65] These findings weretors, such as Q angle, joint laxity, INW, NWI, ACLshown for all jumps and all fatigue levels. Quadri-cross-sectional area and changes in concentration ofceps and hamstring activities were similar in bothcirculating hormones, is fairly weak. However, themales and females.evidence in support of sex differences is much

Wojtys et al.[78] investigated the effect of ham-stronger regarding some of the dynamic stability

string and quadriceps muscle fatigue on anteriorfactors, especially PTTSA, muscle activity pattern,

tibial translation and muscle reaction time in six time to peak torque and muscle stiffness.male and four female healthy subjects. Subjectsundertook knee examination, arthrometer measure- 8. Conclusions and Future Researchments of tibial translation, subjective functional as-

ACL injury is a common injury which frequentlysessment and an anterior tibial translation stress testoccurs in sport. ACL injuries occur most frequentlybefore and after fatiguing exercise. The muscle re-in non-contact situations with the athlete in single

cruitment patterns of the hamstrings, quadriceps andfoot stance, forceful valgus knee collapse with knee

gastrocnemius demonstrated no change between fa-close to full extension accompanied with external or

tigued and non-fatigued tests of anterior tibial trans-internal rotation of the tibia. The situation appears to

lation. However, the results showed a 32.5% in-occur in two particular types of movement: landing

crease in anterior tibial translation for the fatigued from a jump on one leg and rapid change of direc-test compared with the non-fatigued test. Muscle tion initiated on one leg, such as a side-cutting andresponses originating in the spinal cord and cortical cross-cutting manoeuvre. ACL injury occurs as alevel exhibited significant slowing after fatigue of result of a lack of stability provided by the dynamicthe hamstrings and quadriceps. These findings sug- and passive stability mechanisms of the knee. Cur-gest that muscle fatigue alters the neuromuscular rent evidence suggests that the greater incidence ofresponse to anterior tibial translation, which, in turn, ACL injury in females is due to sex differences inmay reduce the dynamic stability of the knee. In the dynamic stabilising structures rather than thegeneral, the results of studies on the effect of fatigue passive stability structures. Consequently, future re-on lower extremity kinematics and kinetics suggest search should investigate these factors and, in par-that fatigue may increase the risk of ACL injury. ticular, the interaction between the factors that affectHowever, there is currently no evidence to suggest dynamic stability of the knee.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 16: A Risk-Factor Model for Anterior Cruciate Ligament Injury

426 Hughes & Watkins

17. Otago L, Neal RJ. Netball pivot landings: is there a correct way?AcknowledgementsIn: Abstracts of the Australian Conference of Science andMedicine in Sport; 1997 Oct 7-10; Canberra. Bruce (ACT):Sports Medicine Australia, 1997: 262-3The authors have not received any financial benefit for the

18. Miller MDM, Cooper DE, Warner JJP. Review of sportsresearch in this review. The authors have no potential con-medicine and arthroscopy. Philadelphia (PA): WB Saunders,flicts of interest relevant to the contents of this review.1995

19. Bartold SJ. Injury driven change to the fundamental designparameters of the Australian rules football boot. In: Abstractsof the Australian Conference of Science and Medicine inReferencesSport; 1997 Oct 7-10; Canberra. Bruce (ACT): Sports1. Faegin JA. Isolated anterior cruciate injury. In: Faegin JA,Medicine Australia, 1997: 52-3editor. The crucial ligaments. New York: Churchill Living-

stone, 1988: 15-23 20. Arendt EA, Dick R. Knee injury patterns among man and2. Johnson RJ. Prevention of anterior cruciate ligament injuries. In: women in collegiate basketball and soccer. Am J Sports Med

Faegin JA, editor. The crucial ligaments. New York: Churchill 1995; 23: 694-701Livingstone, 1988: 349-56 21. Chandy TA, Grana WA. Secondary school athletic injury in

3. Smith BA, Livesay GA, Woo SLY. Biology and biomechanics boys and girls: a three-year comparison. Phys Sportsmed 1985;of the anterior cruciate ligament. Clin Sport Med 1988; 12: 13: 314-6637-46 22. Gray J, Taunton JE, McEnzie DC, et al. A survey of injuries to

4. Gottlob CA, Baker CL, Pellissier JM, et al. Cost effectiveness of the anterior cruciate ligament of the knee in female basketballanterior cruciate ligament reconstruction in young adults. Clin players. Int J Sports Med 1985; 6: 314-6Orthop 1999; 367: 378-82

23. Ferretti A, Papandrea P, Conteduca F, et al. Knee ligament5. Griffin LY, Angel J, Albohm MJ, et al. Noncontact anterior

injuries in volleyball players. Am J Sports Med 1992; 20 (2):cruciate ligament injuries: risk factors and prevention strategy.

