ARTICLE IN PRESS

Journal of Biomechanics 43 (2010) 1503–1508

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jbiomech

Journal of Biomechanics

0021-92

doi:10.1

n Corr

E-m

www.JBiomech.com

Evidence of residual force enhancement for multi-joint leg extension

Daniel Hahn n, Wolfgang Seiberl, Susanne Schmidt, Katrin Schweizer, Ansgar Schwirtz

Department of Biomechanics in Sports, Faculty of Sport Science, Technische Universitat Munchen, Connollystr. 32, 80809 Munich, Germany

a r t i c l e i n f o

Article history:

Accepted 20 January 2010Force enhancement is a well accepted property of skeletal muscle and has been observed at all

structural levels ranging from single myofibrils to voluntarily activated m. quadriceps femoris in vivo.

Keywords:

Force–time history

Force potentiation

Eccentric

Muscle activation

In vivo

90/$ - see front matter & 2010 Elsevier Ltd. A

016/j.jbiomech.2010.01.041

esponding author. Tel.: +49 89 28924583; fa

ail address: d.hahn@sp.tum.de (D. Hahn).

a b s t r a c t

However, force enhancement has not been studied for multi-joint movements like human leg

extension; therefore knowledge about its relevance in daily living remains limited. The purpose of this

study was to determine whether there is force enhancement during maximal voluntary multi-joint leg

extension. Human leg extension was studied (n=22) on a motor driven leg press dynamometer where

external reaction forces under the feet as well as activity of 8 lower extremity muscles were measured.

In addition, torque in the ankle and knee joints was calculated using inverse dynamics. The steady-state

isometric force, joint torques, and muscle activation after active stretch (201 stretch amplitude at 601/s)

were compared with the corresponding values obtained during isometric reference contractions. There

was consistent force enhancement during and following stretch for both forces and joint torques.

Potentiation during stretch reached values between 26% and 30%, while a significant force enhancement

of 10.5–12.3% and 4.3–7.4% remained 0.5–1 and 2.5–3 s after stretch, respectively. During stretch, EMG

signals of m. gastrocnemius medialis and lateralis were significantly increased, while following stretch

all analyzed muscles showed the same activity as during the reference contractions. We conclude from

these results that force enhancement exists in everyday movements and should be accounted for when

analyzing or modelling human movement.

& 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Eccentric contractions of activated muscles cause a transientincrease in force during stretch (Hill, 1938; Joyce et al., 1969;Edman et al., 1978) as well as a persistent increase in force in thesteady-state phase following stretch, when compared with apurely isometric contraction at corresponding muscle length. Thisso called residual force enhancement (RFE) has been observed forover 50 years (Abbott and Aubert, 1952) and is a property of allmuscles and preparations (Sugi, 1972; Edman et al., 1982; Herzogand Leonard, 2000; Rassier et al., 2003; Joumaa et al., 2008; Mehtaand Herzog, 2008). From a more functional point of view, it isimportant to ascertain whether RFE occurs in everyday move-ments when muscles are working in situ as part of theneuromuscular system.

Several studies investigated RFE during in vivo experiments onhumans (Cook and McDonagh, 1995; De Ruiter et al., 2000; Leeand Herzog, 2002; Oskouei and Herzog, 2005; Hahn et al., 2007;Pinniger and Cresswell, 2007; Tilp et al., 2009). These studies havebeen conducted using electrical stimulation as well as voluntaryactivation. Furthermore, muscles of different sizes, from small(Cook and McDonagh, 1995), medium sized (Pinniger and

ll rights reserved.

x: +49 89 28924582.

Cresswell, 2007), to large (Hahn et al., 2007; Seiberl et al., 2008)have been analyzed. For electrically stimulated muscles thefindings on RFE agree well with the results found in vitro (Cookand McDonagh, 1995; De Ruiter et al., 2000; Lee and Herzog,2002; Pinniger and Cresswell, 2007). Similarly, RFE of 4–16% hasbeen observed for voluntary contractions at different activationlevels (Lee and Herzog, 2002; Oskouei and Herzog, 2005; Pinnigerand Cresswell, 2007; Tilp et al., 2009). However, while themagnitude of RFE for human adductor pollicis was directly relatedto activation level (Oskouei and Herzog, 2005, 2006), no suchdependency was found for human plantar flexor muscles(Pinniger and Cresswell, 2007; Tilp et al., 2009). Furthermore,RFE of voluntarily activated muscles was not always long lasting,but reached isometric level 2 s after stretch (Hahn et al., 2008;Tilp et al., 2009). Finally, RFE during voluntary contractions wasneither directly dependent on stretch velocity nor on stretchamplitude (Lee and Herzog, 2002; Hahn et al., 2007; Tilp et al.,2009).

