altering the length-tension relationship with eccentric exercise

Sports Med 2007; 37 (9): 807-826REVIEW ARTICLE 0112-1642/07/0009-0807/$44.95/0

© 2007 Adis Data Information BV. All rights reserved.

Altering the Length-TensionRelationship with Eccentric ExerciseImplications for Performance and Injury

Matt Brughelli and John Cronin

School of Exercise, Biomedical and Health Sciences, Edith Cowan University, Joondalup, WesternAustralia, Australia

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8071. Eccentric Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8082. The Length-Tension Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808

2.1 Single Fibre (Sarcomere) Force-Length Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8092.2 Whole Muscle Force-Length Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8102.3 Single Joint Torque-Angle Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810

3. Studies Reporting a Shift in Optimum Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8113.1 Dynamometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8123.2 Curve Fitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8133.3 Equipment and Testing Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8133.4 Exercise Methodology and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814

4. Mechanisms for the Shift in Optimum Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8164.1 First and Second Shift in Optimum Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8164.2 Theory of Sarcomereogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8174.3 Theory of Passive Tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818

5. Implications for Athletic Performance and Injury Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8195.1 Muscle Injury Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8195.2 Eccentric Exercise and Athletic Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820

6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823

The effects of eccentric exercise on muscle injury prevention and athleticAbstractperformance are emerging areas of interest to researchers. Of particular interestare the adaptations that occur after a single bout, or multiple bouts of eccentricexercise. It has been established that after certain types of eccentric exercise, theoptimum length of tension development in muscle can be shifted to longer musclelengths. Altering the length-tension relationship can have a profound influence onhuman movements. It is thought that the length-tension relationship is influencedby the structural makeup of muscle. However, the mechanism responsible for theshift in optimum length is not readily agreed upon. Despite the conflict, several

808 Brughelli & Cronin

studies have reported a shift in optimum length after eccentric exercise. Unfortu-nately, very few of these studies have been randomised, controlled trainingstudies, and the duration of the shift has not yet been established. Nonetheless, thisadaptation may result in greater structural stability at longer muscle lengths andconsequently may have interesting implications for injury prevention and athleticperformance. Both contentions remain relatively unexplored and provide thefocus of this review.

1. Eccentric Exercise

Over the last 10 years, interest in eccentric exer-cise and its adaptations has grown substantially.There are a number of reviews dedicated to theeffects of eccentric exercise on muscle damage,[1-4]

delayed onset of muscle soreness (DOMS),[5,6] therepeated bout effect,[6-8] injury prevention,[9,10] ath-letic performance[11,12] and rehabilitation.[13] The in-terest can be attributed to the unique physiologicaland mechanical properties of eccentric muscle con-

0

20

40

60

80

100

120

140

160

0˚ 10˚ 20˚ 30˚ 40˚ 50˚ 60˚ 70˚ 80˚ 90˚Angle

Tor

que

(Nm

)

PrePost

Fig. 1. Schematic representation of a shift in the optimum length-tension relationship. Pre = pre-exercise; Post = post-exercise.

tractions. For example, skeletal muscles are capableof producing the greatest magnitudes of force,[14] injury and performance, the discussion of whichwith little metabolic effort[15,16] when contracting provides the focus of this article. First, the length-eccentrically. The high force levels and/or longer tension relationship at a single fibre, whole musclemuscle lengths[17] are thought to be responsible for and single joint levels are briefly discussed. Second-the damage caused to the contractive,[3,6] connec- ly, the literature that has reported a shift in optimumtive[18,19] and cytoskeletal elements of muscle.[20-22] length after eccentric exercise is critiqued. Thirdly,Muscle damage is also associated with decreased the proposed mechanisms responsible for the shiftforce capabilities,[23] inflammation[24] and impair- are described. Finally, the implications of this right-ment of the excitation-contraction coupling pro- ward shift with regards to athletic performance andcess.[25-27] All forms of exercise can cause muscle injury prevention is elaborated upon.damage; however, only eccentric exercise has beenfound to induce severe stiffness and soreness in the 2. The Length-Tension Relationshipdays following the eccentric exercise bout.

An interesting adaptation occurs after certain The length-tension relationship plays a very im-types of eccentric exercise that affects the mechani- portant role in the function of skeletal muscle. Thecal properties of muscle. After a single bout of magnitude of force a muscle can generate dependseccentric exercise, the length-tension relationship of on its length, velocity and stimulation.[28,29] Whenmuscle can be altered. The optimum length of peak creating a length-tension curve, muscle velocity andtension will occur at longer lengths, thus shifting the stimulation are held constant. Force levels are thenlength-tension curve to the right (see figure 1). This plotted against each muscle length, thus creating thephenomenon may have interesting implications for length-tension curve. The length-tension curve

© 2007 Adis Data Information BV. All rights reserved. Sports Med 2007; 37 (9)

Length-Tension Relationship with Eccentric Exercise 809

gives valuable information on the lengths that canproduce the greatest or least amount of tension.From this information, the regions of the ascendinglimb, descending limb and optimum length can bedetermined (see figure 2a). Optimum length is oftenused to describe sarcomere length (see section 2.1),muscle length (see section 2.2), or joint angle (seesection 2.3), and tension is often used to describeforce (see sections 2.1 and 2.2), or torque (seesection 2.3).

Note that the length-tension curves for the singlefibre (sarcomere), whole muscle and single joint allhave different shapes. At the single fibre level (seefigure 2a), only the cross-bridge interaction betweenmyosin and actin contributes to the shape of thecurve and the development of active force within thesarcomere. At a whole muscle (see figure 2b) andsingle joint level (see figure 2c), the curve has moreof a smooth shape because of the variations inmuscle, tendon and joint designs, and the contribu-tion of all these factors to the length-tension curve.

There are many inconsistencies in the literatureregarding the length-tension relationship. Most ofthe confusion revolves around the reporting of forceand torque interchangeably. A brief treatise of force-length curves (see sections 2.1 and 2.2) and torque-angle curves (see section 2.3), and how they aremeasured should clarify some of the conjecture inthe area.

2.1 Single Fibre (Sarcomere)Force-Length Curve

The isometric force-length curve is generated bymaximally stimulating a single muscle fibre(sarcomere) over a range of lengths and measuringforce outputs. In this situation, velocity is held con-stant at zero and stimulation is held constant atmaximum levels. Maximum force levels are plotted

100

a

50

0

1.27

For

ce (

%)

1.7 2.0 2.2 3.6

Passiveforce

Descendinglimb

Ascendinglimb

Plateauregion

1.7

2.0 2.2

Sarcomere length (μm)

60

80b

40

20

0

0

For

ce (

N)

0.5 1.0 1.5 2.52.0

Passivetension

Combinedtension

Activetension

Length (cm)

100

150c

50

0

Tor

que

(Nm

)

0˚ 20˚ 40˚ 80˚60˚

Optimum length

Active andpassive tension

Angle

Fig. 2. Schematic representations of (a) single fibre force-lengthcurves; (b) whole muscle force-length curve; and (c) torque-anglecurve. Active and passive tension curves as well as the ascending,plateau and descending regions are indicated on the single fibreand whole muscle curves.

against each starting length to generate the force-length curve. The active force levels produced and magnitude of overlap between active and myosinthe shape of the curve are a direct result of the filaments (see figure 2a). Larger magnitudes of



force can be produced with greater numbers of 2.2 Whole Muscle Force-Length Curve

cross-bridge attachments. Altering the length of theThe force-length curve of whole muscle is gener-fibre affects the number of attached actin and myo-

ated similar to that of the single fibre. The wholesin cross bridges[22,29] and subsequent force develop-muscle is maximally stimulated over a range of

ment.discrete lengths. Muscle velocity is again held con-

Gordon et al.[30] was the first to use isolated, stant at zero, i.e. isometric contraction. Maximumintact single skeletal muscle fibres to generate the isometric force is plotted against muscle length, thusforce-length curve. As detailed in figure 2a, the creating the whole muscle force-length curve. Maxi-maximum percentage of tension is plotted against mum force at each muscle length is due to bothsarcomere length. At short sarcomere lengths active and passive forces. At shorter lengths, all the

force is due to active components (cross-bridges),(<1.7μm), the double overlap between the actin andand at longer lengths most of the force is due to themyosin filaments results in depressed force levels.passive components.[31]