203-7J Am Acad Orthop Surg 2000; 8 (3): 141-50

24. Paulos LE. Why Failures Occur Symposium: revision ACL6. Huston LJ, Greenfield ML, Wojtys EM. Anterior cruciate liga-surgery. Presented at the American Orthopedic Society forment injuries in the female athlete: potential risk factors. ClinSports Medicine, 18th annual meeting; 1992 Jul 6–9; SanOrthop 2000; 372: 50-63Diego

7. Satku K, Kumar VP, Ngoi SS. ACL injuries: to counsel or to25. Malone TR, Hardaker WT, Garrett WE, et al. Relationship ofoperate? J Bone Joint Surg Br 1986; 68: 458-61

gender to anterior cruciate ligament injuries in intercollegiate8. Kannus P, Jarvinen M. Conservatively treated tears of the ACL:basketball players. J South Orthop Assoc 1993; 2: 36-9long term results. J Bone Joint Surg Am 1987; 69: 1007-12

26. Lindenfeld TN, Schmitt DJ, Hendy MP, et al. Incidence of9. Irvine LB, Glasgow MM. The natural history of the meniscus ininjury in indoor soccer. Am J Sports Med 1994; 22 (3): 354-71the anterior cruciate insufficiency. J Bone Joint Surg Am 1992;

74: 403-5 27. Ireland ML. Special concerns of the female athlete. In Fu FH,Stone DA, editors. Sports injuries: mechanism, prevention and10. Noyes FR, Mooar PA, Mathews DS, et al. The symptomatictreatment. 2nd ed. Philadelphia (PA): Williams & Wilkins,ACL-deficient knee. J Bone Joint Surg Am 1983; 65: 154-741994: 153-8711. McNair P, Marshall R, Matheston J. Important features associat-

ed with acute anterior cruciate injury. N Z Med J 1993; 103: 28. Gwinn DE, Wilckens JH, McDevitt ER, et al. The relative537-9 incidence of Anterior cruciate ligament injury in men and

12. Mykelbust G, Maehlum S, Engbretsen L, et al. Registration of women at the United States naval academy. Am J Sports Medcruciate ligament injuries in Norwegian top level team hand- 2000; 28: 98-102ball: a prospective study covering two seasons. Scand J Med 29. Lysens R, Steverlynck A, Van den Auweele Y, et al. TheSci Sports 1997; 7: 289-92 predictability of sports injuries. Sports Med 1984; 1: 6-10

13. Yu B, Kirkendall DT, Garrett WE. Anterior cruciate ligament 30. Lorenzton R. Intrinsic factors. In: Ditrix A, Knuttgen HG, Tittelinjuries in female athletes: anatomy, physiology and motor K, editors. The Olympic book of sports medicine. Oxford:control. Sports Med Arthrosc Rev 2002; 10: 58-68 Blackwell Scientific, 1988

14. Boden BP, Dean GS, Feagin JA, et al. Mechanisms of anterior31. Nigg BM. Extrinsic factors. In: Ditrix A, Knuttgen HG, Tittel K,

cruciate ligament injury. Orthopedics 2000; 23: 573-8editors. The Olympic book of sports medicine, Oxford:

15. Olsen OE, Mykelbust G, Engebretsen L, et al. Injury mecha-Blackwell Scientific, 1988

nisms for anterior cruciate ligament injuries in team handball:32. Watkins J. Structure and function of the musculoskeletal sys-a systematic video analysis. Am J Sports Med 2004; 32 (4):

tem. Champaign (IL): Human Kinetics, 19991002-1233. Hungerford DS, Barry M. Biomechanics of the patellofemoral16. Hume PA, Steele JR. Injury prevention strategies in netball: are

joint. Clin Orthop 1979; 144: 9-15Australian players heeding the advice? In: Abstracts of theAustralian Conference of Science and Medicine in Sport; 1997 34. Horton MG, Hall TL. Quadriceps femoris angle: normal valuesOct 7-10; Canberra. Bruce (ACT): Sports Medicine Australia, and relationships with gender and selected skeletal measures.1997: 306-7 Phys Ther 1989; 69: 897-901

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 17: A Risk-Factor Model for Anterior Cruciate Ligament Injury