Although there is wide evidence that RFE exists for voluntarilyactivated human muscles, there have been no studies that haveinvestigated the presence of RFE in an everyday multi-jointmovement. Therefore the purpose of this study was to determinewhether RFE occurs during and following stretch when perform-ing a multi-joint leg extension. We hypothesize that RFE will bepresent at the level of external reaction forces as well as jointtorques, revealing a relevance of RFE for human movement.

ARTICLE IN PRESS

00

200400600800

100012001400160018002000A

B

100

110

120

130

FS3FS2FS1DS

isometric contraction stretch contraction

exte

rnal

reac

tion

forc

e [N

]

hip joint

t fle

xion

ang

le [°

]

1 2 3 4 5 6 7 8

D. Hahn et al. / Journal of Biomechanics 43 (2010) 1503–15081504

2. Methods

2.1. Subjects

Twenty-two healthy male subjects (27.076.1 yr, 1.8170.05 m and

76.776.1 kg) with no history of ankle, knee, or hip joint injury or neurological

disorder participated in the study. Written informed consent was obtained from

all subjects before participation in the experiments and the study was approved by

the Human Ethics Committee of the local institute.

2.2. Experimental set-up

Bilateral leg extensions were performed using a motor-driven leg press

dynamometer (IsoMed2000, D&R Ferstl GmbH, Germany). Subjects were placed on

the dynamometer with the vertical backrest reclined to 501. The pelvis was

secured by a safety belt and the upper body by two safety belts and two shoulder

pads. In relation to a vertical axis, the footrest was rotated 151 towards plantar

flexion. Foot placement was 0.1 m above the dynamometers seat (Fig. 1).

According to previous investigations (Lee and Herzog, 2002; Hahn et al., 2007),

stretches with a mean amplitude of 17.871.71 were performed at an angular

velocity of 601/s from the initial start position of 77.574.51 to the end position of

95.374.11 knee flexion, where 01 knee flexion refers to the straight leg. Individual

acceleration and translational velocities of the motor-driven leg press footrest that

resulted in the desired angular velocity of 601/s for each subject were calculated

according to Hahn et al. (2005).

2.3. Force measurement, assessment of muscle activation, and movement analysis

External reaction forces (Fext) were measured for each leg by two custom made

tri-axial force plates mounted to the leg press footrest. Muscle activation of Mm.

rectus femoris (RF), vastus medialis (VM), vastus lateralis (VL), biceps femoris (BF),

gastrocnemius medialis (GM), m. gastrocnemius lateralis (GL), soleus (SO), and

tibialis anterior (TA) was obtained using surface electromyography (EMG). Bipolar

EMG electrodes (Ambus Blue Sensor P-Electrodes, Denmark) were placed on the

muscles with an inter-electrode distance of approximately 20 mm, while a single

reference electrode was attached to the patella. Skin preparation and electrode

placement were done according to the SENIAM recommendations (Hermens et al.,

1999). Based on the markerset of the Newington–Helen Hayes Model (Charlton

et al., 2004) a six-camera VICON MX-3 Motion-Analysis-System (Oxford Metrics,

UK) served for measuring lower extremity kinematics. Due to the recumbent

position in the leg press, the markerset was slightly modified so that the markers

on the left and right posterior spina illiaca were placed on the left and right iliac

crest, respectively. Furthermore, 8 additional markers were attached to the force

plates to assess their position. All data were synchronized and collected at a

sample rate of 1000 Hz (forces and EMG signals) and 250 Hz (kinematic data).

2.4. Experimental protocol

In three training sessions prior to the test protocol subjects were familiarized

with the procedures and trained to perform maximum voluntary contractions.