As the length of the sarcomere increases to the steepThe whole muscle force-length curve has a dif-portion (1.7–2.0μm), the actin filaments from one

ferent shape to the single fibre force-length curve.side of the sarcomere no longer interfere with theThe magnitude of active force generated is not duecross-bridge formations on the other side of thesolely to the interaction of cross-bridges as is the

sarcomere, and force increases. The region of thecase for the single fibre force-length curve. The

curve where active force is first measured (1.27μm),different shape is due to the contractile properties

until maximum force levels (2.0μm) is called theand design of whole muscles and tendons.[29] Whole

ascending limb of the force-length curve. The pla- muscles are made up of a variety of fibre types andteau region (2.0–2.2μm) of the curve is where maxi- sarcomere lengths. This variety means that the dif-mum force levels are attained and held constant. At ferent fibres will have different optimum lengths ofthis range, the greatest amount of actin and myosin force development, which smoothes the curve andoverlap occurs. Although there is greater overlap broadens the plateau region.[29,31] Architecture alsofrom 2.2 to 2.0μm, no additional cross-bridges are plays a role as the number of sarcomeres in series,

parallel, or angle of pinnation contribute to forceformed. This is due to a bare region (0.2μm indevelopment and help to shape the force-lengthlength) at the centre of the myosin molecule that iscurve. Tendon fibres in series with muscle fibresdevoid of cross-bridges.[29] The plateau region isalso contribute to the shape of the curve. Tendonalso known as the ‘optimum length’ of the force-compliance can affect force development by al-length curve. The descending limb region of thelowing simultaneous muscle shortening and tendonforce-length curve occurs from the sarcomerelengthening. All of these factors lead to a smoother

lengths of 2.2–3.65μm. At sarcomere lengths offorce-length curve with an extended peak re-

>3.65μm, active force levels drops to zero as there isgion.[29,32]

no overlap between actin and myosin. Notice in

figure 2a that there is also a passive tension curve.2.3 Single Joint Torque-Angle Curves

As the muscle is stretched to greater lengths, passive

tension increases dramatically. The elements re- At the single fibre level, sarcomere force is plot-sponsible for passive tension lie outside of the cross- ted against sarcomere length, whereas at the wholebridges and do no need active stimulation. muscle level, muscle force is plotted against muscle



length. Conversely, at the single joint level, jointtorque is plotted against joint angle. Torque is deter-mined by the magnitude of muscle force and mo-ment arm relative to the joint.[29,31] Both muscleforce and the length of the moment arm changethroughout a joint movement and affect torque de-velopment. They can both be graphed against jointangle. Joint torque is the product of these two mea-surements. Peak torque and peak force can actuallyoccur at different joint angles.[29] Muscle force canonly be measured at this level if the moment armwas determined at each discrete joint angle.

The shape of the single joint torque-angle curveis unique to that of the single fibre or whole muscleforce-length curves. Joint movements typically in-volve multiple muscle groups crossing the joint andproducing torque, thus affecting the shape of thecurve. Other factors that affect the shape of thecurve include the constraints of the joint itself, andrange of motion. These differences lead to a differ-ent shaped curve. The torque-angle curve is smooth-er and flatter than the muscle-force curves. Noticealso that the peak of the curve is extended andoccurs at longer lengths (see figure 2c).

3. Studies Reporting a Shift inOptimum Length

Several studies have reported a shift in optimumlength after eccentric exercise (table I). This phe-nomenon has been observed in animal and humanstudies. Animal studies have reported shifts in singletoad fibres,[33,34] single frog fibres,[14,35] rat fibres[36]

and individual motor units in cats.[37] Acute shiftshave also been reported in human plantar flex-ors,[38-40] forearms,[41] quadriceps,[42,43] tricepsbrachii,[44] biceps brachii[45,46] and hamstrings.[47,48]

Muscle architecture, in terms of pinnation angle andfibre type, does not appear to have an affect on theshift in optimum length after eccentric exercise. Thegreatest shifts have occurred in the human quadri-ceps[42] and biceps brachii.[45]


Tab

le I

. In

fluen

ce o

f ec

cent

ric e

xerc

ise

inte

nsity

, vo

lum

e an

d m

uscl

e le

ngth

on

the

mag

nitu

de o

f th

e sh

ift in

opt

imum

leng

th

Stu

dyE

xerc

ise

inte

nsity

Exe

rcis

e vo

lum

eM

uscl

e le

ngth

Shi

ft m

agni

tude

(°)

Jone

s et

al.[3

8]LI

, w

alki

ng b

ackw

ards

on

trea

dmill

,M

L, 2

h du

ratio

nM

L, t

oe-t

o-he

el a

ctio

n3.

9 (I

MM

), t

ricep

s su

rae

1.3

km/h

Whi

tehe

ad e

t al

.[39]

LI,

wal

king

bac

kwar

ds o

n tr

eadm

ill,

ML,

1h

dura

tion

ML,

toe

-to-

heel

act

ion

4.4

(IM

M),

tric

eps

sura

e3.

5 km

/h

Whi

tehe

ad e

t al

.[40]

MI,

wal

king

bac

kwar

ds o

n tr

eadm

ill,

ML,

1h

dura

tion

ML,

toe

-to-

heel

act

ion

6 (I

MM

), 7

.6 (

2h p

ost)

,>

3.5k

m/h

with

5–1

0kg

wei

ght

belt

tric

eps

sura

e

Cla

rk e

t al

.[48]

MI,

body

wei

ght

exer

cise

to

failu

re,

LV,

2–3

sets

of

5–8

reps

ML,

90

to <

30°

of k

nee

exte

nsio

n6.

5 (4

wk

post

), h

amst

rings

Nor

dic

ham

strin

g w

ith p

artn

er

Bro

cket

t et

al.[4

7]M

I, bo

dyw

eigh

t ex

erci

se t

o fa

ilure

,H

V,

12 s

ets

of 6

rep

s =

76

tota

lM

L, 9

0 to

<30

° of

kne

e ex

tens

ion

7.7

(IM

M),

8.5

(4d

pos

t),

Nor

dic

ham

strin

gre

psha

mst

rings

Pet

titt

et a

l.[44]

HI,

max

ecc

entr

ic e

lbow

con

trac

tions

,M

V,

3 se

ts o

f 25

rep

s =

75

tota

lM

L, 8

0–13

0° o

f el

bow

ext

ensi

on10

.4 (

IMM

), t

ricep

s br

achi

iis

okin

etic

dyn

amom

eter

reps

Bow

ers

et a

l.[42]

MI,

step

-dow

n ex

erci

seH

V,

12 s

ets

of 2

0 re

ps =

240

tot

alLL

, pl

atfo

rm h

eigh

t se

t at

the

pat

ella

15.4

(IM

M),

10.

5 (4

d po

st),

reps

quad

ricep

s

Pra

sart

wut

h et

MI,

ecce

ntric

elb

ow e

xten

sion

HV

, 40

–160

tot

al r

eps

LL,

60°

to f

ull e

lbow

ext

ensi

on16

.7 (

2h p

ost)

, 14

(1d

pos

t),

al.[4

6]bi

ceps

bra

chii

Phi

lippo

u et

al.[4

5]H

I, m

axim

um e

ccen

tric

elb

owM

V,

2 se

ts o

f 25

rep

s =

50

tota

lLL

, 50

–170

° of

elb

ow e

xten

sion

16 (

1d p

ost)

, 18

(2d

pos

t),

exte

nsio

nre

psbi

ceps

bra

chii

HI =

hig

h in

tens

ity; H

V =

hig

h vo

lum

e; IM

M =

imm

edia

tely

pos

t-ex

erci

se; L

I = lo

w in

tens

ity; L

L =

long

mus

cle

leng

th; L

V =

low

vol

ume;

max

= m

axim

um; M

I = m

oder

ate

inte

nsity

;M

L =

mod

erat

e m

uscl

e le

ngth

; M

V =

mod

erat

e vo

lum

e; p

ost

= p

ost-

exer

cise

; re

ps

= r

epet

ition

s.


A pilot study by Clark et al.[48] reported a sus- odex® III (Biodex Medical, Inc., Shirley, NY,tained shift in optimum length after 4 weeks of USA)[42,47,48] or Kin-Com (Chattecx Corp., Inc.,eccentric exercise. This is the only training study, to Hixson, TN, USA),[44,45] were the most frequentlyour knowledge, that has reported a sustained shift in used to measure joint torques and angles. Theseoptimum length. Brockett et al.[47] and Bowers et machines have an attached lever arm that is con-al.[42] both observed a sustained shift after two sepa- trolled by a motor. The angular velocity of the leverrate eccentric exercise sessions, which were separat- arm is controlled by the machine. The subjects caned by 8 days. The shifts lasted for 18[47] and 24[42]

either push or pull against the lever arm to measuredays, respectively. Although many authors have ar- torque capabilities. Isokinetic dynamometers alsogued that eccentric exercise can lead to a sustained typically have an isometric function that allows forshift in optimum length,[9,47,49,50] there are currently isometric torque measurements. Regardless ofno randomised/controlled training studies that have whether the muscle contraction is isometric orreported a sustained shift in optimum length with isokinetic, torque levels can be measured over aeccentric exercise. The reader should be aware of range of discrete joint angles. Peak torque and jointthis major limitation in the literature.

angle can then be plotted against each other toThe following sections will review the human generate the torque-angle curve.

studies that have directly measured a shift in opti-There are two methods for measuring the rangemum length. Both men and women were used as

of peak torque levels. The first (and more timesubjects in these nine studies.[38-40,42,44-48] Overall, 60consuming) method involves measuring maximummen and 40 women participated. All subjects wereisometric contractions over a range of discrete jointbetween 18 and 37 years of age. With the exceptionangles. The second involves measuring peak torqueof Clark et al.,[48] who used amateur Australianduring a continuous concentric contraction, which isRules football players, all of the subjects were clas-an easier and faster way of generating torque-anglesified as being recreationally active. The reader

should be aware of these limitations and delimita- curves (a maximum isokinetic muscle contractiontions when interpreting the results and developing can be completed in only a few seconds). Thisconclusions regarding shift in optimum length. The second method was first utilised only recently.[47]

following sections of this discussion will focus on More studies have since utilised isokinetic dyna-dynamometry, subjects, muscle groups, eccentric mometers to generate in vivo torque-angleexercise protocols, measurement protocols, magni- curves.[42,44,48] In all these studies, peak torque wastude of the shift and duration of the shift. Such an measured for each contraction and plotted againstapproach will expose some of the delimitations and ankle angle; the curves were then used to determinelimitations associated with the literature in this area. the optimum angle of peak torque.