A Risk-Factor Model for Anterior Cruciate Ligament Injury 427

35. Hsu RWW, Himeno S, Coventry MB, et al. Normal axial 53. Liu SH, Al-Shaikh R, Panossian V, et al. Primary immunolocal-alignment of the lower extremity and load bearing distribution ization of estrogen and progesterone target cells in the humanat the knee. Clin Orthop 1990; 255: 215-27 anterior cruciate ligament. J Orthop Res 1996; 14: 526-33

36. Woodland LH, Francis RS. Parameters and comparisons of the 54. Liu SH, Al-Shaikh R, Panossian V, et al. Estrogen effects thequadriceps angle of college-aged men and women in the cellular metabolism of the anterior cruciate ligament: a poten-supine and standing positions. Am J Sports Med 1992; 20: tial explanation for female athletic injury. Am J Sports Med208-11 1997; 25: 704-9

37. Guerra JP, Arnold MJ, Gajdosik RL. Q angle: effects of isomet- 55. Slauterbeck J, Clevenger C, Lundberg W, et al. Estrogen levelsric quadriceps contraction and body position. J Orthop Sports alters the failure load of the rabbit anterior cruciate ligament. JPhys Ther 1994; 19 (4): 200-4 Orthop Res 1999; 17: 405-8

38. Herrington L, Nester C. Q-angle undervalued? The relationship 56. Yu B, Liu SH, Hatch JD, et al. Effect of estrogen on cellularbetween Q-angle and medio-lateral position of the patella. Clin metabolism of the anterior cruciate ligament. Clin OrthopBiomech 2004; 19 (10): 1070-7 1999; 336: 229-38

39. Shambaugh JP, Klein A, Herbert JH. Structural measures as 57. Wojtys EM, Huston LJ, Lidenfeld TN, et al. Association be-predictors of injury in basketball players. Med Sci Sports tween the menstrual cycle and anterior cruciate ligament inju-Exerc 1991; 23: 522-7 ries in female athletes. Am J Sports Med 1998; 26 (5): 614-9

40. Williams PL, Bannister LH, Berry MM, et al., editors. Gray’s 58. Wojtys EM, Huston LJ, Boynton MD, et al. The effect ofanatomy. Edinburgh: Longman, 1995 menstrual cycle on anterior cruciate ligament injuries in wo-

men as determined by hormone levels. Am J Sports Med 2002;41. Shelbourne KD, Davis TJ, Klootwyk T. The relationship be-30 (2): 182-8tween intercondylar notch width of the femur and the inci-

dence of anterior cruciate ligament tears. Am J Sports Med 59. Slauterbeck JR, Fuzie SF, Smith MP, et al. The menstrual cycle,1986; 26: 402-8 sex hormones, and anterior cruciate ligament injury. J Athl

Train 2002; 37 (3): 275-842. Charlton WPH, St John TA, Ciccotti MG, et al. Differences infemoral notch anatomy between men and women: a magnetic 60. Nunley RM, Wright D, Renner JB, et al. Gender comparison ofresonance imaging study. Am J Sports Med 2002; 30 (3): patella tendon tibial shaft angle with weight bearing. Res329-33 Sports Med 2003; 11 (3): 173-85

43. Uhorchak JM, Scoville CR, Williams GN, et al. Risk factors 61. Malinzak RA, Colby SM, Kirkendall DT, et al. A comparison ofassociated with noncontact injury of the anterior cruciate liga- knee joint motion patterns between men and women in select-ment: a prospective four-year evaluation of 869 West Point ed athletic tasks. Clin Biomech 2001; 16: 438-45cadets. Am J Sports Med 2003; 31 (6): 831-42 62. Chappell DJ, Yu B, Kirkendall DT, et al. A comparison of knee

44. Souryal TO, Moore HA, Evans P. Bilaterality in anterior cruci- kinetics between male and female recreational athletes in stop-ate ligament injuries: associated intercondylar notch stenosis. jump tasks. Am J Sports Med 2002; 30 (2): 261-7Am J Sports Med 1988; 16: 449-54 63. Salci Y, Kentel BB, Heycan C, et al. Comparison of landing

45. Lund-Hanssen H, Gannon J, Engbretsen L, et al. Intercondylar maneuvers between male and female college volleyball play-notch width and the risk of anterior cruciate ligament rupture: a ers. Clin Biomech 2004; 19 (6): 622-8case control study in 46 female handball players. Acta Orthop 64. James CR, Sizer PS, Starch DW, et al. Gender differencesScand 1994; 65: 529-32 among sagittal plane knee kinematics and ground reaction