Subjects were instructed to increase their leg extension forces to the maximum

level as fast as possible, and then to maintain the maximal force for the duration

of the test. The start and finish of each contraction was clearly announced

and verbal encouragement was given at all times. The actual test protocol

started with two maximum voluntary isometric contractions at initial start

position to define a 95% trigger for the stretch contractions. Thereafter, three

Fig. 1. Experimental setting in the leg press dynamometer. The figure shows a

subject placed on the dynamometer seat with reflective markers attached to the

lower extremity and the force plates, as well as EMG attached to analyzed lower

limb muscles.

isometric–eccentric–isometric and three purely isometric contractions at the end

position were performed in a randomized order. The 95% threshold was used to

guarantee maximal voluntary contraction level as well as to control fatigue

because the leg press motor did not start eccentric leg extensions when 95% of

individual leg extension force was not reached. Isometric and stretch contractions

were maintained for 5 s so that following stretching, muscles were kept activated

for approximately 4 s to ensure a steady-state force following the stretch-transient

force response (Fig. 2). Subjects were given as much rest as required, but a

minimum rest of 3 min was enforced at all times.

2.5. Modelling

Kinematics and kinetics of the knee and ankle joints were calculated using

inverse dynamics. To account for inertial properties the anthropometric model

was scaled to individual weight and body height by linear regression (Zatsiorsky

et al., 1984). In addition, Zatiorsky’s segment inertia parameters were adjusted

according to de Leva (1996).

2.6. Data analysis

The data from the right leg were used for data analysis. Dependent variables

were the resultant Fext, plantar flexion torque, and knee extension torque. After

inverse dynamic calculations but prior to data analysis, forces and joint torques

were smoothed by a 4th order Butterworth low pass filter with cut-off frequencies

of 20 and 8 Hz, respectively. Similar to previous studies (Hahn et al., 2007),

parameters were determined at four specific instances in time: during stretch

(DS), forces/torques were analyzed at the end of the stretch, and just before the

motor started to decelerate the eccentric movement. Additionally, forces/torques

following stretch (FS) were assessed over 500 ms intervals from 0.5 to 1 s (FS1),

1.5 to 2 s (FS2), and 2.5 to 3 s (FS3) after stretch, respectively (Fig. 2). Forces and

10

20

70

80

90

ankle joint

knee joint

dors

al e

xten

sion

/ jo

in

time [s]0 1 2 3 4 5 6 7 8

Fig. 2. Exemplar external reaction force (A) and corresponding hip, knee, and

ankle joint angle (B) time histories for a purely isometric contraction (grey lines)

and isometric–eccentric–isometric stretch contraction (black lines). Note that at

the beginning of the contractions prior to the stretch, joint angles change due to

deformation of soft tissues and seat cushion until a steady state is reached. Note

further that the isometric force during the first phase of the stretch contraction

exceeds the corresponding purely isometric force since external reaction force of

multi-joint leg extension decreases with increasing leg flexion. The dotted lines

indicate the instances in time used for data analysis.

ARTICLE IN PRESS

130

140

VC

]

external reaction force

D. Hahn et al. / Journal of Biomechanics 43 (2010) 1503–1508 1505

torques of the isometric–eccentric–isometric contractions were normalized

relative to the corresponding parameters of the isometric reference contractions.

To account for the effect on RFE caused by potential differences in joint

positions when comparing the different conditions, linear interpolation was used.

This is based on the shape of the force–length relationship (F–l-r) of multi-joint leg

extension. For multi-joint leg extension there is less force when the leg is flexed

more and the decline in force is almost linear for the range of motion considered in

this study (Hugh-Jones, 1947; Hahn, 2008).

Muscle activation level of EMG signals was analyzed for all muscles at the

same instances in time and for both contraction conditions. Muscle activation level

was calculated after band pass filtering (10–500 Hz) and calculation of root mean

square (RMS) values of the EMG signals (for a 500 ms period). Subsequently, RMS

values at DS and FS1–FS3 were normalized per muscle per subject to the peak RMS

value that occurred during the MVC at the end position.

2.7. Statistics

After determining the best trials of each subject (i.e. the trial with the greatest

leg extension force), mean values 7SD were calculated and served for statistical

analysis. Normality of the data was tested using the Kolmogorov–Smirnov test and

a two way repeated ANOVA and Bonferroni–Holm post hoc comparisons were

used to identify significant differences at defined instances in time and between

contraction conditions. Statistical analysis was performed using SPSS 17.0 for

Windows (SPSS, Chicago, IL) and alpha level was set to pr0.05.

DS90

100

110

120

Forc

e no

rmal

ized

[% o

f M

instance in time

external reaction force corrected

FS1 FS2 FS3

Fig. 3. External reaction force at different instances in time during (DS) and

following (FS) stretch. External reaction force was normalized relative to the

corresponding values of the isometric reference contraction and represents

the magnitude of force enhancement. Grey columns show original data, while

the white columns are corrected for differences in joint kinematics between

contraction conditions. Both original as well as corrected values differ significantly

(po0.01) from corresponding isometric reference. Figure shows mean value7SD.