The shift in optimum length can be measured3.1 Dynamometry directly or indirectly with this equipment. Three

studies reported a decrease in isometric peak torqueat short lengths, and an increase in isometric peakAll the human studies that have directly mea-torque at longer lengths after eccentric exer-sured a shift in optimum length after eccentric exer-cise.[33,43,51] This would be considered an indirectcise measured tension at the single joint level, thusmeasure of a shift in the optimum length since thejoint torque was measured and plotted against jointoptimum length was not directly measured.angle. Isokinetic dynamometers, such as the Bi-



3.2 Curve Fitting 3.3 Equipment and Testing Protocol

Direct measures of optimum length shifts can be Brockett et al.,[47] Clark et al.,[48] and Bowers etdetermined by utilising curve fitting procedures. By al.[42] utilised a Biodex® III isokinetic dynanometerfitting a curve to a range of peak torque values at to measure knee joint angle and peak torque values

during maximum concentric contraction. They alldiscrete joint angles, a torque-angle curve can beused a constant velocity of 60° per second duringgenerated. Curve fitting is a convenient way ofknee extension and knee flexion. The ‘zero angle’identifying the angle of peak torque. The most fre-was set at 90° of leg flexion for Brockett et al.,[47]

quently used curve fitting procedures are the Gaus-thus an increase in degrees meant an increase in

sian and quadratic (fourth order) polynomial. Inlength (joint angle). Conversely, Clark et al.[48] and

order to use a Gaussian fitting curve, the data mustBowers et al.[42] set their zero angles at full leg

be normally distributed. However, the shape of the extension. Either five[48] or seven[47] maximum con-single joint torque-angle curve is not symmetrical, centric contractions were averaged and used as datathus the data are not normally distributed. Only in these studies.datapoints >75–90% peak torque can be used to Pettitt et al.[44] and Philippou et al.[45] used a Kin-generate the curve. Datapoints above these values Com isokinetic dynamometer to measure torque andhave more of a normal distribution. Thus, only the corresponding joint angles. Philippou et al.[45] had

their subjects perform maximum isometric contrac-top of the torque-angle curve is used to identify thetions of the elbow flexors (biceps brachii). Theangle of peak torque. The Gaussian fitting curve iscontractions were measured over a range of fivethe most commonly used length-tension fittingjoint angles (50, 70, 90, 140 and 160°) in randomcurve in the literature for both human and animalorder. The subjects performed two maximum volun-

studies. Jones et al.[38] and Whitehead et al.[39,40]

tary isometric contractions of 3-second durations atfitted their datapoints above 75% peak torque. Sev-

each joint angle. The better of the two trials waseral isometric contractions were measured over an used for the data collection. The subjects in theankle range of 50–90°. Others have used Gaussian study of Pettitt et al.[44] performed three maximumfitting curves for datapoints >90% peak torque dur- concentric contractions of elbow extension (tricepsing isokinetic concentric contractions.[42,47] brachii). The repetition with the highest torque value

was used to determine optimum length.The greatest concern with the Gaussian curvePrasartwuth et al.[46] had their subjects performfitting is that it does not describe changes in the

maximum isometric contractions on an isometricentire torque-angle curve, since only the peak ismyograph, which measured peak torque and jointconsidered. For this reason, some researchers useangle. The average of two maximum isometric con-quadratic polynomial curve fitting for their datatractions (biceps brachii) was measured at each dis-

collection.[45,46] Like the Gaussian curve, the anglecrete joint angle. Joint angle measurements started

of peak torque is easily identified. However, with at 60° and progressed in 10° increments to 150°. Inthe polynomial fitting curve, changes can be de- addition, during each contraction, paired-pulse stim-scribed across the entire range of motion. Another ulation was used to evoke a superimposed twitch. Itadvantage of this procedure is that the data do not was reported that when voluntary force is reduced,need to be normally distributed. paired-pulse stimulation is an appropriate stimulus



for estimations of voluntary activation. Peak torque jects performed the toe-to-heel action in one legwas plotted against each discrete joint angle. (experimental), while the other leg (control) landed

with the toe and heel simultaneously. When the toeWhitehead et al.[39,40] and Jones et al.[38] used acustom-made adjustable chair with a steel frame to and heel land together, eccentric contraction is limit-measure torque and joint angle. The steel frame ed. All three studies inclined their treadmills to 13°.supported a footplate that was attached to a rotating Jones et al.[38] used a very low intensity protocol (1.3axis. Torque was measured with four strain gauges km/h stepping rate), which was carried out for a longattached to an axle supporting the footplate. Torque duration (2 hours). Whitehead et al.[39] had theirwas measured isometrically from 90 to 50° in 5°[38] subjects step at a higher rate (3.5 km/h), whichor 10°[39,40] increments. Ankle angle was defined as increased the intensity of the eccentric contractions.the angle between the footplate and the shins. All Whitehead et al.[40] had their subjects step at athree studies determined the torque-angle curve by slightly higher stepping rate (>3.5 km/h). However,using double pulse stimulation of the tibial nerve, the main difference between the studies was thatusing 1ms pulses 20ms apart. It was reported that a Whitehead et al.[40] had their subjects carry an addi-torque-angle curve constructed this way was easier tional 5–10kg weight belt. The additional load andto control and similar in shape and optimum length step rate increased the intensity of their eccentricto those taken from maximum voluntary contrac- exercise protocol, which may explain why they hadtions.[35]

a greater magnitude of shift in optimum length.

Brockett et al.[47] and Clark et al.[48] used the3.4 Exercise Methodology and Results

same exercise (Nordic hamstring) and reported sim-ilar shifts (7.7 and 6.3°). The Nordic hamstringUp until this point there has been no discussionexercise involves a subject kneeling on the groundon the magnitude of the shift in optimum length after(or a board), and while maintaining a constant andeccentric exercise. Table I lists the nine studies inopen hip angle, the subject slowly lowers their bodyorder from the smallest to largest shifts in optimumtowards a prone position. In this exercise, the ham-length. There are three main factors that influencestrings are isolated and control the lowering bodythe magnitude of the shift:while contracting eccentrically. Brockett et al.[47]1. eccentric exercise intensity;[40,44,45]

used a custom-made 2m long wooden board with2. eccentric exercise volume;[42,46,47]

upholstered areas for the chest and knees. The an-3. length of the muscle during eccentric exer-kles of the subjects were stabilised to the board withcise.[42,45,46]

ankle straps. Clark et al.[48] used the same exercise,Jones et al.[38] and Whitehead et al.[39,40] usedexcept the subjects were placed on the ground andsimilar eccentric exercise protocols and thus report-had a partner. The subjects placed a towel belowed similar shift magnitudes (3.9, 4.4 and 6°). Theirtheir knees and had a partner apply pressure to theirsubjects walked backwards on an inclined movingheels to ensure that their feet stayed in contact withtreadmill, which was meant to eccentrically strainthe ground throughout the movement. The Nordicthe triceps surae. However, the exercise protocolshamstring exercise can be classified as a submax-were not exactly the same, and thus may explain theimal exercise intensity, but Brockett et al.[47] used adifferences in their shifts. For the experimental legs,very high exercise volume (12 sets of 6 repetitions).the subjects were instructed to step backwards withThis volume was meant to acutely damage the ham-a toe-to-heel action. This technique ensured that thestrings, and then measurements would be taken totriceps surae was contracted eccentrically. The sub-



determine how long the shift in optimum length subjects in the study by Philippou et al.[45] per-formed maximum eccentric contractions of the el-would last. The subjects in the study by Clark etbow flexors to 170° elbow extension. Pettitt et al.,[44]al.[48] performed 2–3 sets of 5–8 repetitions. Thison the other hand, had their subjects perform maxi-protocol was repeated 1–3 times per week for 4mum eccentric contractions of the triceps to 130°weeks. This protocol started with a low volume thatelbow flexion. It should be noted that the range ofincreased over a 4-week period. The objective wasmotion of elbow flexion is much less than elbownot to cause acute damage, but to slowly increase theextension. The design of the elbow joint will notintensity and volume of eccentric exercise. Fromallow the triceps brachii to be fully stretched. Thus,these two studies, it appears that a high-volumethe differences in shifts may be due to the muscleeccentric protocol can cause an acute shift in opti-length operating range during eccentric contraction.mum length. If volume is increased overtime, acuteFrom these two studies, it appears that very littlemuscle damage can be avoided, while a shift indamage occurs at short lengths (0–80° elbow exten-optimum length may occur over a 4-week period.sion),[44] where damage is observed at longer lengthsImportant technical notes were to slowly lower the(80–130° elbow extension),[44] and damage is thebody while maintaining contraction of the ham-greatest at the longest lengths (50–170° elbow flex-strings for as long as possible.ion).[45]