46. Ireland ML, Balantyne BT, Little K, et al. A radiographic force characteristics during a rapid sprint and cut maneuver.analysis of the relationship between the size and shape of the Res Q Exerc Sport 2004; 8: 31-9intercondylar notch and anterior cruciate ligament injury. Knee 65. Decker MJ, Torry MR, Wyland DJ, et al. Gender differences inSurg Sports Traumatol Arthrosc 2001; 9 (4): 200-5 lower extremity kinematics, kinetics and energy absorption

47. Schichendantz MS, Weiker GG. The predictive value of radi- during landing. Clin Biomech 2003; 18: 662-9ographs in the evaluation of unilateral and bilateral anterior 66. Ford KR, Myer GD, Hewett TE. Valgus knee motion duringcruciate ligament injuries. Am J Sports Med 1993; 21: 110-3 landing in high school female and male basketball players.

48. LaPrade RF, Burnett QM. Femoral intercondylar notch stenosis Med Sci Sports Exerc 2003; 35 (10): 1745-50and correlation to anterior cruciate ligament injuries: a pro- 67. Zeller BL, McCrory JL, Kibler WB, et al. Differences in kine-spective study. Am J Sports Med 1994; 22: 198-203 matics and electromyographic activity between men and wo-

49. Teitz CC, Lind BC, Sacks BM. Symmetry of femoral notch men during the single-legged squat. Am J Sports Med 2003; 31width index. Am J Sports Med 1997; 25: 687-90 (3): 449-56

50. Hutchinson MR, Ireland ML. Knee injuries in female athletes. 68. Beard DJ, Dodd CA, Trundle HR. Proprioception enhancementSports Med 1995; 19: 288-302 for anterior cruciate ligament deficiency. J Bone Joint Surg Br

1994; 75: 654-951. Huston LJ, Wojtys EM. Neuromuscular performance character-istics in elite female athletes. Am J Sports Med 1996; 24: 69. Cowling EJ, Steele JR. Is lower limb muscle synchrony during427-36 landing affected by gender? Implications for variations in ACL

injury rates. J Electromyogr Kinesiol 2001; 11: 263-852. Muneta T, Takakuda K, Yamamoto H. Intercondylar notchwidth and its relation to the configuration and cross-sectional 70. Wojtys EM, Huston LJ, Taylor PD, et al. Neuromuscular adap-area of the anterior cruciate ligament. Am J Sports Med 1985; tations to isokinetic, isotonic and agility training programs.25: 69-72 Am J Sports Med 1996; 24 (2): 187-92

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)

Page 18: A Risk-Factor Model for Anterior Cruciate Ligament Injury

428 Hughes & Watkins

71. Granata KP, Wilson SE, Padua DA. Gender differences in active 76. Chappell JD, Herman DC, Knight BS, et al. Effect of fatigue onmusculoskeletal stiffness, part I: quantification in controlled knee kinetics and kinematics in stop-jump tasks. Am J Sportsmeasurements of knee joint dynamics. J Electromyogr Kinesi- Med 2005; 33 (7): 1022-9ol 2002; 12 (2): 119-26

77. Fagenbaum R, Darling WG. Jump landing strategies in male and72. Granata KP, Padua DA, Wilson SE. Gender differences in activefemale college athletes and implications of such strategies formuscle stiffness, part II: quantification of leg stiffness duringanterior cruciate ligament injury. Am J Sports Med 2003; 31functional hopping tasks. J Electromyogr Kinesiol 2002; 12

(2): 127-35 (2): 233-4073. Wojtys EM, Huston LJ, Shock HJ, et al. Gender differences in

78. Wojtys EM, Wylie BB, Huston LJ. The effects of fatigue onmuscular protection of the knee in torsion in size-matched

neuromuscular function and anterior tibial translation inathletes. J Bone Joint Surg Am 2003; 85 (5): 782-9healthy knees. Am J Sports Med 1996; 24 (5): 615-2174. Kanehisa H, Okuyama H, Ikegawa S, et al. Sex difference in

force generation capacity during repeated maximal knee exten-sions. Eur J Appl Physiol 1996; 73: 557-62

Correspondence and offprints: Gerwyn Hughes, Depart-75. Hakkinen K, Kraemer WJ, Newton RU. Muscle activation andforce production during bilateral and unilateral concentric and ment of Sports Science, University of Wales Swansea, Sin-isometric contractions of the knee extensors in men and wo-

gleton Park, Swansea, SA2 8PP, UK.men at different ages. Electromyogr Clin Neurophysiol 1997;E-mail: [email protected]: 131-42

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (5)