150

knee joint torque

3. Results

3.1. Potentiation during stretch

Our results showed significantly (po0.001) enhanced Fext of28.477.4% at the end of stretch. Similarly, we found enhanced kneejoint torque of 26.379.8% as well as potentiated ankle joint torqueof 30.1717.1% (po0.001, Table 1). After correction of Fext due toless flexion in the hip, knee, and ankle joint of about 2.172.41,2.171.51, and 1.471.71, respectively, a significant (po0.001)potentiation of 23.177.3% (po0.001) remained (Fig. 3).

3.2. Residual force enhancement

Both Fext and torques in the knee (MK) and ankle (MA) jointwere consistently higher in the isometric phase followingstretching compared with the isometric reference contraction.For FS1 we found significant RFE of 10.578.9% (po0.001),12.279.3% (po0.001), and 12.3719.5% (po0.05) for Fext, kneejoint torque, and ankle joint torque, respectively. For FS2, valuesof RFE reached 8.577.7% (Fext, po0.001), 9.478.9% (MK,po0.001), and 8.3719.0% (MA, po0.05), whereas for FS3, RFEwas 6.876.2% (Fext, po0.001), 7.478.1% (MK, po0.001), and

Table 1External reaction forces and joint torques for isometric and stretch contractions at

different instances in time.

Contraction condition Force [N] or torque [Nm] at instant in time

DS FS1 FS2 FS3

External reaction force

Iso–ecc–isoa 12707214 10727171 10147176 9577164

Isometricb 9937182 9797190 9397178 8997163

Knee joint torque (MK)

Iso–ecc–iso. 196729 167726 157727 150726

Isometric 156726 150726 145726 140723

Ankle joint torque (MA)

Iso–ecc–iso 157737 129736 122737 115735

Isometric 122732 118740 114737 110734

Values are mean7SD. See the Methods section for description of different

instances in time (DS, FS1–FS3).

a Isometric–eccentric–isometric contraction involving stretch.b Purely isometric contraction.

4.3712.1% (MA, po0.05), (Figs. 3 and 4). The decrease of RFE overtime from FS1 to FS3 was significant (po0.01) for all dependentvariables (Table 1). Using linear interpolation to account forreduced leg flexion following stretch (0.772.41, 0.871.21, and0.471.91 for the hip, knee, and ankle joint; po0.01), we foundthat Fext as measured at FS1–FS3 slightly overestimates RFE byabout 2.273.3% (Fig. 3).

3.3. Muscle activation

At the end of stretch, muscle activation (%MVC) of GM andGL muscles was significantly higher (102.5735.0% (GM), 98.1730.9% (GL), both po0.001) in comparison with the isometricreference contraction (68.3719.2% (GM), 62.7718.9% (GL)). For

DS90

100

110

120

130

140

Torq

ue n

orm

aliz

ed [%

MV

C]

instance in time

ankle joint torque

FS1 FS2 FS3

Fig. 4. Knee (grey columns) and ankle joint (white columns) torques at different

instances in time during (DS) and following (FS) stretch. Values are normalized to

the corresponding isometric joint torques to represent potentiation during and

following stretch contractions. Potentiation showed significance (po0.05) for

both joints and all instances in time. Figure shows mean value7SD.

ARTICLE IN PRESS

Table 2Muscle activation of 8 lower limb muscles for isometric and stretch contractions at different instances in time.

Muscle Contraction condition Activation level [%MVC] at instant in time

DS FS1 FS2 FS3

Rectus femoris Iso–ecc–isoa 92718 85715 84715 79715

Isometricb 86710 8479 8479 81711

Vastus medialis Iso–ecc–iso 93726 82720 78720 74719

Isometric 79716 80716 77717 77718

Vastus lateralis Iso–ecc–iso 92713 8378 80710 78712

Isometric 8678 8678 8376 7779

Biceps femoris Iso–ecc–iso 75726 68725 66725 64723

Isometric 73722 75719 76719 71720

Gastrocnemius medialis iso–ecc–iso 103735 75725 69725 67726

Isometric 68719 72723 67721 64721

Gastrocnemius lateralis Iso–ecc–iso 98731 61723 55721 55726

Isometric 63719 64722 60722 57721

Soleus Iso–ecc–iso 102733 77727 70725 67730

Isometric 73714 74716 69717 67715

Tibialis anterior Iso–ecc–iso 68726 51734 54735 55735

Isometric 49732 53731 57734 53736

Values are mean7SD. See the Methods section for description of different instances in time (DS, FS1–FS3).

a Isometric–eccentric–isometric contraction involving stretch.b Purely isometric contraction.