Pettitt et al.[44] and Philippou et al.[45] utilised anBowers et al.[42] reported a 15.4° shift after a step-isokinetic dynamometer (Kin-Com) to perform

down exercise that eccentrically contracted themaximum eccentric contractions for their exercisequadriceps. The subjects stepped up to a platformprotocols. The elbow flexors (biceps brachii)[45] and(height set at the patella) with their left leg (control),the elbow extensors (triceps brachii)[44] were used tofollowed by their right leg (experimental). Onceperform maximum eccentric contractions. The non-both feet were placed on top of the platform, thedominant arms were utilised in both protocols. In thesubjects slowly stepped down with their left legstudy of Pettitt et al.,[44] the subjects performed threeleading. The subjects were instructed to control thesets of 25 maximum eccentric repetitions (50–170°)movement with their right leg, which remained onat a velocity of 60° per second. This study investi-the platform during the step down. This allowed thegated the effects of eccentric exercise performed atquadriceps of the right leg to contract eccentrically.short muscle lengths (0–80°) and long muscleThe subjects performed 12 sets of 20 step-downs forlengths (80–130°). The short-length group did nota total of 240 steps. This study used a very highshow a shift in optimum length after eccentric exer-volume of 240 total eccentric contractions. Also, thecise. The long-length group did show a significantexercise was designed to have the eccentric contrac-shift in optimum length (10.4°). The eccentric exer-tions occur at long muscle lengths. The subjectscise protocol used a much higher intensity than thestepped down from a platform while their experi-previous studies. The eccentric contractions weremental leg contracted eccentrically to control theperformed with maximum voluntary effort. Thismovement.could explain why the shift was higher than the

previous studies. However, Philippou et al.[45] used a Prasartwuth et al.[46] also reported a large shift invery similar exercise protocol to Pettitt et al.,[44] but optimum length (16.7°) after eccentrically trainingreported a much greater shift in optimum length the elbow flexors (biceps brachii). A pulley wheel(16.1° vs 10.4°). In fact, the subjects performed 25 was designed for the eccentric exercise protocol.fewer repetitions (two sets of 25 repetitions). The The axis of rotation of the pulley wheel was aligned



with the subjects’ elbow joints. The subjects low- 4. Mechanisms for the Shift inOptimum Lengthered a weight attached to the wheel from 60° to full

elbow extension. Initially, five sets of ten repetitionsSeveral mechanisms have been proposed for thewere performed. The sets were continued until max-

rightward shift after eccentric exercise, which in-imum voluntary torque fell by 40%. The total

clude a partial transformation of active contractiveamount of repetitions varied from 50 to 160. The elements into passive elastic elements;[14] damage toeccentric contractions were performed from 60° to the myotendinous attachments;[41] and damage to thefull elbow extension. The combination of high vol- calcium handling structures.[27,52] It has further beenume and long muscle lengths in this and the previous proposed that two different shifts occur after eccen-study probably contributed to the large shifts in tric exercise. The first shift is acute and appears after

the muscle has been damaged, and the second shiftoptimum length (15.4[42] and 16.7°[46]).is due to an adaptation that occurs over a period ofFrom these nine studies, a few tentative conclu-time (10 days to 8 weeks).[2,8,42] The mechanismssions can be made:proposed for this second shift include an increase in

• high-intensity eccentric exercise results in sarcomeres in series (sarcomereogenesis)[53] and angreater shifts in optimum length; increase in passive tension at longer muscle

lengths.[12]• muscles that are eccentrically contracted at long-

er lengths result in greater shifts in optimum4.1 First and Second Shift in Optimum Lengthlength;

• high-volume eccentric exercise results in greaterStructural muscle damage has been observed im-

shifts in optimum length; mediately after eccentric exercise. Electron micro-• the combinations of high-intensity/long muscle scopic examinations have shown Z-line streaming

length or high-volume/long muscle lengths pro- (sarcomeres out of register with each other), regionsof overstretched half sarcomeres and t-tubule dam-duced the greatest shifts in optimum length;age.[4,54,55] The disrupted sarcomeres are thought to• muscle architecture does not seem to affect thebe responsible for the first acute shift in optimumshift in optimum length;length. The descending limb of the length-tension

• eccentric exercises that involve long muscle curve is thought to be a region of instability, wherelengths and either high intensity or high volume force levels decrease. It has been proposed by Mor-should be developed to produce the greatest gan[53] that during active muscle lengthening whenshifts in optimum length; the myofilaments are stretched onto the descending

limb, some of the weaker sarcomeres will be stretch-• it may be possible to produce a sustained shift ined more than others. These sarcomeres will becomeoptimum length after 4 weeks of eccentric exer-progressively weaker until there is no overlap be-cise;tween myofilaments. When the eccentric contrac-

• muscle damage may not need to be induced for tions are repeated, more sarcomeres will be overex-this adaptation (shift in optimum length) to occur. tended from weakest to strongest. At the end of eachStudy investigators should keep these factors in contraction, as the muscle relaxed, a number ofmind when developing eccentric exercise proto- sarcomeres may not reintegrate, thus becoming dis-cols. rupted. The disrupted sarcomeres will be scattered at



random along muscle fibres, which increases series eccentric contractions. Thus, the muscle will main-compliance. The acute shift in optimum length after tain stability at longer muscle lengths.eccentric exercise is thought to result from an in- It has been argued that the number of sarcomerescrease in series compliance, due to disrupted in series is highly plastic and can be altered withsarcomeres.[2,53]

training.[32,57,58] Herzog et al.[59] found that length-It has also been proposed that a second shift tension properties of the rectus femoris were differ-

occurs after eccentric exercise. This delayed shift is ent between runners and cyclists. With running, theargued to be a result of either a mechanical[13,56] or rectus femoris experiences a stretch-shortening cy-cellular[10,53] adaptation in the muscle. Despite the cle, where high force levels were required at longarguments, there is a growing interest in the addi- lengths during both eccentric and concentric con-tional stability at longer muscle lengths due to the tractions. Cycling requires high forces at shortershift in optimum length. One of the proposed mech- muscle lengths during concentric contractions. Con-anisms for this second shift is an addition of sequently, it would be expected that the runnerssarcomeres in series (sarcomereogenesis) after ec- have more sarcomeres in series than the cyclists.centric exercise.[2,10,53] Many believe that a more The study reported that runners produce peak torquecompliant muscle will avoid the unstable regions at long lengths, and cyclists produce peak torque at(descending limb) of the length-tension curve dur- short muscle lengths.[59]

ing further eccentric exercise.[2,50] The other pro-Direct evidence for sarcomereogenesis has beenposed mechanism for the second shift is an increase

observed in several animal and human studies. Mostin passive tension at longer muscle lengths. Thisof the studies reported that sarcomereogensis oc-adaptation is thought to occur after a period ofcurred after a static stimulus (passive static stretch-eccentric training (6–8 weeks). The belief is thating or shortening). However, there are a few studieseccentric exercise leads to greater passive stiffness,that have shown an increase in serial sarcomereor spring-like qualities.[12,56] LaStayo et al.[13] arguednumber after eccentric exercise. Lynn and Mor-that an increase in stiffness at longer lengths willgan[60] were the first to report direct evidence of anincrease force production before failure. Stability atincrease in the number of sarcomeres in series afterlonger muscle lengths may prevent active muscleeccentric exercise. Rats were trained to run uphillstrain injuries and possibly enhance athletic per-(concentric group) or downhill (eccentric group) onformance.a treadmill for 5 days. Running downhill caused themuscles to contract eccentrically as they controlled4.2 Theory of Sarcomereogenesisthe movement, while running uphill was predomi-nantly a concentric exercise. The number ofSarcomereogenesis is the addition of sarcomeressarcomeres in series (vastus intermedius) increasedin series within a muscle fibre and is thought to havein the downhill group. Since then, four more studiesan important role in maintaining the relationshiphave reported an increase in serial sarcomere num-between sarcomere length and joint angle. The num-ber after eccentric exercise.[61-64] Butterfield et al.[63]ber of sarcomeres in series is thought to be analso reported an increase in sarcomeres in seriesadaptable property of muscle. After sarcomere-(vastus intermedius) after trained rats ran downhillogenesis has occurred, sarcomere length will befor 10 days. More surprisingly, the group of rats thatshorter for a given muscle length.[53] This adaptationran uphill decreased the number of sarcomeres inis thought to keep the myofilaments off the descend-

ing limb of the length-tension curve during future series in the vastus intermedius muscle.