98

100

isometric contraction stretch contraction

°]

D. Hahn et al. / Journal of Biomechanics 43 (2010) 1503–15081506

all other muscles, no differences in activation between contrac-tion conditions were observed. Following stretch there was asignificant decrease of the MVC level over time for VM, RF, GM,and GL, whereas MVC level of VL, GM, and SOL muscles showed asignificant decline over time during the purely isometric referencecontraction (Table 2).

DS88

90

92

94

96

knee

join

t fle

xion

[

instance in timeFS1 FS2 FS3

Fig. 5. Knee joint flexion angles during purely isomeric contraction (white

squares) and isometric–eccentric–isometric (black squares) stretch contraction

at different instances in time during (DS) and following (FS) stretch. In comparison

with the purely isometric contractions, joint flexion during stretch contractions

were significantly (po0.01) reduced at all instances in time. Figure shows mean

value7SD.

4. Discussion

This is the first study to show that RFE is present during multi-joint leg extension, which is a typical everyday movementpattern. Thus the results indicate that RFE is a functionallyrelevant characteristic of human skeletal muscle in daily activity.

4.1. Force enhancement (FE) during stretch

In agreement with recent studies (Lee and Herzog, 2002; Hahnet al., 2007; Tilp et al., 2009), but in contrast with former findings(Westing et al., 1988; Dudley et al., 1990; Webber and Kriellaars,1997), Fext and joint torques at the end of stretch exceeded thecorresponding isometric values of about 26–30%. There areseveral factors to take into account that contribute to thesedifferences.

First, as shown by experiments on isolated muscle prepara-tions, eccentric FE is a basic property of skeletal muscle (Katz,1939; Joyce et al., 1969). Hence, the absence of FE during stretchin vivo has often been associated with neural inhibition (Westinget al., 1991; Webber and Kriellaars, 1997; Babault et al., 2001;Beltman et al., 2004). However, from studies that found FE duringstretch for voluntary activation there is evidence that neuralinhibition can be reduced by training (Aagaard et al., 2000;Spurway et al., 2000) or depends on the activity level of subjects(Amiridis et al., 1996). While our subjects were stronglyfamiliarized to perform maximal voluntary eccentric contractionsthere is no such information for subjects of studies who did notfind FE during stretch. Secondly, not all the potentiation duringstretch that we found seems to be due to muscular properties.Kinematic analysis showed that joint angles differed in compar-

ison between contraction conditions (Fig. 5) and result from thefact that all parameters during stretch were analyzed prior to theend of stretch. The correction by linear regression thereforereveals that Fext as measured during stretch overestimates forcepotentiation by about 5.474.3%. Similarly, we assume thatpotentiation at the joint torque level was also overestimated.This assumption is based on the analysis of torques before andafter stretch during the stretch contractions, which reveal that alljoints are acting on the descending limb of their multi-jointtorque–angle relation (po0.001). However, as joint torques havenot been calculated during the isometric contractions at the initialknee flexion angle, it was not possible to do linear interpolations

ARTICLE IN PRESS

060708090

100110

255

270

285

300

315

330 isometric contraction stretch contraction

COP

knee flexion

flexi

on a

ngle

[°] /

pos

ition

of C

OP

[mm

]

time [s]1 2 3 4 5

Fig. 6. Variations in centre of pressure (COP) during different contraction

conditions but identical foot placement on the force plate. The figure shows time

histories of the vertical COP position (upper trace) and the corresponding knee

joint angles (lower trace) during a purely isometric contraction (grey lines) and a

isometric–eccentric–isometric stretch contraction (black lines) for a single subject.

Note that there is a major difference in COP positions during the stretch phase of

the isometric–eccentric–isometric contraction.

-150

10

20

30

40

50

60

70

80

90

r = 0.74

FE [%

]

COP shift [% of foot length]-10 -5 0 5 10 15 20 25

Fig. 7. Correlation between the shift of COP position and force enhancement (FE)

during stretch (r=0.74, po0.001). COP shift is normalized to foot length, which

was calculated as the vertical distance between toe and heel markers. Positive

values represent an upwards shift of COP position towards the toe marker,

whereas negative values represent a COP shift towards the heel.

D. Hahn et al. / Journal of Biomechanics 43 (2010) 1503–1508 1507

in order to obtain quantitative estimates. Further it has to betaken into consideration that the correction by linear regressiondoes not account for changes in tendon and fibre length due toincreased forces or torques. Therefore, it cannot be excluded thatadditional tendon and fibre length changes affect the actualmuscular contribution to FE during stretch.

Third, FE during stretch was affected by technique variationsdue to shifting of the centre of pressure (COP) under the foot(Fig. 6) in conjunction with changing the line of action (LOA) ofFext. During stretch, COP shifted 3.678.5% of foot length(calculated as the vertical distance between the toe and heelmarkers) towards the ball of the foot when compared to theisometric reference contraction (po0.05). Similarly, the LOA ofFext during stretch was 1.672.11 lower than during purely

isometric muscle action (po0.05). This change of the LOAsignificantly correlated with the shift of the COP (r=0.75,po0.001) as well as with potentiation of Fext (r=0.46, po0.05),knee joint (r=0.46, po0.05), and ankle joint torque (r=0.53,po0.01). Additionally, the COP shift correlated significantly withpotentiation of Fext (r=0.56, po0.01) and ankle joint torque(r=0.74, po0.001) during stretch (Fig. 7).

4.2. Residual force enhancement

We found significantly enhanced Fext as well as joint torques for allinstances in time after muscle stretch. Since joint torques found inthis study are similar to those from single-joint studies (Worrell et al.,2001; Maganaris, 2003; Seiberl et al., 2008), we assume that all jointswere acting with maximum effort. Thus the observed RFE seems toreflect mainly muscular properties. The level of RFE ranged from12.3% at time period FS1 to 4.3% at FS3, which is similar to valuesreported in other studies investigating voluntarily activated humanmuscles (Pinniger and Cresswell, 2007; Seiberl et al., 2008; Tilp et al.,2009). But again, the magnitude of RFE was affected by differing jointkinematics. This was probably caused by the enhanced Fext afterstretch, since when there is more force, soft tissues and seat cushionof the dynamometer are deformed more, leading to slightly moreextended joint positions. As shown for Fext (Fig. 3) we assume that RFEat the joint torques level is most likely also slightly overestimated.However, in contrast with potentiation during stretch, RFE was notaffected by COP movement or changes of the LOA of Fext.

In agreement with the majority of reports in literature (Abbottand Aubert, 1952; Herzog et al., 2006), RFE could be detected aslong as the muscles were voluntarily activated. In contrast, thereare recent findings on voluntarily activated muscles, that did notshow RFE more than 2 s after stretch (Tilp et al., 2009). Althoughour results showed RFE for all time periods analyzed followingstretch, we also found that the magnitude of RFE decreasedsignificantly over time. Tilp et al. (2009) associate such abehaviour with a missing true steady state following stretch.However, to our knowledge there is no study on maximumvoluntarily activated large or medium sized human muscles thatexamined the duration beyond stretch until a steady state isreached, but there is evidence that a steady state could be reachedfor submaximal contractions (Pinniger and Cresswell, 2007).Alternatively, the absence of a long term RFE was suggested tobe due to activation/inhibition properties of large muscles (Seiberlet al., 2008; Tilp et al., 2009), but Seiberl et al. (2008) could notconfirm this assumption when analyzing muscle activation forpurely isometric as well as isometric contractions followingstretch. The latter is in accordance with the results of this study.

5. Conclusion

Over the last decade RFE has been observed for a variety ofmuscle preparations and is therefore thought to be a basic propertyof skeletal muscle. In any case, this is the first investigation todemonstrate that RFE up to 12% is present in multi-joint legextension, which is an important everyday movement pattern. Thusit seems reasonable to account for RFE in models of humanmovement, in motor control studies, and in studies of humanperformance. In order to incorporate the characteristics of RFE intobiomechanical models, further research is needed since everydaymovements are performed with submaximal muscle activations anda variety of ROMs. From a functional point of view the magnitudeand duration of RFE depending on different submaximal activationlevels and stretch amplitudes is therefore of major interest for amore complete description of the force enhancement phenomenon.Furthermore, it would be interesting to investigate if the occurrence

ARTICLE IN PRESS

D. Hahn et al. / Journal of Biomechanics 43 (2010) 1503–15081508

of RFE is beneficial in terms of metabolic cost. This assumption isbased on results of different studies (Oskouei and Herzog, 2005;Hahn et al., 2007), which found that the same magnitude of forcecan be achieved with reduced activation and that enhanced forceswere reached without additional muscle activation, respectively.