Furthermore, Whitehead et al.[39] found that if rated anatomically in fish, making them a goodeccentric exercise is followed by previous concen- model for study. Sarcomeres that are exposed totric exercise, muscle damage and the shift in opti- greater strains may express titins with larger elasticmum length was greater than if eccentric exercise segments. Differences in titin expression have alsowas performed alone. The authors concluded that been reported in humans. Weight lifters, power lift-concentric-based training programmes produce ers and sprinters expressed a greater percentage ofchanges in the muscles that make them more prone titin-1 isoforms than non-athletes.[69] The subjects into injury caused by eccentric exercise. Athletes such the athletic groups would be expected to experienceas triathletes and marathon runners, who regularly strain more often and to a greater degree.perform both concentric and eccentric muscle con- The titin filament has a major role in the develop-tractions for a prolonged period of time may benefit ment of passive tension. However, the role of thefrom these findings. Programmes should be devel- titin in muscular performance during explosive ac-oped to help these athletes enhance their perform- tivities is unknown. It is possible that the titin isance, while preventing muscle damage induced by capable of storing and releasing elastic energy.[13,70]

eccentric exercise.[39]McBride et al.[69] reasoned that a more elastic mus-cle (due to altered titin isoforms) would lead to

4.3 Theory of Passive Tension greater power production, and proposed that explo-sive training can alter titin expression. In this study,

Another proposed mechanism for the secondsubjects performed explosive squat-jump training

shift in optimum length after a period of eccentricfor 8 weeks with either 30% or 80% of their maxi-exercise (6–8 weeks) is a greater contribution of themum back squat. There was no change in titinpassive elements at longer muscle lengths.[13,56]

isoform expression after the training period. Howev-More passive tension at longer muscle lengths caner, the training stimulus did not induce a greatalso shift the length-tension curve to the right. Ec-amount of muscle strain, which would be needed forcentric exercise can cause disruption to the musclesa mechanical adaptation to occur.passive components. Cytoskeletal proteins, such as

Despite the claims that eccentric training candesmin and titin, play a significant role in the struc-alter titin expression, currently there are no trainingture and function of the sarcomere.[29] After eccen-studies that have examined the effects of eccentrictric exercise, disruption and degradation occurs totraining on titin expression. There are, however,the desmin and titin proteins.[22,29] Titin content hasstudies that have observed increases in passive stiff-been reduced by up to 30% in human subjects (vas-ness at longer muscle lengths after eccentric exer-tus lateralis) after a single bout of eccentric exer-cise. Increases in passive stiffness have been report-cise.[65] A protective adaptation has been suggesteded in animal[56,71] and human training studies[71,72]to occur that strengthens the cytokeletal proteins,after eccentric exercise. Labeit et al.[73] reported thatand prevents them from being damaged in the fu-passive force enhancement was caused by a stifferture. Barash et al.[66] reported an increase in desminmolecular titin. It was also reported that the titincontent 7 days after eccentric exercise-induced mus-becomes more sensitive to calcium at longercle damage in rats. Severe or continuous musclelengths.[73,74] Since the titin filament is responsiblestrain is thought to be a stimulus for adaptations infor the majority of passive tension and stiffness attitin expression.[67,68] Spiers et al.[68] reported thatlonger muscle lengths,[29,73] this could be an indirectsarcomere strain amplitude and titin isoform size

were correlated in fish (carp). Fibre types are sepa- indicator that an adaptation has occurred. Even



though the torque-angle curve is taken when a mus- proposed that hamstring injuries most often occurcle is active and contracting maximally, both pas- when they are being actively lengthened,[82,83] orsive and active elements contribute to the curve. If when they need to switch from an eccentric contrac-eccentric exercise can cause an increase in passive tion to a concentric contraction after being activelytension at longer lengths, the length-tension curve lengthened.[84] The long head of the biceps femorisshould be shifted to the right. This adaptation to the is the most often injured hamstring muscle.[83] Un-titin filament could be the mechanism for the second like the semitendinosus and semimembranous, theshift in optimum length. biceps femoris is actively lengthened throughout the

It should be noted that this increase in passive latter half of the swing phase during sprinting.[83]

tension after a training period with multiple eccen- This makes the biceps femoris more susceptible totric training sessions, should not be confused with muscle strain injuries.the well documented increase in passive tension that Many believe that athletes who produce peakoccurs immediately and up to 10 days after a single torque at shorter muscle lengths are more likely tobout of eccentric exercise. The increase in passive get injured.[9,10,47,50] A shorter optimum lengthtension immediately after eccentric exercise is due would mean that more of the muscles operatingto disruption of the excitation-coupling process. The range would be on the descending limb of theincrease in passive tension that occurs after an ec- length-tension curve. Brockett et al.[50] explored thiscentric-based training programme of 6–8 weeks is idea by measuring optimum lengths in athletes whothought to be caused by an increase in the contribu- have had previous hamstring injuries in one legtion of the passive elements. (experimental) versus their uninjured leg (control).

The mean optimum angle was 12.7° shorter for the5. Implications for Athletic Performanceinjured leg, although strength values (hamstring/and Injury Preventionquadriceps ratio) were similar. Based on these find-ings, the authors suggested that optimum length may

5.1 Muscle Injury Prevention be a greater risk factor for muscle strain injuries thanstrength ratios. The conclusions were that a shorterThere is a growing interest in the effects ofoptimum length would place an athlete at greatereccentric exercise on muscle strain injuries, specifi-risk for muscle strain injuries.cally for the hamstrings.[9,50] The design of the bi-

Another interesting adaptation that occurs after aarticulate hamstring muscles places them at a highsingle bout of eccentric exercise is called the ‘re-risk of injury. During the late swing phase in run-peated bout effect’. If a second bout of eccentricning, the simultaneous actions of hip extension andexercise is performed that is similar to the first,knee flexion actively stretch the hamstrings tomuscle damage is significantly reduced.[6,18,85] Thisgreater lengths. Because of the prevalence of ham-adaptation can last for several weeks, and possiblystring injuries in sport, finding methods of decreas-up to 6 months.[85] The repeated bout effect caning injury rates is a high priority. Hamstring strainoccur even if there is slight muscle damage.[18] Theinjuries alone account for 6–16% of all injuries inrightward shift in the length-tension curve is consid-Australian Rules football,[75] soccer,[76] basket-ered one of the possible mechanisms for the repeat-ball,[77] cricket[78] and rugby union.[79,80] In addition,ed bout effect.[7] Several authors believe that duringthe risk of re-injuring the hamstrings is even highereccentric exercise, damage occurs on the descending(12–31%).[76,81] Previous injury is considered one of

the greatest risk factors for re-injury. It has been limb of the length-tension curve.[42,50,53] The de-



scending limb is the region on the curve that is search of this nature has proved problematic becausebeyond the optimum length of myofibril overlap. multiple spotters are needed for heavy eccentricSome have suggested that muscle strain injuries lifts; specialised and expensive equipment is oftenoccur on the descending limb of the length-tension required; there is a risk of DOMS; and there is a riskcurve.[9,50] By shifting the optimum length to greater of injury. These limitations, unfortunately, led prac-lengths, the descending limb may not be reached titioners and scientists to suggest that eccentric exer-during subsequent eccentric activities, i.e. sprinting, cise should only be performed by advanced athletes.kicking and jumping. However, given the resurgent interest in eccentrics,

Although there is a growing interest in eccentric exercises are being developed for subjects over aexercise for preventing muscle strain injuries, very range of abilities. The challenge for clinicians andfew studies have actually assessed injury rates after strength and conditioning practitioners is to be morea period of eccentric exercise. There currently are inventive in their prescription of eccentric exercise.only three published studies[49,86,87] and one unpub- Much of the literature addressing injury and ath-lished study[9] that have reported a reduction in letic performance uses the Nordic hamstring curl orhamstring injuries with eccentric exercise (table II).

similar derivations. Such exercises have limitations.It should be noted, however, that >700 athletes were

First, it is an open-chain/bilateral exercise. It is verymonitored during these four studies, and they all

likely that the stronger leg (longer optimum length)reported significant drops in hamstring injury rates.

will be strained more than the weaker leg (shorteroptimum length). Both Brockett et al.[47] and Clark

5.2 Eccentric Exercise andet al.[48] reported differences in optimum length be-

Athletic Performancetween legs. After 4 weeks of training with the Nor-dic hamstring exercise, Clark et al.[48] reported thatDespite the recent interest in eccentric exercisethe imbalance became larger between the legs. Itand the length-tension relationship, there is a gap inwas thought that the leg with a longer optimumthe literature on the effects of eccentric exercise onlength would receive more strain and continue toathletic performance. Intuitively, one would thinkadapt to a longer optimum length, while the otherthat if the mechanical properties of muscle areleg would not adapt as much. Secondly, this exercisechanging, then the functional properties would alsois a single joint exercise. The hamstrings are a bi-change. Also, since human movement typicallyarticulate muscle group, with both hip extension andtakes place when the muscles are on or near theirknee flexion functions. Injuries most frequently oc-optimum lengths,[29,57] performance may be alteredcur when the hamstrings are being actively stretchedif the optimum length is shifted. However, this is notby simultaneous hip flexion and knee extensionthe case for all muscles. Some muscles operate(during running or kicking). Multi-joint exercisessolely on the descending or ascending limbs of thewould probably be more effective at improving per-length-tension curve.[29]

formance and preventing muscle strain injuries inIt has been known for some time that eccentricthe lower body. Thirdly, the subjects are unable tocontractions can produce the greatest amounts ofsupport their own bodyweight at around 30° of kneeforce, recruit fast twitch fibres with little effort, andextension. As the subjects lower their bodyweight,place greater strain on the muscle to induce favour-gravity becomes a factor, and is a major factor atable adaptations.[14,15,29,31] Many have suggested that30°. If the effects of gravity could be manipulated, itmore research should be performed with eccentric-would be possible to go beyond 30°, since this isbased training programmes.[13,29,32] However, re-


Length-T

ension Relationship w

ith Eccentric E

xercise821

© 2007 A

dis D

ata

Info

rma

tion

BV. A

ll righ

ts rese

rved

.Sp

orts M

ed

2007; 37 (9)

Table II. Preventing hamstring strain injuries in sport

Study Subjects Eccentric exercise groups Exercise protocol Results

Askling et al.[49] Competitive soccer Yo group (n = 15), Yo curl 10wk, 4 sets of 8 reps of additional Yo group: 3 of 15 received a

players (n = 30) Yo training hamstring injury

Control group (n = 15) 10wk, regular training Control group: 10 of 15 received a

hamstring injury

Proske et al.[9]a Australian Rules Additional eccentric exercises: Pre-season training (not detailed) 2001 control season: 16 hamstring

football players straight leg deadlifts, knee curls, NH injuries

(n = NR)

2002 experimental season: 5

hamstring injuries

2003 experimental season: 2

hamstring injuries

Gabbe et al.[86]b Amateur Australian NH group (n = 114) 12wk, 5 sessions of additional NH NH group: 4% of players sustained a

Rules football players exercise (12 sets of 6 reps) hamstring injury

(n = 220)

Control group (n = 106) 12wk, regular training Control group: 13.2% of the players

sustained a hamstring injury

Brooks et al.[87] English Premier rugby S group (n = 296) Concentric and eccentric exercise S group: 1.1 hamstring injuries per

union players 2002–4 (weekly basis) 1000h

(11 clubs participated)

SSS group (n = 288) Additional hamstring stretching SSS group: 0.59 hamstring injuries

(weekly basis) per 1000h

SSN group (n = 400) Additional NH exercise 2–3 sets of SSN group: 0.39 hamstring injuries

6–7 reps (weekly basis) per 1000h

a Preliminary results.

b Pilot study.

NH = Nordic hamstring; NR = not reported; reps = repetitions; S = strengthening; SSN = stretching, strengthening and NH; SSS = static stretching and strengthening; Yo = Yoyo

hamstring.

822B

rughelli & C

ronin

© 2007 A

dis D

ata

Info

rma

tion

BV. A

ll righ

ts rese

rved

.Sp

orts M

ed

2007; 37 (9)

Table III. Enhancing athletic performance with eccentric exercise

Study Subjects Eccentric exercise groups Exercise Results

protocol type

Askling et al.[49] Competitive soccer players Yo group (n = 15) 10wk, 4 sets of 8 30m sprint Yo: 2.4% ↓ in sprint times (ES = 0.8)(n = 30) reps (Yo)

Control group (n = 15) 10wk, regular 30m sprint Control ↔training

Benn et al.[72] Physically active men (n = 9) SL control leg 10wk, 2–3 times/wk VJ SL: 6.6% ↑ (ES = 0.2)and women (n = 22). Eachserved as own control

SLS experimental leg 10wk, 2–3 times/wk VJ SLS: 8.6% ↑ (ES = 0.25)

Clark et al.[48] Amateur Australian Rules NH group (n = 9) 2–3 sets of 6–8 VJ NH: 6.6% ↑ (no SD reported)football players (n = 9). No reps, 1–3 times/wkcontrol group

Colliander and Physically active men (n = 34) Con group, Con only isokinetic 12wk, 3 times/wk VJ Con group ↔Tesch[88] dynamometer

Ecc/Con group, Con/Ecc isokinetic 4–5 sets of 12 max VJ Ecc/Con group: 7.9% ↑ (ES: 0.92)dynamometer reps

LaStayo et al.[13]a High-school basketball players HE group 8wk, 3 times/wk for VJ HE: 8% ↑ (no SD reported)(n = NR) 30 min

Hop Hz HE: 12% ↑ (no SD reported)

Control group Regular training VJ ↔

Hop Hz ↔

Mjolsnes et al.[89] Competitive soccer players NH group 10wk H/Q ratio NH: 11% ↑ in H/Q ratio(n = 23)

NH exercise 2–3 sets of 6–12reps (NH)

LC group 10wk LC ↔

LC exercise 2–3 sets of 5–10reps (LC)

a Preliminary results.

Con = concentric; Ecc = eccentric; ES = effect size; HE = high-force eccentric ergometer; H/Q = hamstring-quadriceps; Hz = frequency; LC = leg curl; max = maximum; NH =Nordic hamstring; NR = not reported; reps = repetitions; SD = standard deviation; SL = single-leg squat; SLS = single-leg squat with extra stretch load; VJ = vertical jump; Yo =Yoyo hamstring; ↑ indicates increase; ↓ indicates decrease; ↔ indicates no change.


around the same joint angle that most people pro- ability,[12,48,72] optimum hopping frequency[12] andoverall strength (table III). However, of these stud-duce peak torque during isokinetic testing. Fourthly,ies, two were unpublished,[11,12] one was a pilotafter a few weeks of this exercise, subjects are ablestudy[48] and one used a combination of eccentric/to lower their bodyweight beyond 30° (personalconcentric exercise versus concentric only.[88] Givenobservations). If an imbalance in optimum length isthese limitations, there would seem a need for aoccurring between legs it would not be appropriategreat deal more research in this area.to overload this exercise, which is the normal proce-

dure for training athletes, until this imbalance has6. Conclusionsbeen corrected.

Given the development of exercises and equip-There has been a re-emergence of interest in

ment that safely overload the eccentric contractileeccentric exercise over the last decade. The ability to

ability of muscle, greater interest in the shift inshift the optimum length with eccentric exercise

optimum length after eccentric exercise and its ef-could have several implications for performance and

fects on athletic performance may become of in-injury prevention. Altering the mechanical proper-

creasing interest. How the shift in optimum length ties of muscle could have profound effects on athlet-affects functional and athletic performance, i.e. stiff- ic potential. Also, by allowing the muscle to operateness/compliance, storage and release of elastic ener- and maintain stability at longer lengths could de-gy, fast stretch-shortening cycle performance, slow crease injury rates. Several studies have reportedstretch-shortening cycle performance, kinetics either an increase in athletic performance or a de-(force, work, power, rate of force development, and crease in injury rates with eccentric exercise. De-kinematics (acceleration, peak velocity) need to be spite the recent interest in eccentric exercise, thereinvestigated. It would appear that the shift causes are a few limitations in the literature. Currently,active stiffness to decrease at shorter muscle lengths there are no randomised, controlled training studies(due to sarcomereogenesis)[2,10,53] and passive stiff- that have reported a shift in optimum length afterness to increase at longer muscle lengths (due to titin eccentric exercise. Only one training study (pilot)elasticity).[11,12,56] A more compliant muscle-tension reported a shift after 4 weeks. There is also a lack ofunit is capable of storing more elastic energy, while studies on athletic performance after eccentric exer-a stiffer muscle-tension unit is capable of producing cise. Most studies in the past only performed heavya faster rate of power output. If the muscle is more eccentric exercise or plyometrics and examined thecompliant at the beginning of the stretch, it would be effects on performance. There is only one type ofpossible to store more elastic energy. Also, if stiff- exercise that is currently being used to prevent inju-ness increases at the end of the stretch, more energy ries and enhance performance, i.e. the Nordic ham-could be released at a higher rate. Thus, perform- string exercise or Yoyo hamstring curl. This type ofance of the stretch-shortening cycle would be great- exercise has a few inherent flaws. Being a bilaterally enhanced. It would be possible that each adapta- and open-chain exercise, the hamstrings may not betion would have a role in enhancing athletic per- worked equally between legs. Thus, imbalancesformance. may occur over time. New exercises need to be

developed that do not have these limitations.The training studies that have assessed athleticperformance after eccentric exercise have found that From the literature and a functional standpoint,performance does improve. Eccentric exercise has lower-body eccentric exercises that are designed tobeen shown to improve sprint times,[49] jumping increase the optimum length, prevent muscle inju-