Conflict of interest statement

None of the authors has any conflict of interest.

References

Aagaard, P., Simonsen, E.B., Andersen, J.L., Magnusson, S.P., Halkjaer-Kristensen, J.,Dyhre-Poulsen, P., 2000. Neural inhibition during maximal eccentric andconcentric quadriceps contraction: effects of resistance training. Journal ofApplied Physiology 89 (6), 2249–2257.

Abbott, B.C., Aubert, X.M., 1952. The force exerted by active striated muscle duringand after change of length. Journal of Physiology 117, 77–86.

Amiridis, I.G., Martin, A., Morlon, B., Martin, L., Cometti, G., Pousson, M., vanHoecke, J., 1996. Co-activation and tension-regulating phenomena duringisokinetic knee extension in sedentary and highly skilled humans. EuropeanJournal of Applied Physiology and Occupational Physiology 73 (1–2), 149–156.

Babault, N., Pousson, M., Ballay, Y., Van Hoecke, J., 2001. Activation of humanquadriceps femoris during isometric, concentric, and eccentric contractions.Journal of Applied Physiology 91 (6), 2628–2634.

Beltman, J.G., Sargeant, A.J., van Mechelen, W., de Haan, A., 2004. Voluntaryactivation level and muscle fiber recruitment of human quadriceps duringlengthening contractions. Journal of Applied Physiology 97 (2), 619–626.

Charlton, I.W., Tate, P., Smyth, P., Roren, L., 2004. Repeatability of an optimisedlower body model. Gait and Posture 20 (2), 213–221.

Cook, C.S., McDonagh, M.J., 1995. Force responses to controlled stretchesof electrically stimulated human muscle–tendon complex. ExperimentalPhysiology 80 (3), 477–490.

de Leva, P., 1996. Adjustments to Zatsiorsky–Seluyanov’s segment inertiaparameters. Journal of Biomechanics 29 (9), 1223–1230.

De Ruiter, C.J., Didden, W.J.M., Jones, D.A., De Haan, A., 2000. The force–velocityrelationship of human adductor pollicis muscle during stretch and the effectsof fatigue. Journal of Physiology 526 (3), 671–681.

Dudley, G.A., Harris, R.T., Duvoisin, M.R., Hather, B.M., Buchanan, P., 1990. Effect ofvoluntary vs. artificial activation on the relationship of muscle torque to speed.Journal of Applied Physiology 69 (6), 2215–2221.

Edman, K.A., Elzinga, G., Noble, M.I., 1978. Enhancement of mechanicalperformance by stretch during tetanic contractions of vertebrate skeletalmuscle fibres. Journal of Physiology 281, 139–155.

Edman, K.A., Elzinga, G., Noble, M.I., 1982. Residual force enhancement afterstretch of contracting frog single muscle fibers. Journal of General Physiology80 (5), 769–784.

Hahn, D., 2008. Force–length properties of leg extension and their implications forstrength diagnostics. In: Proceedings of the XXVI International Conference onBiomechanics in Sports, Seoul, Korea.

Hahn, D., Schwirtz, A., Huber, A., 2005. Anthropometric standardisation ofmultiarticular leg extension movements: a theoretical study. Isokinetics andExercise Science 13, 95–101.

Hahn, D., Seiberl, W., Amereller, M., 2008. Influence of fibre type distribution onforce enhancement during and after muscle stretch. Journal of Biomechanics41 (Suppl. 1) S142–S142.

Hahn, D., Seiberl, W., Schwirtz, A., 2007. Force enhancement during and followingmuscle stretch of maximal voluntarily activated human quadriceps femoris.European Journal of Applied Physiology 100 (6), 701–709.

Hermens, H.J., Freriks, B., Merletti, R., Hagg, G., Stegeman, D.F., Blok, J., Rau, G.,Disselhorst-Klug, C., 1999. SENIAM 8: European Recommendations for SurfaceElectroMyoGraphy. Roessingh Research and Development, Enschede.

Herzog, W., Lee, E.J., Rassier, D.E., 2006. Residual force enhancement in skeletalmuscle. Journal of Physiology 574 (Pt 3), 635–642.