12. Lindstedt S, Reich T, Keim P, et al. Do muscles function asries and enhance athletic performance should in-adaptable locomotor springs? J Exper Biol 2002; 205 (Pt 15):

clude as many of the following recommendations as 2211-613. LaStayo P, Woolf J, Lewek M, et al. Eccentric muscle contrac-possible: multi-joint movements; high/moderate ec-

tions: their contribution to injury, prevention, rehabilitation,centric force activities; muscles operating at long and sport. J Orthop Sports Phys Ther 2003; 33 (10): 557-71lengths; closed-chain exercises; high/moderate ve- 14. Katz B. The relation between force and speed in muscular

contraction. J Physiol 1939; 96: 45-64locity; easy to implement exercises; cost-effective15. Bigland-Ritchie B, Woods J. Integrated electromyogram and

exercises; high volume (if an acute shift is desira- oxygen uptake during positive and negative work. J Physiol1976; 260 (2): 267-77ble); and unilateral exercises. Exercise design

16. Lastayo P, Reich T, Urquhart M, et al. Chronic eccentric exer-should take a multi-joint and multi-movement ap-cise: improvements in muscle strength can occur with little

proach. This will allow greater prevention against demand for oxygen. Am J Physiol 1999; 276 (2 Pt 2): R611-1517. Newham D, Jones D, Ghosh G, et al. Muscle fatigue and painactive strain injuries to a larger number of muscles,

after eccentric contractions at long and short length. Clin Sciand possibly improve performance. 1988; 74: 553-7

18. Clarkson P, Tremblay I. Exercise-induced muscle damage, re-pair, and adaptation in humans. J Appl Physiol 1988; 65 (1):

Acknowledgements 1-619. Ebbeling C, Clarkson P. Exercise-induced muscle damage and

adaptation. Sports Med 1989; 7: 207-34The authors received no funding for the preparation of this20. Lieber R, Schmitz M, Mishra D, et al. Contractile and cellulararticle and have no conflicts of interest directly relevant to its

remodeling in rabbit skeletal muscle after cyclic eccentriccontents. contractions. J Appl Physiol 1994; 77 (4): 1926-34

21. Lieber R, Thornell L, Friden J, et al. Muscle cytoskeletal disrup-tion occurs within the first 15 min of cyclic eccentric contrac-tion. J Appl Physiol 1996; 80 (1): 278-84References

22. Friden J, Lieber R. Eccentric exercise-induced injuries to con-1. Allen D. Eccentric muscle damage: mechanisms of early reduc-tractile and cytoskeletal muscle fibre components. Acta Physi-tion of force. Acta Physiol Scand 2001; 171 (3): 311-9ol Scand 2001; 171 (3): 321-62. Proske U, Allen T. Damage to skeletal muscle from eccentric

23. MacIntyre D, Reid W, Lyster D, et al. Presence of WBC,exercise. Exerc Sport Sci Rev 2005; 33 (2): 89-104decreased strength, and delayed soreness in muscle after ec-3. Clarkson P, Hubal M. Exercise-induced muscle damage incentric exercise. J Appl Physiol 1995; 80 (3): 1006-13humans. Am J Phys Med Rehabil 2002; 81 (11 Suppl.): S52-69

24. Nosaka K, Clarkson P. Changes in indicators of inflammation4. Morgan D, Allen D. Early events in stretch-induced muscleafter eccentric exercise of the elbow flexors. Med Sci Sportsdamage. J Appl Physiol 1999; 87: 2007-15Exerc 1996; 28 (8): 953-615. MacIntyre D, Reid W, McKenzie D, et al. Delayed muscle

25. Ingalls C, Warren G, Williams J, et al. E-C coupling failure insoreness: the inflammatory response to muscle injury and itsmouse EDL muscle after in vivo eccentric contractions. J Applclinical implications. Sports Med 1995; 20: 24-40Physiol 1998; 85 (1): 58-676. Newham D, Jones D, Clarkson P, et al. Repeated high-force

26. Warren G, Ingalls C, Lowe D, et al. Excitation-contractioneccentric exercise: effects on muscle pain and damage. J Appluncoupling: major role in contraction-induced muscle injury.Physiol 1987; 63 (4): 1381-6Exerc Sport Sci Rev 2001; 29 (2): 82-77. McHugh M. Recent advances in the understanding of the repeat-

27. Warren G, Lowe D, Sayes D, et al. Excitation failure in eccen-ed bout effect: the protective effect against muscle damagetric contraction-induced injury of mouse soleus muscle. Jfrom a single bout of eccentric exercise. Scand J Med SciPhysiol 1993; 468: 487-99Sports 2003; 13 (2): 88-97

28. Bobbert M, Van Soest A. Why do people jump the way they do?8. Proske U, Morgan D. Muscle damage from eccentric exercise:Exerc Sport Sci Rev 2001; 29 (3): 95-102mechanism, mechanical signs, adaptation and clinical applica-

29. Leiber L. Skeletal muscle structure, function and plasticity. 2ndtions. J Physiol 2001; 537 (2): 333-45ed. Philadelphia (PA): Lippincott Williams and Wilkins, 20029. Proske U, Morgan D, Brockett C, et al. Identifying athletes at

30. Gordon A, Huxley A, Julian F, et al. The variation in isometricrisk of hamstring strains and how to protect them. Clin Expertension with sarcomere length in vertebrate muscle fibers. JPharmacol Physiol 2004; 31 (8): 546-50Physiol 1966; 184: 170-9210. Morgan D, Proske U. Popping sarcomere hypothesis explains

31. Enoka R. Neuromechanics of human movement. 3rd ed. Cham-stretch-induced muscle damage. Clin Exper Pharmacol Physi-paign (IL): Human Kinetics, 2002ol 2004; 31 (8): 541-5

11. Lindstedt S, LaStayo P, Reich T, et al. When active muscles 32. Rassier D, Macintosh B, Herzon W, et al. Length dependence oflengthen: properties and consequences of eccentric contrac- active force production in skeletal muscle. J Appl Physioltions. News Physiol Sci 2001; 16: 256-61 1999; 86 (5): 1445-57



33. Talbot J, Morgan D. Quantitative analysis of sarcomere non- 50. Brockett C, Morgan D, Proske U, et al. Predicting hamstringuniformities in active muscle following a stretch. J Muscle Res strain injury in elite athletes. Med Sci Sports Exerc 2004; 36Cell Motility 1996; 17 (2): 261-8 (3): 379-87

51. McHugh M, Tetro D. Changes in the relationship between joint34. Wood S, Morgan D, Proske U, et al. Effects of repeated eccen-angle and torque production associated with the repeated bouttric contractions on structure and mechanical properties of toadeffect. J Sports Sci 2003; 21 (11): 927-32sartorius muscle. Am J Physiol 1993; 265 (3 Pt 1): C792-800

52. Balnave C, Allen D. Intracellular calcium and force in single35. Morgan D, Claflin D, Julian F, et al. The effects of repeatedmouse muscle fibres following repeated contractions withactive stretches on tension generation and myoplasmic calciumstretch. J Physiol 1995; 488 (Pt 1): 25-36in frog single muscle fibres. J Physiol 1996; 497 (Pt 3): 665-74

53. Morgan D. New insights into the behavior of muscle during36. Butterfield T, Herzog W. Is the force-length relationship aactive lengthening. Biophys J 1990; 57: 209-21useful indicator of contractile element damage following ec-

centric exercise? J Biomech 2005; 38 (9): 1932-7 54. Newham D, Mills K, Quigley B, et al. Pain and fatigue afterconcentric and eccentric muscle contractions. Clin Sci 1983;37. Brockett C, Morgan D, Gregory J, et al. Damage to different64: 55-62motor units from active lengthening of the medial gas-

trocnemius muscle of the cat. J Appl Physiol 2002; 92 (3): 55. Friden J, Sjostrom M, Ekblom B, et al. A morphological study1104-10 of delayed muscle soreness. Experientia 1981; 37: 506-7

38. Jones C, Allen T, Talbot J, et al. Changes in the mechanical 56. Reich T, Lindstedt S, LaStayo P, et al. Is the spring quality ofproperties of human and amphibian muscle after eccentric muscle plastic? Am J Physiol Regul Integr Comp Physiolexercise. Eur J Appl Physiol Occup Physiol 1997; 76 (1): 2000; 278 (6): R1661-621-31 57. Leiber L, Friden J. Functional and clinical significance of skele-

39. Whitehead N, Allen T, Morgan D, et al. Damage to human tal muscle architecture. Muscle Nerve 2000; 23: 1647-66muscle from eccentric exercise after training with concentric 58. Herzog W, ter Keurs H. Force-length relation of in-vivo humanexercise. J Physiol 1998; 512 (Pt 2): 615-20 rectus femoris muscles. Eur J Physiol 1988; 11 (6): 642-7

40. Whitehead N, Weerakkody N, Gregory J, et al. Changes in 59. Herzog W, Guimaraes A, Anton MG, et al. Moment-lengthpassive tension of muscle in humans and animals after eccen- relations of rectus femoris muscles of speed skaters/cycliststric exercise. J Physiol 2001; 533 (2): 593-604 and runners. Med Sci Sports Exerc 1991; 23 (11): 1289-96

41. Saxton J, Donnelly A. Length-specific impairment of skeletal 60. Lynn R, Morgan D. Decline running produces more sarcomeresmuscle contractile function after eccentric muscle actions in in rat vastus intermedius muscle fibers than does incline run-man. Clin Sci 1996; 90 (2): 119-25 ning. J Appl Physiol 1994; 77 (3): 1439-44

42. Bowers E, Morgan D, Proske U, et al. Damage to the human 61. Lynn R, Talbot J, Morgan D, et al. Differences in rat skeletalquadriceps muscle from eccentric exercise and the training muscles after incline and decline running. J Appl Physioleffect. J Sports Sci 2004; 22 (11-12): 1005-14 1998; 85 (1): 98-104

43. Byrne C, Eston R, Edwards R, et al. Characteristics of isometric 62. Butterfield T, Herzog W. The magnitude of muscle strain doesand dynamic strength loss following eccentric exercise-in- not influence serial sarcomere number adaptations followingduced muscle damage. Scand J Med Sci Sports 2001; 11: eccentric exercise. Eur J Physiol 2006; 451 (5): 688-700134-40 63. Butterfield T, Leonard T, Herzog W, et al. Differential serial

44. Pettitt R, Symons D, Eisenman P, et al. Eccentric strain at long sarcomere number adaptations in knee extensor muscles of ratsmuscle length evokes the repeated bout effect. J Strength Cond is contraction type dependent. J Appl Physiol 2005; 99 (4):Res 2005; 19 (4): 918-24 1352-8

45. Philippou A, Bogdanis G, Nevill A, et al. Changes in the angle- 64. Koh T, Herzog W. Eccentric training does not increaseforce curve of human elbow flexors following eccentric and sarcomere number in rabbit dorsiflexor muscles. J Biomechisometric exercise. Eur J Appl Physiol 2004; 93 (1-2): 237-44 1998; 31 (5): 499-501

46. Prasartwuth O, Allen T, Butler J, et al. Length-dependent 65. Trappe T, Carrithers J, White F, et al. Titin and nebulin contentchanges in voluntary activation, maximum voluntary torque in human skeletal muscle following eccentric resistance exer-and twitch responses after eccentric damage in humans. J cise. Muscle Nerve 2002; 25: 289-92Physiol 2006; 571 (Pt 1): 243-52 66. Barash I, Peters D, Friden J, et al. Desmin cytoskeletal modifi-

47. Brockett C, Morgan D, Proske U, et al. Human hamstring cations after a bout of eccentric exercise in the rat. Am Jmuscles adapt to eccentric exercise by changing optimum Physiol Regul Integr Comp Physiol 2002; 284 (4): R958-63length. Med Sci Sports Exerc 2001; 33 (5): 783-90 67. Wang K, McCarter R, Wright J, et al. Regulation of skeletal

48. Clark R, Bryant A, Culgan J, et al. The effects of hamstring muscle stiffness and elasticity by titin isoforms: a test of thestrength training on dynamic jumping performance and segmental extension model of resting tension. Proc Natl Acadisokinetic strength parameters. Phys Ther Sport 2005; 6: 67-73 Sci 1991; 88: 7101-5

49. Askling C, Karlsson J, Thorstensson A, et al. Hamstring injury 68. Spiers I, Akster H, Granzier H, et al. Expression of titin isoformsoccurrence in elite soccer players after preseason strength in red and white muscle fibres of carp exposed to differenttraining with eccentric overload. Scand J Med Sci Sports 2003; sarcomere strains during swimming. J Comp Physiol 1997;13 (4): 244-50 167: 543-51



69. McBride J, Triplett-McBride T, Davie A, et al. Characteristics 81. Crosier J. Factors associated with recurrent hamstring injuries.of titin in strength and power athletes. Eur J Appl Physiol Sports Med 2004; 34: 681-952003; 88 (6): 553-7

82. Garrett W. Muscle strain injuries. Am J Sports Med 1996; 24:70. McGuigan M, Sharman M, Newton R, et al. Effect of explosive S2-8

resistance training on titin and myosin heavy chain isoforms in83. Thelen D, Chumanov D, Hoerth M, et al. Hamstring muscletrained subjects. J Strength Cond Res 2003; 17 (4): 645-51

kinematics during treadmill sprinting. Med Sci Sports Exerc71. Pousson M, van Hoeche J, Goubel F, et al. Changes in elastic 2005; 38: 108-14

characteristics of human muscle induced by eccentric exercise.84. Verrall G, Slavotinek J, Barnes P, et al. Clinical risk factors forJ Biomech 1990; 23: 343-8

hamstring muscle strain injury: a prospective study with corre-72. Benn C, Forman K, Mathewson D, et al. The effects of serial

lation of injury by magnetic resonance imaging. Br J Sportsstretch loading on stretch work and stretch-shortening per-Med 2001; 35 (6): 435-9formance in the knee musculature. J Orthop Sports Phys Ther

1998; 27: 412-22 85. Nosaka K, Clarkson P. Muscle damage following repeated bouts

of high force eccentric exercise. Med Sci Sports Exerc 1995;73. Labeit D, Watanabe K, Witt C, et al. Calcium-dependent molec-27 (9): 1263-9ular spring elements in the giant protein titin. Proc Natl Acad

Sci 2003; 100 (23): 13716-21 86. Gabbe B, Branson R, Bennell K. A pilot randomised controlled74. Labeit S, Kolmerer B. Titins: giant proteins in charge of muscle trial of eccentric exercise to prevent hamstring injuries in com-

ultrastructure and elasticity. Science 1995; 270 (5234): 293-6 munity-level Australian Football. J Sci Med Sport 2006; 9

(1-2): 103-975. Arnaso A, Sigurdsson S, Gudmundsson A, et al. Risk factors forinjuries in football. Am J Sports Med 2004; 32 Suppl. 1: 5-16S 87. Brooks J, Fuller C, Kemp S, et al. Incidence, risk, and preven-

tion of hamstring muscle injuries in professional rugby union.76. Woods C, Hawkins R, Maltby S, et al. The Football AssociationMedical Research Programme: an audit of injuries in profes- Am J Sports Med 2006; 34 (8): 1297-306sional football -analysis of hamstring injuries. Br J Sports Med 88. Colliander E, Tesch P. Effects of eccentric and concentric2004; 38: 36-41

muscle actions in resistance training. Acta Physiol Scand77. Meeuwisse W, Sellmer R, Hagel B, et al. Rates and risks 1990; 140: 31-9

of injury during intercollegiate basketball. Am J Sports Med89. Mjolsnes R, Arnason A, Osthagen T, et al. A 10-week random-2003; 31: 379-85

ized trial comparing eccentric vs concentric hamstring strength78. Orchard J, James T, Alcott E, et al. Injuries in Australian cricket

training in well-trained soccer players. Scand J Med Sci Sportsat first class level 1995/1996 to 2000/2001. Br J Sports Med

2004; 14 (5): 311-72002; 36: 270-5

79. Brooks J, Fuller C, Kemp S, et al. Epidemiology of injuries inEnglish professional rugby union, part 1: match injuries. Br J Correspondence: Matt Brughelli, School of Exercise,Sports Med 2005; 39: 757-66

Biomedical and Health Sciences, Edith Cowan University,80. Brooks J, Fuller C, Kemp S, et al. Epidemiology of injuries in

100 Joondalup Drive, Joondalup, WA 6027, Australia.English professional rugby union, part 2: training injuries. Br JSports Med 2005; 39: 767-75 E-mail: [email protected]


altering the length-tension relationship with eccentric exercise

Documents