Herzog, W., Leonard, T.R., 2000. The history dependence of force production inmammalian skeletal muscle following stretch–shortening and shortening–stretch cycles. Journal of Biomechanics 33 (5), 531–542.

Hill, A., 1938. The heat of shortening and dynamic constants of muscle.Proceedings of the Royal Society of London, Series B: Biological Sciences 126,136–195.

Hugh-Jones, P.H., 1947. The effect of limb position in the seated subject on abilityto utilize the maximum contractile force of the limb muscles. Journal ofPhysiology 105, 332–346.

Joumaa, V., Leonard, T.R., Herzog, W., 2008. Residual force enhancement inmyofibrils and sarcomeres. Proceedings of the Royal Society of London, SeriesB: Biological Sciences 275 (1641), 1411–1419.

Joyce, G.C., Rack, P.M., Westbury, D.R., 1969. The mechanical properties of catsoleus muscle during controlled lengthening and shortening movements.Journal of Physiology 204 (2), 461–474.

Katz, B., 1939. The relation between force and speed in muscular contraction.Journal of Physiology 96, 45–64.

Lee, H.D., Herzog, W., 2002. Force enhancement following muscle stretch ofelectrically stimulated and voluntarily activated human adductor pollicis.Journal of Physiology 545 (Pt 1), 321–330.

Maganaris, C.N., 2003. Force–length characteristics of the in vivo humangastrocnemius muscle. Clinical Anatomy 16 (3), 215–223.

Mehta, A., Herzog, W., 2008. Cross-bridge induced force enhancement? Journal ofBiomechanics 41 (7), 1611–1615

Oskouei, A.E., Herzog, W., 2005. Observations on force enhancement insubmaximal voluntary contractions of human adductor pollicis muscle.Journal of Applied Physiology 98 (6), 2087–2095.

Oskouei, A.E., Herzog, W., 2006. Force enhancement at different levels of voluntarycontraction in human adductor pollicis. European Journal of AppliedPhysiology 97 (3), 280–287.

Pinniger, G.J., Cresswell, A.G., 2007. Residual force enhancement after lengtheningis present during submaximal plantar flexion and dorsiflexion actions inhumans. Journal of Applied Physiology 102 (1), 18–25.

Rassier, D.E., Herzog, W., Wakeling, J., Syme, D.A., 2003. Stretch-induced,steady-state force enhancement in single skeletal muscle fibers exceedsthe isometric force at optimum fiber length. Journal of Biomechanics 36 (9),1309–1316.

Seiberl, W., Hahn, D., Kreuzpointner, F., 2008. Sonographic determination offascicle length during and after active muscle stretch in vivo. Journal ofBiomechanics 41 (Suppl. 1), S140–S140.

Spurway, N.C., Watson, H., McMillan, K., Connolly, G., 2000. The effect of strengthtraining on the apparent inhibition of eccentric force production in voluntarilyactivated human quadriceps. European Journal of Applied Physiology andOccupational Physiology 82 (5–6), 374–380.

Sugi, H., 1972. Tension changes during and after stretch in frog muscle fibres.Journal of Physiology 225 (1), 237–253.

Tilp, M., Steib, S., Herzog, W., 2009. Force–time history effects in voluntarycontractions of human tibialis anterior. European Journal of AppliedPhysiology 106 (2), 159–166.

Webber, S., Kriellaars, D., 1997. Neuromuscular factors contributing toin vivo eccentric moment generation. Journal of Applied Physiology 83 (1),40–45.

Westing, S.H., Cresswell, A.G., Thorstensson, A., 1991. Muscle activation duringmaximal voluntary eccentric and concentric knee extension. European Journalof Applied Physiology and Occupational Physiology 62 (2), 104–108.

Westing, S.H., Seger, J.Y., Karlson, E., Ekblom, B., 1988. Eccentric and concentrictorque–velocity characteristics of the quadriceps femoris in man. EuropeanJournal of Applied Physiology and Occupational Physiology 58 (1–2), 100–104.

Worrell, T.W., Karst, G., Adamczyk, D., Moore, R., Stanley, C., Steimel, B., Steimel, S.,2001. Influence of joint position on electromyographic and torque generationduring maximal voluntary isometric contractions of the hamstrings andgluteus maximus muscles. Journal of Orthopaedic and Sports Physical Therapy31 (12), 730–740.

Zatsiorsky, V.M., Aruin, A.S., Selujanow, W.N., 1984. Biomechanik des mens-chlichen Bewegungsapparates. Sportverlag, Berlin.